TEAM FLY
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other
Applications with HTML/OS
by D.M. Silverberg
ISBN: 1930082177
Top Floor Publishing © 2002 (387 pages)
For programming and HTML novices, this guide will have you building Web sites
effectively and affordably.
Companion Web Site
Table of Contents
Advanced Web Sites Made Easy—The Simple Way to Build Databases and
Other Applications with HTML/OS
Preface
Part I - Getting Started
Chapter 1
- Introduction
Chapter 2
- Logging into Your Copy of HTML/OS
Chapter 3
- Your First Program
Chapter 4
- Your First Web Database Program
AM
FL
Y
Part II - Programming Basics
Chapter 5
- Underlays, Inlays, and On-Click Overlays
Chapter 6
- Variables, Conditionals, and Loops
Chapter 7
- Debugging Techniques
Chapter 8
- Building Text Editors
Chapter 9
- Building Login Pages
TE
Chapter 10 - HTML Forms Processing
Part III - Database Programming
Chapter 11 - The Web Database
Chapter 12 - Building Query Pages
Chapter 13 - Building Database Reports
Chapter 14 - Building Database Editors
Chapter 15 - Database Networking
Part IV - E-Commerce Programming
Chapter 16 - Designing E-Commerce Systems
Chapter 17 - Building Product Navigation Pages
Chapter 18 - Building Product Detail Pages
Chapter 19 - Building Shopping Cart Pages
Chapter 20 - Building Checkout Pages
Chapter 21 - Building Back-End Management Systems
Appendix A - HTML/OS Resources
Appendix B - Major HTML Tags
Appendix C - The Next Generation: Web-Based Products
Appendix D - HTML/OS Tag Reference Guide
Index
List of Figures
-1Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Advanced Web Sites Made Easy—The Simple Way
to Build Databases and Other Applications with
HTML/OS
by D.M. Sliverberg
Copyright © 2002 D.M. Silverberg
All rights reserved. Printed in the United States of America. No part of this book may be used or
reproduced in any form or by any method, or stored in an electronic storage system, without prior
written permission of the publisher except in the case of brief quotations in critical articles and
reviews. Copying any part of this book for any purpose other than your own personal use is a
violation of United States copyright law.
SAN#: 299-4550
Top Floor Publishing
8790 W. Colfax #107
Lakewood, CO 80215
Feedback to the author: <[email protected]>
Sales information: <[email protected]>
The Top Floor Publishing Web Site: http://TopFloor.com/
The HTML/OS companion Web site: http://dev.aestiva.com/advanced/
Library of Congress Catalog Card Number: XXXXXXX
ISBN: 1-930082-17-7
This book is sold as is, without warranty of any kind, either express or implied, respecting the
contents of this book, including but not limited to implied warranties for the book's quality,
performance, merchantability, or fitness for any purpose. Neither the author nor Top Floor Publishing
and its dealers and distributors shall be liable to the purchaser or any other person or entity with
respect to liability, loss, or damage caused or alleged to have been caused directly or indirectly by
this book.
04 03 02 6 5 4 3 2 1
Trademark and service mark terms used in this book are the property of the trademark holders. Use
of a term in this book should not be regarded as affecting the validity of any trademark or service
mark.
Here's what readers are saying:
"No other product or web development tool provides the level of flexibility and power provided by
HTML/OS. I'm glad there's now a book on this amazing product."
—Charles J. Cangialosi, VP, Electronic Communications,
Cooper Leder Marketing, www.cooperleder.com
-2-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
"Nothing else comes close to the performance, flexibility and easiness of use of HTML/OS.
Advanced Web Sites Made Easy is simple, clear, and easy to understand with lots of examples and
exercises."
—Yann Rapetto, Chief Information Officer,
Alese Technology, Inc., www.alese.com
"Advanced Web Sites Made Easy is a wonderful tool for connecting the programming dots. The book
covers a good range of topics and areas, is well documented, contains illustrations, good exercises
to implement what you learn and has a lot of meat. There is something for everyone, touching on
topics useful for programmers at different levels and skills."
—Marvin Ellis, President,
www.Equestrian.com
"This first book covering Aestiva HTML/OS provides a long-needed resource and a solid steppingstone to the world of HTML/OS programming. The novice will find grounding in the first four
introductory chapters, and subsequent chapters build the reader's confidence by introducing simple
(and not-so-simple, but easy) tasks, then encouraging learning through experimentation by
presenting possibilities and thorough exercises, to which end the reader receives a 30-day online
trial of the software—ample time to learn to develop sophisticated applications."
—Brigitte Botnick, Owner,
Fauna Art Studios, www.fauna-art.net
"Advanced Web Sites Made Easy is right on target. Not only does it cover the ins and outs of using
HTML/OS but it serves as a great introduction and review of basic programming concepts and
methods. It is invaluable to anyone who wants an advanced web site."
—D. Schuyler Kuhl, Computer Assistance Services,
www.computerassistanceservices.com
Here's what HTML/OS users are saying:
"When I got HTML/OS, I had no knowledge of HTML or programming. Now I have one of the largest
vitamin stores on the Web. And I built it myself."
—David Greenwalt, President,
www.ThePowerStore.com
"Our real-time tracker allows tens of thousands of people to simultaneously monitor runners in our
marathon. Thanks to HTML/OS and its high-speed database engine our Web site and race day
options are number one in the world."
—Alice Schneider, VP Computing,
The New York City Marathon, www.nyrrc.org
"Because of HTML/OS, I've leapfrogged my competitors."
—Curtis Palmer, Web Developer,
www.palmerinternet.com
"I built a content management system for Globix.com, one the backbones of the Internet. I saved
them about a quarter of a million dollars."
—Khanh Vo
"We've been in business for over 100 years. We chose HTML/OS for one reason and one reason
only. It's more powerful, easier to use, and more reliable than anything else on the market."
—Tom Sweet
www.kingarthurflour.com
-3-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
"I built an entire computer retail business powered on the HTML/OS engine and then sold it for over
$10,000,000. I can sincerely say it would not have happened without HTML/OS."
—Bryn Kaufman, Former Owner,
CMP Express
"You could spend hundreds of thousands of dollars on SQL Server 7, ColdFusion and a team of
Web developers. But why do that when you get more power and more reliability with HTML/OS. Plus
you can do the development yourself."
—Don Nicholes, President,
The Diamond Lane
ACKNOWLEDGMENTS
The road leading to this book has been in the making since 1995, when the seeds of HTML/OS were
first sown. Along the way hundreds of friends, colleagues, customers, vendors, and associates have
provided invaluable suggestions, criticisms, ideas, challenges, and labor. I thank everyone who has
offered a suggestion, presented a criticism, volunteered an idea, challenged me, or committed time
and labor to the cause of Web-based computing and HTML/OS.
A special thanks goes to the talented team who helped me create Advanced Web Sites Made Easy.
My heartfelt thanks goes to: Missy Ramey whose executive decisions, artistic creativity and industry
experience made this book magically come together; J.W. (Jerry) Olsen whose impressive
leadership as project manager and development editor made this book exceed expectations; Syd
Jones whose hard work ensured chapters made sense; Pablo Collins, Paul Forsyth, and Matt
Esquivel for their invaluable proofing of the endless number of code examples; Joann Woy for her
indexing work; Susan Woolfolk for proofreading; Steve Winterkorn, Pamela Punzulan, and Don
Nicholes for their help in the construction of the companion Web site for this book.
Finally, I would like to thank Peter Kent whose support and suggestions made this book possible.
ABOUT THE AUTHOR
Dave Silverberg is a founding member of Aestiva, the maker of HTML/OS. Dave is currently the
Director of Technology for Aestiva, overseeing commercial product groups and Web technology
research. Prior to founding Aestiva, Dave managed a consumer-oriented online serviced called LA
ONLINE. He has been involved in online technology development since 1985. Dave has an
undergraduate degree in physics from Clarkson University and a graduate degree in physics from
the University of California, San Diego.
-4-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 1: Introduction
Overview
This book is about bringing sanity to the world of advanced Web-site construction. Now that the dotcom mania is over, it's time to simplify your outlook on building advanced Web sites. It's no longer
acceptable to employ 10 or 20 people to build a business Web site. During the dot-com days money
flowed from well-financed venture capitalists like water down the Nile. But those days are toast.
Today's, advanced Web development has the same goals, but this time it's not rich dot-coms
knocking on the door. Millions of small businesses and fiscally responsible corporations are running
the show.
Making advanced Web development straightforward for the legions of people wanting advanced
Web sites is what Advanced Web Sites Made Easy is all about. Large development staffs are over.
Whether you're a graphics designer, business owner, computer consultant, or programmer, you want
to know how to single-handedly build sites with databases, shopping carts, download and upload
areas, staff areas, membership areas, discussion areas, and sites with high degrees of functionality.
And that's not all. You may also want to build Web-based applications. After all, the next generation
of programs will be on the Web. You might want to build a Web-based sales and contactmanagement system, a document-management and workflow system, an intranet with file
management and collaborative messaging, a Web-based accounting package, a financial reporting
system, a company portal, or any number of industry-specific, Web-based applications.
You might want to build that Web site for yourself, a customer, or your boss. You might want to build
a commercial product and sell it to thousands of people. Or maybe your goal is simply to advance
your Web-development capabilities. Whatever it is you want to do, you need a development
environment that's easy to understand and deployable all across the Web.
Most people believe these kinds of projects are way over their head—even the professionals. You
probably do too. If you're like most people, you've put off building advanced Web sites. It seems
every day you need to learn another new technology. The whole endeavor seems far too
complicated. In fact, I wouldn't be surprised if you're a bit suspicious of the title of this book. After all,
you might wonder how the word "advanced" and the word "easy" could be in the same book title. A
walk down the computer aisle of any bookstore reinforces the notion that advanced Web
construction is complicated. A peek inside any book on the subject will tell you so. A talk with a friend
or colleague would support your conclusion too.
But advanced Web site construction isn't complicated if you're using the right approach. You don't
need to be a highly experienced programmer. You don't need to learn the latest technologies. If you
follow the approach detailed in this book, you'll be able to build sites as advanced and even more
advanced than those dot-com Web developers of yore. You'll be able to satisfy the development
requests of all of those small businesses and fiscally responsible corporations knocking on your door.
In this chapter, I'll explain the overall approach used by HTML/OS, the product described in this book,
and compare it to the approach most people follow. You'll see why HTML/OS simplifies advanced
development. You'll also see why the approach produces sites more powerful than ever. The dotcom days may be over, but advanced Web development has just begun.
The Legacy Beast
Let's start with a misconception: Building advanced Web sites has to be tough. This notion is
accepted as gospel in most Web development circles. Why? Because most Web developers still use
-5-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
the approach popularized by the dot-com era, which indeed is tough. The approach is called systems
integration. It requires you to be well versed in intricate and complex technologies, many of which
can be understood only be those trained in computer science.
The systems-integration approach is about building Web sites by integrating multiple software
packages: a database engine, integration tools, programming tools, and so on. The approach is
based on the belief that it's smart to link legacy (pre-Web) systems directly with newer Web-based
systems. Otherwise, the argument goes, you'd have to throw out the legacy systems, which would
cost you more than integrating them with new, Web-based systems.
But this point of view doesn't take into account the high cost and complexity of systems integration.
The systems-integration road is filled with potholes. System and version incompatibilities, hosting
compatibility problems, the need to learn interface languages and protocols, potential upgrade
nightmares, and the need to learn the intricacies of multiple products are just some of the challenges.
Writing applications in these environments is incredibly complicated. The systems-integration
approach is not elegant, simple, or beautiful. It's not a smart way to build an advanced site. Sites
built this way are typically 50 times the cost, less reliable, and harder to maintain. So much for saving
a couple of bucks. It's the brute force way to build advanced sites, and because Web developers are
told over and over they have no other option, the systems-integration approach is still the dominant
way advanced Web sites are constructed.
The systems-integration approach is so complicated it has single-handedly given rise to dozens of
new technologies and well over a thousand books on various aspects of the subject, all designed to
tame the beast; but the beast has not been tamed. If anything, systems integration has become
more complex as new technologies are introduced on top of, underneath, and along side the others.
It's become a multibillion-dollar business backed by many of the world's largest software
powerhouses. Divide and conquer should be their motto. Thankfully, nobody is forcing you to follow
the systems-integration approach.
A Web Approach
As you may have guessed, this book is not another attempt to simplify the world of systems
integration. It is not a patch or a fix to what should be obvious to most—that systems integration is a
dead end. There is no sense in spending so much money and time saving old systems or integrating
them with the Web. It's a lot easier and less costly for customers to simply transfer their data to new
Web-based systems and ready themselves for a new Web-based world.
This book readies you for the next generation of the Web and, at the same time, frees you from the
systems-integration nightmares that poison Web development. Your future is simple once you leave
the legacy world behind and enter the Web world. Web-based business applications, advanced Web
sites, e-commerce systems, even commercial products are much easier to build and more powerful
when there's no legacy stuff to muck up everything. Using HTML/OS, you can build a customer login
page with five lines of code. You can build a Web-based text editor with 10 lines of code. You can
build a Web-based database editor with 15 lines of code. This book includes such code examples.
When the Web is set free to do its thing, it blows away the old school.
Advanced Web Sites Made Easy disregards systems integration. Words used in the systemsintegration world, such as DCOM, CORBA, .NET, SOAP, SQL, and OLE, aren't solutions. They're
the reasons why everything is so complicated.
Here, you'll learn how to place HTML/OS tags in HTML documents. That's the crux of it. And the
beauty is you won't be any worse off than had you spent months learning the systems-integration
approach. On the contrary, you'll gain capabilities surpassing those using these complex systemsintegration technologies.
You'll find the single-product, Web-centric approach of HTML/OS so easy, you'll be amazed as you
wander your way through discussions of applications you previously thought too tough to build on
-6-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
your own. Armed with only knowledge of HTML tags and some rudimentary programming skills,
you'll find you already have the ability to build sites more advanced than those being built by people
with far more education and experience.
Your first step will be to access your preinstalled copy of HTML/OS. (For those purchasing a copy of
HTML/OS, it's no more difficult. HTML/OS installation is included with purchase.) This step is
described in Chapter 2, Logging into Your Copy of HTML/OS. You'll be provided a URL that
accesses your preinstalled copy of HTML/OS over the Web. Once on that copy of HTML/OS, you'll
have everything you need. You won't need external database engines, because HTML/OS includes
its own high-performance database engine. You don't need integration tools either. It'll just be you,
your HTML documents, and your copy of HTML/OS.
One of the first things you'll discover about HTML/OS is that it's Web based. You don't install
software on your own PC. You work over the Web. When you log into your copy of HTL/OS, you'll
see a desktop with icons. The icons launch Web-based programs you've built and others included
with the desktop for building and maintaining Web-based products. HTML/OS is a Web computer.
You can add icons to your desktop, install products, and write programs, as with a computer, only
everything is Web based.
Another nice feature of HTML/OS is you don't need to be a server administrator or to know anything
about Web servers. You don't need FTP or Telnet either. You can do everything through the
HTML/OS desktop over the Web.
What will really get you excited, however, is the way HTML/OS simplifies advanced Web
development. To drive home the point, consider the case of building a database-driven shopping-cart
system. In the world of Web development, shopping carts are considered advanced applications.
Consider the difference between building one with Perl, perhaps the most popular language on the
Web, and building one with HTML/OS. A basic database-driven shopping-cart system written in Perl
takes 5,000 or more lines of intricate programming spread across dozens of program libraries and
program files. With HTML/OS, you'll need a few dozen lines of easy instructions placed in the halfdozen HTML documents needed for the site. That's a 250-to-1 ratio. You might say advanced Web
development with HTML/OS is more than 250 times as easy as Perl. And that doesn't include the
benefits that come from not having to set up a database server and maintain it, or that HTML/OS is a
lot easier to write.
HTML/OS is simplicity itself. It gives you the ability to go where you haven't been able to go before; it
gives you the ability to do your work faster, and the ability to build sites and complete projects you
would never have otherwise attempted. If you're learning HTML/OS to help with your own business,
HTML/OS means new business opportunities, because with it you can build features you would not
have otherwise built. If you're learning HTML/OS because you want to become a better Web
developer, you'll be able say "yes" to almost anything asked of you. It also means you can singlehandedly compete against systems-integration teams by simply offering your potential customers
custom features they would be reluctant to provide.
Placing HTML/OS Tags in Documents
Developing advanced Web sites is a matter of placing HTML/OS tags inside HTML documents. The
HTML/OS tags are also called Otags, which stands for Overlay tags. You'll see why in a moment. An
example Web page is shown here:
<html>
<title>Test Page</title>
<font size=5><b>My Dynamic Web Page</b></font><br><br>
<font size=4>Did you know...<ul>
-7-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<li> 345 multiplied by 12 is equal to <<345*12>>.
<li> We're <<TIMEFROM("01/01/2000","days")>> days into the new
millenium.
<li> It's <<IF ISWEEKDAY(TODAY) THEN DISPLAY "not" /DISPLAY /IF >>
the weekend.<br>
</ul></font>
</html>
This Web page is a standard HTML document except for the programming that appears in bold.
HTML/OS programming instructions are placed within << and >> characters. These program
segments are called Overlays. The tags within them are called Overlay tags. The page shown here
has three Overlays, each containing a single HTML/OS tag, called an Otag. Building dynamic Web
pages is a matter of adding Overlays to them. The first Overlay does a math calculation. The second
calculates the number of days since Jan 1st, 2000. The last displays the word "not" in the document
if the current day is a weekday.
The Spreadsheet Analogy—Overlay programming is somewhat analogous to spreadsheet
programming. Whereas spreadsheets are composed of one or more spreadsheet pages containing
macros applied to specific cells, HTML/OS Web sites are composed of one or more Web pages
containing macros applied to specific locations within the HTML document.
You can use Overlays to program any part of a Web page. You use them to change the content of
any HTML document sent to a visitor of a site at the instant the page is requested. This ability to
change documents as they're requested by browsers is what separates simple Web sites from
advanced ones. Advanced Web sites, for this reason, are also known as on-the-fly Web sites.
In the current example, three Overlays calculate some numbers and text. Overlays can also
generate HTML tags, links, and any content you want to place dynamically in an HTML document.
You can even use them to change the parameters you pass to Java applets or place Javascripts in
documents or exchange data with Flash applications.
In addition, Overlays can execute in response to a user clicking a hypertext link or a button in an
HTML form. For example, suppose you have a Web page with an order form. When the user clicks
the Complete Order button on the form, you want to add an entry to a database and send out a
confirmation e-mail. To accomplish this, you add an Overlay to the HTML document that includes
instructions for performing these two tasks. You'll learn how to perform these kinds of tasks, and
more, later in this book.
The ability to transform a static Web site into an on-the-fly Web site by merely placing Overlays in
HTML documents, is one of the elegant aspects of HTML/OS. No matter how advanced your site
becomes, the process of creating the site is still a matter of placing Overlays in HTML documents.
Overlays can contain HTML/OS tags that read, write, and search databases. They can send out email. They can perform calculations and display images. They can move files around on the server.
They'll do anything you program them to do. You can define all the dynamic tasks that need to be
performed as a user navigates a Web site by placing instructions in the Overlays you put in your
Web pages.
Once you know the rules for placing Overlays in HTML documents and know how to write the
instructions contained inside them, you can begin building advanced Web sites. A detailed
description of the different ways to place Overlays in documents appears in Chapter 5, Underlays,
Inlays, and On-Click Overlays. Writing instructions is covered in Chapter 6, Working with Variables.
-8-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Simplicity is Power
The single-product, Web-centric approach used by HTML/OS has many benefits. As you'll see in the
sections that follow, HTML/OS is more than an easy-to-program environment. It's not just a pretty
face. It's also fast and well suited to the server environments prevalent on the Web. The baggage of
the systems-integration world has been dropped. You're in the Web world now.
Speed
HTML/OS runs fast. By eliminating the overhead of inter-product communication, speeds are actually
increased. And HTML/OS uses a high-level language, not a low-level language, so code runs faster.
Let's say, for example, you have a database with 500,000 records and you want to write a Web page
that searches the database and displays the first 20 results. When using HTML/OS, this task
requires two tags: DBFIND to search the database and LAYOUT to lay out the results on the page.
That's it. You type these high-level tags directly into the HTML document used to display the page.
Exactly such a page is discussed in Chapter 4, Your First Web Database Program. You can write the
code for the page in a minute. Because the page contains only two tags, most of the processing
occurs inside the HTML/OS engine, a highly efficient, compiled C++ application—the language of
choice for the fasted products in the world. You'll be happy to know HTML/OS renders search results
in about 10 or 20 milliseconds. (A millisecond is one-thousandth of a second). Of course, the transfer
of the document from the server to your browser will take a lot more time. Since HTML/OS is so fast,
chances are you won't feel the time HTML/OS takes to perform the search.
Compare this to systems integration, in which case the program needs to compose a database
message and pass it to the remote database. Then it must wait for the reply. Sockets need to be
opened and closed. Transaction limits may need to be tested. Database drivers need to be accessed.
(Separate database engines need to be set up, tested, and maintained too.) Multiple messages may
need to be passed to the database server, so this process might have to be repeated. The overall
speed of the page depends on many factors. The efficiency of a driver, the speed of the external
database engine, how well the database programming is written, and the efficiency of the integration
product selected, are all important factors. The systems-integration approach can take 100 times as
long.
Unlimited Transactions
The HTML/OS database engine has no built-in transaction limits. Most database servers limit the
number of users at the same time. The limitations stop you from taking full advantage of the
capabilities of your hardware, unless you purchase special licenses often costing more than the
computers on which the database servers run. HTML/OS has no built-in transaction limits. You're
limited only by the capacity of your hardware.
Web Compatibility and Portability
Another advantage of HTML/OS is its superior Web-compatibility and portability. Web-compatibility is
the ability to run a Web application in the various environments popular on the Web. Not only must
an application be compatible with the various browsers on the market, it should also be compatible
with the heterogeneous array of server environments. Most Web sites run on servers using some
version of Unix, Linux, or Windows. Many hosting accounts are shared. On a practical note, many
server administrators don't give their hosting clients the ability to reconfigure their servers to fit their
specific needs. All of this leads to problems for organizations using multiproduct solutions.
Web sites built with systems integration rarely run at shared hosting accounts except when the
hosting service is the one providing the products you must integrate to. The simpler single-product
-9-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
approach used by HTML/OS suffers no such limitation. HTML/OS applications run in shared hosting
accounts.
Product Sharing—Some hosting services provide you with database engines and integration
products you can share with other customers. While this offers a temporary solution you can use
until you get your own database engine, the solution locks you into the hosting service. The large
amount of systems integration needed to link to the various products installed at the hosting
provider often creates an insurmountable cost if you decide to move your site later. In addition,
hosts reserve the right to upgrade or suspend the maintenance of those products at any time.
Furthermore, shared products can be adversely affected by others using them. This puts your site
at constant risk. It's best to control your own server-side software and build sites so they can be
moved, if necessary.
Portability is also a concern. Portability, means the ability to move a Web site from one kind of
hardware or operating system to another. The systems-integration approach requires the matching
of hardware and software components. Each software component must be compatible with the other
and each must be compatible with the hardware. The result is a system of hardware and software
that is best left alone. The Web site is difficult to move, because changes in hardware and operating
systems require the intervention of experts, a change in software components, and most often, a
reprogramming of the Web site. The HTML/OS single-product approach frees the owner of a Web
site from the loss of Web compatibility and the loss of portability. HTML/OS sites run in standard
hosting accounts. No special configuration requirements are placed on the server administrator. The
servers can use almost any kind of server, such as Unix, Linux, Windows, and MacOS X.
Moving an HTML/OS site from one setting to another is much like moving Web sites with no
programming or dynamic elements. It's a matter of copying the files, documents, and databases in
the site to a new hosting account. The only difference is that you need to install an HTML/OS engine
on the new site. This high level of portability and cross-platform compatibility are not possible when
Web sites are built from multiple, integrated products.
Server-Configuration Independence
The HTML/OS engine is not only portable and cross-platform. It's also immune to serverconfiguration dependence. Applications built with HTML/OS have no connection to the underlying
server. As a result, you can install applications with the point-and-click method. For example,
suppose you take an HTML/OS application, pack it up, and move it to some other hosting account
with an HTML/OS engine. When you unpack the application, it would run the same way. The two
servers could be completely different. Their internal file directories, hardware, and operating system
could be different. One server could be Windows. The other could be MacOS X.
This ability to move HTML/OS applications between different kinds of servers, without modification,
gives you the ability to build and maintain a single application but sell it to people on Unix, Linux, or
Windows server platforms. The topic of building commercial products with HTML/OS is discussed in
Part VI, Building Commercial Products.
Web Networking
The single-product approach used in this book means you'll be able to enjoy some capabilities not
available before—not even to those in the systems-integration world. One of the new things you'll be
able to do is take advantage of Web networking, which is a technology that allows you to network
multiple copies of HTML/OS. Unlike traditional networking, Web networking implies the nodes of the
network (a copy of HTML/OS is called a node) can be anywhere on the Web and use any kind of
hardware or operating system. In other words, Web networking is wide-area and cross-platform.
-10-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
HTML/OS allows you to set up Web networks. The capability is built in. Once you set up a Web
network, files and databases on one Web site can be accessed from another site by preceding the
name of files and databases with node names. For example, if site A, located on a Linux system in
India, has a product database called /db/products, site B, located on a Sun Solaris system in
Texas, could access it by using the name A:/db/products in the HTML documents on their site.
You set up Web networks by using the Control Panel included with each copy of HTML/OS. The
Control Panel gives you the ability to define levels of security, set a node identity for that copy of
HTML/OS, and define links to other nodes in your Web network. Setup is point and click.
Web networking is appropriate in a variety of situations. As an example, consider an international
company with multiple copies of HTML/OS spread throughout the world. If the human resources
department needs access to data stored on a copy of HTML/OS used by the sales department, the
human resources users can use Web networking to access the data. As another example, a
company may want to share data between its intranet and Web site. Web networking provides an
easy solution.
Distributed Execution
AM
FL
Y
With HTML/OS, you can take advantage of the automatic way in which applications can be
distributed across multiple servers. For example, suppose you have an existing Web-based ordering
system you want to modify. You would like to place the order page on a new Web site and leave the
rest of the Web site on the original server. You want to do this because you're developing an intranet
on the new Web site that will be taking over the ordering function. A developer using systems
integration would have a tough time adapting to the new requirements. Distributed execution is
extremely complicated in multiproduct environments, even for the experts. A developer thinking
about satisfying such a request might say the task is impossible, unrealistic, or too much trouble, or
recommend reprogramming the site from scratch.
TE
Distributed execution is built into the HTML/OS engine. When you spread documents between
different servers, they continue to run as if it they were placed on a single server, because HTML/OS
is a distributed state-persistent system, a fancy way of saying HTML/OS doesn't loose its variables
as users on a site move from page to page, even when the pages are on different servers.
Distributed execution gives you, as the HTML/OS developer, the ability to spread applications across
multiple servers when one server can't take the volume of hits, or is simply not convenient.
Summary
In this chapter, we've reviewed the major differences between the systems-integration approach to
building Web sites and the single-product approach taken in this book. As you have seen, the topic
of advanced Web sites has many facets; but the construction process discussed in this book is
simple and powerful. Now you know why this book is titled Advanced Web Sites Made Easy.
In the next chapter, you'll get on a copy of HTML/OS. You'll use its Web-based development
environment. You'll review the tools available to you and begin experiencing for yourself the beauty
of advanced Web development.
-11Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 2: Logging into Your Copy of HTML/OS
As you use this book, you will want HTML/OS in front of you. This chapter tells you how to log on
and run your personal copy of HTML/OS. Then, the chapter acquaints you with the HTML/OS File
Manager and Web-based text editor, two of the applications reviewed in this chapter that you will use
throughout this book. You'll get an overview of the HTML/OS desktop, take a look at the File
Manager, a tool for navigating the files and directories, and learn to use the Web-based text editor.
After that, you'll build some simple Web pages and learn to provide access to them. The chapter also
describes the HTML/OS Control Panel and other programs included with your copy of HTML/OS.
Let's Get Started
Advanced Web Sites Made Easy has reserved a copy of HTML/OS just for you. The copy runs on
the Web, so there's nothing to install on your own PC. All you need is a connection to the Web and
the HTML/OS registration number printed on the HTML/OS Free-Trial Card in the back of your book.
Follow these instructions to activate your copy of HTML/OS:
1.
2.
3.
4.
Open up a browser.
Go to the HTML/OS Free-Trial Web page at http://dev.aestiva.com/freetrial.
Click the Free Trial Account Signup button.
Enter the 10-digit reservation number on your HTML/OS Free-Trial Card along with
your e-mail address and any optional information requested.
Seconds after you sign up a new-account letter will be e-mailed to you. It will contain a URL and
password. Pick up your e-mail and use the URL and password to access your copy of HTML/OS.
The URL goes directly to the login screen of your preinstalled copy of HTML/OS. You're now ready
to work online.
Usage Limitations—You can use your copy of HTML/OS as a place to sharpen your design and
development skills and as a demonstration site for existing and potential customers. But you
cannot use the copy to host a Web site. It's strictly an Aestiva Developer account, a limited hosting
account with no FTP or Telnet, no e-mail accounts, and 100 MB of hard disk space. See your signup agreement at http://dev.aestiva.com/freetrial/policies for details, limitations, and usage policies.
You can access your copy of HTML/OS from anywhere. Share it with friends or use it to show off
your HTML/OS applications. Your copy of HTML/OS will be available to you for a period of 30 days
from the date you first log in. If after 30 days you wish to extend the hosting of your free copy, you
can purchase additional 6-month blocks for a nominal charge of $25.
-12-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.1: This is Your HTML/OS Login Screen.
First Looks at Your Copy of HTML/OS
To log into your copy of HTML/OS go to the URL provided in your new-account e-mail and enter your
password. Then press Enter. You will see the HTML/OS desktop as shown in Figure 2.2. The
desktop is your personal administrative area. It somewhat resembles a computer desktop, but of
course, it runs in a browser. In the next section, you'll read about the HTML/OS desktop and learn
how to use it.
Figure 2.2: The HTML/OS Desktop…
-13-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Your HTML/OS Desktop
Your HTML/OS desktop has two horizontal options bars and a set of application icons. The top
options bar includes the options Menu, New, File Manager, Control Panel, Preferences, and
Logout. The bottom options bar includes a list of application menus. When you first log in, you're
placed on the Main menu.
Your HTML/OS desktop is where you place applications and run them. Use it to organize
applications and store bookmarks to other Web sites. The desktop also comes with a number of
useful utilities and programs preinstalled that help you, the Web developer. We'll discuss them later
in this chapter.
The Top Options Bar
Your top options bar contains fixed links and programmable links you can add and edit. The fixed
links are as follows. They're discussed individually in the sections that follow.
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
Menu
New
File Manager
Control Panel
Preferences
Help
Logout
Menu
The Menu link gives you the ability to add or delete menus from the menu bar. A menu is a set of
icons that launch applications or locations on the Web.
New
The New link is for creating new icons. When you click New, a small window will pop up allowing you
to create a new icon in the menu you are viewing. An example is shown in Figure 2.3. Note that
icons can be links to applications or bookmarks.
-14-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.3: Creating a New Icon or Bookmark.
To set up an application icon, enter a title for your icon and the full name of an HTML document to
run. Click Select if you need to browse your file system. Note that every application icon has an
entry HTML document that runs when it's clicked. This is called the application Launch document.
To set up an icon to a remote Web location, in other words a bookmark, enter a title for your icon and
the full URL of the Web site. Click the Create icon button at the bottom of the page when you are
done.
The New window also includes options to position your icon relative to the others on your menu,
select an image for your icon, and determine how it should be launched. An icon can launch in a new
browser window or in the same window, displacing your desktop when it runs. If your computer has
enough memory and is fast, you will want icons to launch in new windows.
Just below the Launch New Window option is the Security drop-down list. It should be set to
Desktop Only unless you want the application to be accessible to others on the Web. In this case,
set security to Allow Public Access and click Start Link to view the URL you can give out to
anyone wishing to launch the application from the Web
File Manager
The File Manager link gives you the ability to navigate the directories in your system, edit and run
HTML documents, perform uploads and downloads, and perform file-management tasks. A
description of the File Manager is provided in the "File Manager" section later in this chapter.
Control Panel
You use the Control Panel link to install applications, set up networking to other copies of HTML/OS,
configure e-mail settings, and set up system-wide settings used by HTML/OS. The Control Panel is
also described in the "Building Your First Web Pages" section later in this chapter.
-15-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Preferences
The Preferences window allows you to change the appearance of your HTML/OS desktop to fit your
individual tastes. The window is shown in Figure 2.4. It contains a variety of settings, from
determining how to display the titles of icons on your screen to setting screen colors. The settings
are summarized in the following list:
Figure 2.4: You Use the Preferences Window to Change Desktop Preferences.
ƒ
ƒ
Background Themes— Use this to set up different desktop backgrounds and
color combinations. Select from themes already installed or create your own.
Shortcuts— You can set up these special links to launch specific applications.
Shortcuts are positioned between the File Manager and Control Panel on your top
options bar. To modify your shortcuts, click Select and choose one or more icons
from those installed. Note that shortcuts must relate to an existing icon. Use
shortcuts only for highly used applications, because the top options bar is visible,
regardless of which menu you are in.
ƒ
Launch Windows— Use this setting to choose how windows are launched.
HTML/OS launches windows in either of two ways. They can be launched as a new
browser window or as an adjustable JavaScript window. JavaScript windows give
you greater control over the size of the window launched; but sometimes,
depending on the computer you use, they can open slowly. HTML/OS is delivered
with JavaScript windows off. Feel free to change this setting to see what works best
for you.
ƒ
Font and Font Size— These two settings allow you to change the type of font and
font sizes used in icon titles.
ƒ
Maximum Rows and Maximum Columns— These two settings give you the
ability to organize how icons appear on a menu. Use the settings to ensure icons
do not fall off your desktop.
-16-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Help
Help launches an on-line Help window with information about how to use the HTML/OS desktop.
Refer to this pop-up facility when you have questions.
Logout
Click Logout to log out of your HTML/OS desktop.
The Bottom Options Bar
Your bottom options bar contains the names of menus of icons. Menus give you a way to organize
the bookmarks and applications you use. Create and delete menus as needed. When you first log
into HTML/OS, you'll see it comes with the following four menus preinstalled:
ƒ
Main Menu— Place general-purpose applications here.
ƒ
Utilities— Place utilities here.
ƒ
My Apps— Place the applications you create here.
ƒ
My Websites— Place bookmarks to your popular Web sites here.
Each menu on your bottom options bar is marked by a small globe. The globes preceding your
current menu selection serve a function. Click them to toggle your display between a listing of icons
and a text-based display of listings.
Your Icon Marker
Each icon or listing in your display is preceded by a globe or a diamond. They have meaning. Globes
tell you the icon is a bookmark, in other words, a link to a Web site. Diamonds tell you the icon is an
application, in other words, a link to an HTML document.
Diamonds can be red or orange. Red means the public can access the application with a Start-link.
Start-links are explained in the "Start-links" note later in this chapter. The red color warns you the
application is accessible by the public. All other application icons are orange.
Icon Preferences
The Icon Marker is also a link. Click it to edit the preferences for the icon. A window, like that shown
in Figure 2.5, will open.
-17-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.5: In the Edit Icon Window You Can Delete or Move Icons.
The Edit Icon screen is similar to the screen you see when you click New on the top options bar on
the desktop. It includes the same options as the New window with additional options to delete the
icon or move it to another menu. The New Icon screen was described in the "Top Options Bar"
section earlier in this chapter.
Desktop Applications
Your HTML/OS desktop includes useful applications for managing files, editing and running HTML
documents, packing and unpacking files, creating and managing databases, bundling products, and
changing the settings used by HTML/OS. The two applications you'll be using the most throughout
this book are the File Manager and Web-based editor. To a lesser degree, you'll be using the Control
Panel, discussed later in this chapter, and dbConsole, discussed in Chapter 11, What Is a Web
Database?
In preparation for creating your first Web page, let's review how to use the HTML/OS File Manager
and its Web-based text editor. We'll also be discussing Start-links and the Web File System (WFS)
used by HTML/OS. If you find these a bit confusing, don't worry. They're not essential at this point.
They will make more sense later in this chapter when the topic comes up again in the "Giving Access
to Your Three-Page Application" section.
File Manager
The HTML/OS File Manager is a file-management utility that allows you to select, copy, rename,
create, delete, upload, download, and list files on your Web site. You can launch it by clicking File
Manager on the top options bar on your desktop. The File Manager also includes an online help
utility, a search utility, and a Web-based text editor. (See Figure 2.6.)
-18-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.6: From the File Manager, You Can Access Documents on Your Web Site, Edit Them, And Run
Them.
Using the File Manager, you can perform tasks on single files or entire groups of files. The top
options bar has links for creating new files and directories, uploading files, and finding files. You'll
also find a link to an online help utility and a link to show or hide options. When options are hidden,
the options area, just below the options bar, is not shown on the screen.
The options area includes a New Location box and buttons to copy, move, download, and obtain
information on selected files or directories. Buttons in the options area will perform tasks on one file
or on all the files and directories you check.
Your current directory path is listed just below the options area. Below that is your file and directory
list. To copy, move, download, or obtain information on one or more items, place a checkmark
beside the items and click the button for the action you want to perform. To create a new file or
directories or upload a file, use the appropriate link in the options bar.
Multi-File Tasks
As mentioned in the preceding section, the File Manager can copy, move, download, and display
information on multiple files and directories. For example, if you wanted to copy or move some files
and directories into the directory /mywork, you would place a checkmark beside the files and
directories you wanted to copy or move, type /mywork in the New Location box, and click the Copy
or Move button. All the files and directories would copy or move across. Note that the File Manager,
when it copies or moves directories, copies or moves the entire directory tree, meaning all files and
directories inside the specified directory, and all files and directories inside that, and so on.
If you need to download or display information on multiple files, place a checkmark beside each file,
and click Download or Info. When downloading multiple files, a download list will be provided to you.
Click each name you want to download. When getting information on multiple files, each information
section will be listed, one after the next, on the same screen.
-19-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
File List Sorting
At the top of your file list, you'll see headings for each filename, size, and last date modified. Click
these column headers to sort files by filename, size, or date modified. Click again to reverse the sort
order. Note that directories always appear above files regardless of how you sort your file list.
Navigation
The File Manager gives you access to the files in your Web site. The root directory of your Web site
is called Home or /. It can contain files and other directories, which can contain other files and
directories, and so on. To view the contents of a directory, click the directory name on your screen.
To exit a directory, click any directory link in the current directory path.
File Viewing and Editing
When directories and files are listed in the File Manager, each is preceded with an icon. Click the
icon to display the document in the browser. Click the document name to open it in the HTML/OS
File Manager's Web-based editor. The icon preceding a filename tells you the kind of file it is.
Directories are preceded with an icon that looks like a folder. Other icons tell you whether the file is
an HTML document, a text file, an image, a database, a Pack file, a bundled application, or an
unknown file type.
Information
To retrieve information on one or more files and directories, select them and click the Info button in
the options area. You'll see a screen similar to that shown in Figure 2.7. The following information is
provided for each file requested:
Figure 2.7: You Can View File Information From the File Manager.
-20-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
File Path—The full path to your selected file or directory. The file path is the full
location and name of the document. For example, the full path of index.html is
/index.html. The full path of a file called cart.html located in the directory apps is
/apps/cart.html.
ƒ
File Size—The size of your file.
ƒ
Type of Document—A document type, such as image or database. The Type of
Document tells you whether HTML/OS recognizes your file as an image, HTML
document, or a special file.
ƒ
Last Modified—The date and time the file was last changed.
ƒ
File Area—The internal storage area of the file (Public, Private, or Mirror). The File
area tells you where your file is located in the Public or Private tree of the server.
See explanation in the accompanying "Internal Files Areas" note.
ƒ
Internal File Areas—Web sites use a Web File System (WFS) that's a
subdirectory of the server's file system. When using HTML/OS from the Web, you
don't need to think about the underlying directory structure of the server. However,
if you work directly on the server, you should know that a single directory in
HTML/OS has two directories on the server. One is in the Public tree and the other
is in the Private tree. Files are stored in either side, but directories are mirrored in
both sides. The Public tree is usually the server's DocumentRoot (where the
index.html file is located). The Private tree is a directory outside the DocumentRoot.
When HTML/OS creates a new directory, two directories on the server are created,
one on the public side and one on the private side. You can move documents
between these two directories by using the Info option. A conventional Web site
has only a single directory tree, called the DocumentRoot, which causes file
security problems, because all files are, by default, open and public. In HTML/OS,
most files are private. Specifically, if the extension used by a file is in the HTML/OS
Private Extensions List, which is defined in the Control Panel, it is private.
Otherwise it is public. All database files, text files, and HTML files are, by default,
private. GIFs and JPEGs are, by default, public. A technical description of the WFS
can be found in the Knowledge Base on Aestiva's Web site.
ƒ
Start-link Access—A security setting (Allowed or Not Allowed). Start-link Access
tells you whether the public has access to your Web page. Start-link URL is the
URL users would use to launch HTML/OS and run the page. See the
accompanying "Start-links" note.
ƒ
Start-links—Start-links are URLs that can be placed in static Web pages or used
over the Web. They launch the HTML/OS engine and run a specified Web page.
When using HTML/OS on a Web site, whether for the entire site or only an
application in a site, a Start-link accesses the first page in HTML/OS. For security
reasons, Start-links cannot be used until they're allowed in the Control Panel.
ƒ
Start Link—The URL to run the document (if Start-link Access is allowed).
TE
AM
FL
Y
ƒ
Find It
-21Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The File Manager's options bar includes the Find It link. Click this to find a file. Multiple search
options are provided. Searches can be case sensitive; they can span the content of all files, or only
filenames, and you can restrict searches to only files with specific extensions.
Help
Click the Help link on File Manager's options bar to access the File Manager's on-line Help system.
Use Help to find detailed information on using any options in the File Manager. Explanations of the
HTML/OS directory structure and how to use the Web-editor are included.
Web-Based Editor
The Web-based editor used by the File Manager is a text editor for your HTML documents and any
kind of text document. The Web-based editor is launched when you click a filename. It has options to
save, reload, and run your Web page. If you're working over the Web, you'll be using this editor for
much of your work.
Text Box Resizing
Before using the Web-based editor, you'll want to adjust it for your computer screen. To resize the
text area, click the Resize icon at the bottom right under the text box. A window will appear giving
you options to change the width and height of the text area. (See Figure 2.8.) The following table
shows the recommended settings:
Figure 2.8: Your Web-Editor is Used to Edit HTML Documents and Try Them Out.
Monitor Size
Recommended Columns and Rows Settings
640 X 460
Width 73, Height 12
800 X 600
Width 92, Height 20
1024 X 768
Width 120, Height 20
-22-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Reverting to Previous Pages
To revert to a previous version of a page you have been working on, take advantage of your browser.
Use the Back button of your browser or your browser's History feature to find the page you want.
When you find it, click the Web-based editor's Save button (not your browser's Save feature). If you
simply want to revert to the last saved version of a page, click the Web-based editor's Reload button.
Your text area will be filled with the previous version, overwriting all changes.
Running Your Page
The HTML/OS Web-based editor includes a Save/View button. Use this to run your page. Clicking
the Save/View button saves your data and runs it. Click the Back button in your browser to return to
your editor after running your page.The Save/View button comes in handy as you write and test the
documents you create. The ability to make a change to a page, test it, and return to edit the page in
just a few clicks makes programming a snap. Note that your page is saved before it is viewed (run),
because until it's saved, it's stored in your browser, not on the server, where it runs.
If you want to run your page in one window and edit it in another, do the following: Edit the
preferences for your File Manager's icon. Set it to launch as a new window. Then Click the File
Manager icon twice to launch two different windows. Use one window to edit and one to view.
Also note that you can view (run) any HTML document from the File Manager by clicking the icon to
the left of the document's name.
Quitting the Web-Based Editor
Use the HTML/OS Save/Exit button to save what you're working on and return to the File Manager.
Click Quit to quit what you're working on and return to the File Manager, discarding changes made
since the last save or save/exit.
Building Your First Web Pages
Now that you've seen how to use the File Manager and Web-based editor, you can build your first
Web pages. You'll create three simple Web pages that do little other than link to each other. Building
these pages is a good way to learn your way around HTML/OS. Think of these pages as a simple
Web application. Later you'll add a Start-link to the first page so the set of pages can be accessed
from the Web
.
Click File Manager and click New. It's a good idea to set up a directory to work in. Enter the
directory name /work and click the New Dir button. Click New again, enter /work/first.html,
and click the New File button. You're ready to create your first page. Enter the following code into
your page:
<html>
<title>My First Page</title>
THIS IS MY FIRST PAGE<br>
<a href=two.html>My Second Page</a><br>
<a href=three.html>My Third Page</a>
</html>
-23-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Click Save to save your work. Then click Save As and enter the filename /work/two.html and
click Save again. This will create a new file using your first file as a starting point. Edit the document
so it looks as follows:
<html>
<title>My Second Page</title>
THIS IS MY SECOND PAGE<br>
<a href=first.html>My First Page</a><br>
<a href=three.html>My Third Page</a>
</html>
Click Save. Then click Save As, enter the filename /work/three.html and click Save. Edit the
document so it looks as follows:
<html>
<title>My Third Page</title>
THIS IS MY THIRD PAGE<br>
<a href=first.html>My First Page</a><br>
<a href=two.html>My Second Page</a>
</html>
Click Save/Exit. You have now created three Web pages that link to each other in your /work
directory. To run the first page, click the icon to the left of first.html. Test the pages. Go back
and edit if you've made some mistakes.
Giving Access to Your Three-Page Application
Think of your three Web pages as an application. After all, on the Web that's exactly what
applications are—sets of HTML documents. The entrance page of your application, the first page
entered by users, is first.html. This page has links to two.html and three.html. Although
your three Web pages are accessible from the HTML/OS desktop, they are not accessible by the
public. But what if you want to attach these three pages to your site? To accomplish this, you need to
give access to the entrance of your application with the Control Panel.
Click the link back to your desktop. On the top options bar, click the link that launches the Control
Panel. In the Control Panel, select Security. A number of options will appear under Security. Select
Start-link Allow. You'll see an input area like that shown in Figure 2.9.
-24-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.9: Start-Links are Allowed using the Control Panel.
Type the entry /work/first.html on a new line and click Save Settings. Now the page is
accessible with a Start-link. As mentioned in the "Start-links" note earlier in this chapter, a Start-link
is a URL anyone on the Web can use to access the /work/first.html page. The URL starts
HTML/OS and runs the page. There are two ways to determine the URL. First, you can look at the
help screen provided in the Control Panel. Click Turn Help On to do that. I can't tell you exactly how
to write a Start-link in this book, because Start-links depend on how your copy of HTML/OS was
installed. On the other hand, your Help screen knows. It shows you exactly how Start-links are
written for your installation.
Another way to find your Start-link is to use the File Info option in the File Manager. Exit the Control
Panel and click File Manager on the HTML/OS desktop. Find /work/first.html, place a
checkmark beside it, and click Info. The Properties screen for the document will give you the Startlink for the page. To test it, enter the URL in the location line of your browser.
Now that you know how to build Web pages and give users access to them, how about setting up an
icon on your desktop for it too? To do that, exit to your desktop and select New on the top options
bar. Enter a title for your icon and enter the name of your launch document—the document that runs
when the icon is clicked. In this case, it would be /work/first.html. You can type it into the box
manually or use the Select option to navigate your Web site and find it. At this point, you can click
the Create Icon button. If you want to use a different icon as the default, then before clicking Create
Icon, click the Select link located to the right of the Icon entry box. Select an icon and then click
Create Icon. Now you have a new icon that, when clicked, launches your three-page application in a
new window.
The Control Panel
Your Control Panel is where you install new applications, change or view your system settings, set
up networking, and manage your copy of HTML/OS. Control Panel options are organized as follows:
ƒ
E-mail— Configure outgoing e-mail settings and defaults
ƒ
Install— Install, register, and uninstall products
ƒ
System— Change or view system settings
-25-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
Security— Set up access and database security settings
ƒ
Network— Configure Web networking
ƒ
Database— Access special files and logs used by databases
You can open or close each category by clicking it. Open categories are followed by a list of options.
The options are summarized in the following sections. The Control Panel is shown in Figure 2.9.
E-mail
This is where you can test and configure your outgoing e-mail defaults. Autoresponders and
HTML/OS mail tags will not send out e-mail until the outgoing mail server is configured here. You
also configure related e-mail settings here.
Install
HTML/OS applications can be bundled, using a program called Bundle Bee, into application files
known as bb files (pronounced bee-bee). Once bundled, an application can install by using pointand-click into any copy of HTML/OS, regardless of the hardware it is on. This category is for
installing and managing these bundled HTML/OS applications. The category includes the following
four options:
ƒ
Install Product— Use this to install bundled applications, also known as bb files.
To install, select a highlighted product from your file list and install. Products may
be installed in any directory of your choice. In general, it is best to use the install
directory provided by the product or a subdirectory of /apps.
ƒ
Registration— If your product needs to be registered, type in the product code and
registration key provided by the vendor of your product here. Make sure your
product vendor gives you both a product code and key. Both are necessary.
ƒ
Product Uninstall— Use this to uninstall a product. Use it carefully. This option will
uninstall icons on your desktop and will empty all files and subdirectories in the
install directory of the application.
ƒ
History— This screen allows you to review products you have installed or
uninstalled on your Web site.
System
Over 15 system-wide settings are stored in this section—the most important being the password to
your HTML/OS desktop! Step through each, with Help turned on, to learn about each system-wide
option.
Security
Unlike standard Web sites, where pages are accessible from the Web until you make them
inaccessible, HTML/OS keeps everything inaccessible from the Web until you explicitly make it
accessible from the Web or from a page (by virtue of typing a hypertext link or an HTML form) that is
already accessible. This Security category contains access lists for Start-links and variables, secure
server (HTTPS) access information, settings related to Web-network security, and other security
measures.
Network
-26-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
This category has options for setting your node (a node is a copy of HTML/OS) identity, and
configuring the nodes that can access you. The HTML/OS network is a peer-style, wide-area
network. There is no central point of failure. Every node has full control over whom it can see and
who can see it. This category also includes a network error log and testing utilities that come in
handy when setting up Web networks.
Database
This category contains miscellaneous database options. An error log that logs problem database
requests and an option to set up database handles (aliases to databases) is provided. For further
information on database error logging or database handles, see the on-line help included in this
section or the knowledge base on Aestiva's Web site.
Other Applications
At first, you'll see many HTML/OS applications that are unfamiliar to you. The following sections
provide a brief introduction to the more popular applications. Use the pop-up help system included
with each application in addition to the cross-references provided here to educate yourself further
about these products.
Bundle Bee
Bundle Bee is an application for bundling commercial products. It can take a series of Web pages
and convert them into a single, compressed, copy-protected application (a bb file), which can be sold
and installed on any copy of HTML/OS. The application is discussed in Chapter 35, Using Bundle
Bee.
dbConsole
dbConsole is an application used to create databases and import data into them. If you're a
developer, you'll find this program invaluable. dbConsole is discussed in Chapter 11, What Is a Web
Database?
Packit 2.0
If you've heard of Stuffit on the Macintosh, Zip on Windows, or Tar on Unix or Linux, then you
already know what Packit does. This utility allows you to pack multiple files into a single file for
storage or transfer. Packit 2.0 is shown in Figure 2.10.
-27-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 2.10: Packit Backs up Files and Directories or Transports Large Sets of Documents between
Copies of HTML/OS.
Packing files is accomplished through an easy-to-use wizard. First, you select a directory from which
to extract files. All files in your Pack file will be specified relative to this directory. Then you select the
files and subdirectories you want to place in your Pack file. Then you pack it. Like Stuffit, Zip, and
Tar files, Pack files can be e-mailed and copied from one location of the Web to another. They can
be unpacked on any copy of HTML/OS. Note that HTML/OS applications are server-independent.
HTML documents packed on one system can be unpacked and run on another.
Note for Packit 1.0 Users—Packit 1.0 provided a mechanism to attach desktop icons to a pack file.
This option is not supported in Packit 2.0. Instead, the option is available in Bundle Bee, a
complete application-packaging tool that's included with HTML/OS. Bundle Bee is described in
Chapter 35.
Aestiva Shell
Aestiva Shell is a Web-based version of the shells that have become the primary command-line
interface for Unix and Linux systems. If you prefer command- line interfaces to point-and-click
interfaces, you'll enjoy this shell.
Aestiva Shell supports many of the features found in Unix and Linux shells including the following:
ƒ
ƒ
ƒ
ƒ
ƒ
On-line Help
Redirection and pipes
Command aliasing
Program execution
Command history
-28-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Aestiva Shell supports over 20 popular commands such as ls, cat, diff, cp, mv, mvdir,
rm, rmdir, and help. It also includes a text editor (edit) and gives you ability to run HTML/OS
programs. The Aestiva Shell is shown in Figure 2.11.
Figure 2.11: Aestiva Shell is an Optional Command-Line Interface for those Who Want to Keep Their
Fingers on the Keyboard.
Tracer
Aestiva Tracer is a utility that works with the HTML/OS programming tag, TRACE. The application is
described in Chapter 7, Debugging Techniques. Tracer allows you to analyze trace files filled with
history information accumulated during program execution.
WebMail
WebMail is a Web-based e-mail program. The full-featured application can read e-mail from multiple
e-mail boxes, direct e-mail into specific folders (filters), search your e-mail, and maintain a list of email addresses.
Using Freeware
In addition to the applications listed in the preceding sections, numerous freeware applications are
available. Aestiva Freeware Library, which is located in Aestiva's User Center
(http://www.aestiva.com/support) features dozens of applications that can be downloaded and
installed using the Install Product option in your Control Panel.
Exercises
Exercises 1 and 2 test your knowledge of the desktop and using Packit 2.0. Exercises 3 and 4 are
about variable passing —a topic related to the three-page application you created earlier in this
chapter. These latter two exercises show you how variables work with those Web pages and how
they pass between pages or can be passed into them from the Web. Answers to all exercises are
provided on this book's companion Web site as described in the book's Preface.
Exercise 1
Create a desktop menu called mywork. Move any icons you've created into this menu.
Exercise 2
Pack up the /mywork directory into a Pack file. Download it and save to your local computer. Then
take the file and upload it back to your copy of HTML/OS. Unpack it into the directory /mywork2 and
access it from the File Manager. Then create a new icon that launches the page
/mywork2/first.html.
-29-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 3
This exercise is about passing variables between pages. Take the three-page application you
created in this chapter. Place the instructions just below the <html> tag on each page. Place
<<abc=1>><<abc>><br> on the first page. Place <<abc=abc+1>><<abc>><br> on page two,
and <<abc>> on page three. Run the application from the desktop and move between the pages. In
HTML/OS, you display single variables by placing them in << and >> brackets. You also run
instructions by placing them in << and >> brackets. This topic will be expanded on in the coming
chapters. Follow abc as you move from page to page. What happens?
Using the instructions here as a model, write new instructions that reset abc to 0 when first.html
is accessed, adds 2 to abc when two.html is accessed, and adds 3 to abc when three.html is
accessed.
Exercise 4
This exercise is about passing variables to applications when they are launched. You do this by
adding name=value pairs to Start-links. You'll see an example in this exercise. In Exercise 3, you
placed the instruction <<abc=1>><<abc>><br> in first.html. Delete this instruction so the
variable abc is not initialized to 1 when the page is accessed.
Access first.html using a Start-link. The default value of all variables, including abc, is ERROR.
Note how abc displays as ERROR and how abc plus 1 equals ERROR1. This is Okay. It happens
because the plus (+) character is also used to paste text together.
Add to the end of the Start-link the six characters, ?abc=5. Use this extended Start-link to access
first.html. This is what we meant by adding a name=value pair to a Start-link. What happens?
Now enable access to the variable abc in the Control Panel. To do this go into the Control Panel,
click Security and select the option Variable Allow. Type the variable name abc on a line by itself
and click Save Settings. Now try the extended URL again. What happens this time?
Next, change the example so abc is initialized to 100 when the page is launched and is decreased
by 1 every time any of the pages first.html, two.html, or three.html is accessed.
-30-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 3: Your First Program
Overview
In the previous chapter, you got online with HTML/OS. Now you build your first program. Your first
program is simple since the purpose of this chapter is to give you a taste of programming on the
Web. In this chapter, you will build a program that's an adaptation of a computer program called
Hello World.
Hello World is typically the first program students learn in computer science classes. Students learn
how to write a program that displays the words Hello World on the screen.
TE
AM
FL
Y
In keeping with tradition, in this chapter you'll also write a program that prints the words Hello World
on the screen. However, you'll do more. After all, anyone with knowledge of HTML can display Hello
World, because you can do that by simply typing the words into an HTML document. Instead, you'll
write a program that displays Hello World in different colors depending on the color a user selects.
See Figure 3.1. If a user selects Red, Hello World displays in red; if the user selects Blue, it displays
in blue; if the user selects Green, it displays in green.
Figure 3.1: Hello World is Your First Web-Based Computer Program.
Your first step is to create an HTML document with the look you want. This is because developing
programs with HTML/OS is top down. In other words, first you design your Web pages, then you add
the programming. Using the Web editor in the File Manager (see Chapter 2, Logging into Your Copy
of HTML/OS, for an explanation of how to use the File Manager), enter the following code into a Web
page:
<html>
<font color=[Location1]>Hello World</font>
<form action=[Location2]>
<input type=radio name=color value=red>Red<br>
-31Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<input type=radio name=color value=green>Green<br>
<input type=radio name=color value=blue>Blue<br>
<input type=submit name=mybutton value="Go">
</form>
</html>
This page displays the words Hello World on the screen and includes a set of three radio buttons the
user can use to select a color. In this code example, [Location1] and [Location2] appear
where you need to insert the dynamic content. [Location1] is positioned where the contents of a
variable containing a color value is needed. [Location2] is where you should link your HTML form.
When the user clicks the Go button, you want the color selected to appear at [Location1]. That
way, when the page is redisplayed, the words Hello World appear in the color the user selected.
Don't Know HTML?—If the HTML tags in the accompanying Web page are unfamiliar to you, you
will want to bone up on your HTML. Luckily, HTML is not difficult. It's not necessary to understand
most HTML tags. You only need to learn about 15 tags. Most others work only with Netscape or
Internet Explorer browsers, or in only the latest browsers. By sticking to a small set of tags, you'll
be able to learn HTML fast and you'll be learning the same tags used by sites like amazon.com and
yahoo.com that design their sites so they can be used by anyone on the Web. A list of these tags
can be found in Appendix B, Major HTML Tags.
To complete the Hello World program, you need to know a few things about HTML/OS programming.
First, when using HTML/OS, you create dynamic content in documents by placing Overlays in them.
An Overlay is one or more programming instructions surrounded with << and >> characters. For
example, the Overlay <<DISPLAY color /DISPLAY>> prints the contents of the variable color in
the page. The Overlay can also be written as <<color>>. You'll learn more about Overlays in
chapter 5, Overlays, Inlays, and On-click Overlays.
The second thing you need to learn about HTML/OS is that you link HTML forms to Web pages by
setting the ACTION parameter in the HTML form. When users submit their form, the selections they
made are automatically saved to the variable names specified in the HTML form and the user is
directed to the page specified by the ACTION parameter. The page can be written explicitly or you
can use an Overlay to dynamically set it.
In your program, you want to redisplay the page after the user clicks the Submit button. To
accomplish this, set ACTION to <<page>>, which is a tag in HTML/OS that contains the name of the
current Web page. Alternatively, you can set ACTION explicitly to the current page. For example, if
your HTML document is called mypage.html then you can write ACTION="mypage.html" in the
header of your HTML form. You are now ready to complete your first program. Edit the page so it
looks as follows:
<html>
<font color=<<color>>>Hello World</font>
<form action=<<PAGE>>>
<input type=radio name=color value=red>Red<br>
<input type=radio name=color value=green>Green<br>
<input type=radio name=color value=blue>Blue<br>
<input type=submit name=mybutton value="Go">
</form>
-32-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</html>
This revised page has Overlays where [Location1] and [Location2] once stood. An Overlay
that displays the value of the variable color replaces [Location1]. An Overlay that displays the
value of the current page replaces [Location2]. The page tag contains the value of the current
page. Click the Save/View button in your Web-based editor to run your program. Select different
colors and click the Go button. The page will redisplay with the words Hello World in the color you
select.
Understanding the Inner-Workings of Web Programs
It is important to understand the steps that occur in the split second you run a Web page. Unlike
computer programs that aren't Web-based, programs running on the Web are distributed between
the Web server and a browser. The Overlays run on the Web server. The HTML tags are run on the
browser.
When a user clicks a link on a Web page, a request is transmitted from the browser to the Web
server for a specific Web page. The page is then scanned by the HTML/OS engine on the Web
server. HTML/OS reads the document from top to bottom. As it encounters Overlays, it runs them
and uses them to build an HTML-only document, which is then transmitted to your browser. The
original document containing Overlays stays on the server. The browser reads the HTML-only
document received from the Web server and displays it on the screen.
When a user clicks a button in an HTML form, a similar sequence of steps occurs:
1. The variable names and values placed in the HTML form, along with a request for a
Web page are transmitted to the Web server. Notice that each radio button contains
a variable name and a value. When an HTML form is submitted, a variable name and
value pair is transmitted from the browser to the Web server (only the one selected,
of course).
2. The HTML/OS engine on the server takes those values and places them inside the
variable names specified in the HTML form.
3. The HTML/OS engine reads the Overlays in the document and uses them to build an
HTML-only document. It then transmits the document back to your browser. The
browser reads the HTML-only document and displays it on your screen.
As you can see, Web pages with HTML tags and Overlays contain instructions for both the Web
server and the browser. Web programs behave differently than programs not on the Web. They're
double-pass, meaning first the server passes across the document and then the browser passes
across the document. Computer programs not on the Web are single-pass.
The server-side tasks defined in Overlays are performed before client-side tasks. They are used to
change the contents of HTML documents dynamically. By the time the HTML document is rendered
by HTML/OS and transmitted to the browser, the document contains only HTML tags. It no longer
contains Overlays.
To witness this, run your Hello World page using the Save/View button in your Web-based editor.
Then view the source of your Web page with your browser. All the Overlays will be gone. In their
place, you will see the standard HTML. URLs, such as the one you defined for the ACTION in the
header of your HTML form will be replaced with encoded URLs that link back to the Web site. You
will not see any Overlays, because HTML/OS used them on the Web server to render the page.
They are not transferred across to the browser.
-33-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
This is a good thing. Browsers are finicky, and the way they interpret tags can vary from browser
version to browser version, which is why it's best to stick to the most basic HTML tags. On the other
hand, servers browsers do not interpret Overlays, so that Overlays can be used reliably. Reliability is
a necessity when it comes to building advanced Web sites. Although you can depend on Overlays to
work as advertised, you will need to use HTML tags sparingly and cautiously.
The HTML Document as a Template
As you also see, Web pages stored on the Web server are not the same as those received by the
browser. The pages on the Web server act as templates. A single page on the Web server can
render thousands of different pages, because it can include Overlays that extract data from variables
and databases and merge it with the page when requested by a browser.
Such pages will look different depending on the record from the database used or the contents of a
variable. On the other hand, the HTML tags in the page are constant. A single change to the HTML
in the page will change the rendering of any page it produces.
Summary
Although you have only created your first program, you are beginning to see why HTML/OS is so
powerful. It gives you the ability to fully define the dynamic nature of your Web pages, an important
component of advanced Web sites. The HTML outside your Overlays is the fixed portion of the page.
The HTML produced by your Overlays is the dynamic portion of the page. The page stored on the
server acts as a template. The pages transmitted to the users are the pages rendered by HTML
document stored on the server, not necessarily the document itself.
In this chapter, you learned about Hello World, a simple program that's educational but not
particularly useful. In the next chapter, you will write your first database program. There you learn
how to extract data from a database and merge it with an HTML document.
Exercises
After completing the three exercises in the following sections, you will have a better feel for
programming with HTML/OS. As you complete your exercises, you'll see that HTML/OS
programming is different from programming off the Web. You can build sophisticated programs
quickly and with less effort than with conventional off-the-Web (legacy) programming techniques.
Answers to all exercises are provided on this book's companion Web site as described in the book's
Preface.
Exercise 1
The program in this chapter uses radio buttons. Exchange the radio buttons with a pull-down select
box. Note that the code for a pull-down Select box is written as follows:
<select NAME=variable_name>
<option value=value_one>Selection One
<option value=value_two>Selection Two
<option value=value_three>Selection Three
</select>
-34-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 2
Exchange the radio buttons with three hypertext links. Note that HTML/OS allows you to place namevalue pairs in standard hypertext links. For example, you can write the following:
<A HREF=some_page.html NAME=var_name VALUE=var_value>Selection One</A>
When the user clicks the link var_name is set to var_value.
Exercise 3
Build a Hello World program where both the color and font size can be varied by the user. Use two
pull-down select boxes in a single HTML form to accomplish your goal.
-35-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 4: Your First Web Database Program
Overview
After you write the program Hello World, you're ready to go one step further. In this chapter, you will
build a program that allows users to search a contact database and display a search result. Your
program does not require many Overlay tags—only a few; but it does require that you learn how
HTML/OS stores and displays variables, how to use an Overlay tag called DBFIND that searches
databases, and how to use LAYOUT, an Overlay tag for reorganizing information in variables.
This chapter introduces you to the concept of two-dimensional data, an important concept used
throughout HTML/OS. You'll learn how to create a database with a program called dbConsole. You'll
then learn how to write the tag to fill a variable with multiple rows and columns of search results and
how to reorganize that variable so it displays nicely on the page. Figure 4.1 illustrates the program
you will build.
Figure 4.1: Creating a Web Database is Point and Click.
You start by creating a database and populating it with sample data. You use dbConsole, a pointand-click Web application for creating and managing databases, as described in the next section.
Creating a Database
To create a database, click the dbConsole icon on your HTML/OS desktop. Then select
Create. You should see a screen like that shown in Figure 4.1.
A First Look at HTML/OS Databases—The following list contains a quick
introduction to the basic concepts of HTML/OS databases:
1. The HTML/OS database engine is a high-performance engine capable of
serving thousands of users simultaneously.
2. HTML/OS databases look like spreadsheets placed on their sides. Instead
of columns, they have fields. Instead of multiple rows, they have multiple
records. HTML/OS databases can contain over a million records each.
Each record can contain hundreds of fields.
-36-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
3. The first field of an HTML/OS database record is called record. It
contains a number that uniquely identifies the record. The record field is
followed by the fields you set up.
4. In HTML/OS, databases and database tables are the same things,
because HTML/OS database tables do not need to be linked together
with schemas, a technique used by relational databases off the Web to
join multiple database tables together. HTML/OS can join multiple
database tables together at program time, a feature not typically found in
database engines. This topic is discussed in further detail in Chapter 20,
Database Joins.
5. The HTML/OS database engine works with your Web server. It does not
need a separate database server like off-the-Web database engines.
There is no database server to maintain, and you can manage the
database engine easily through a Web-based program called dbConsole.
6. The HTML/OS database engine is built directly into the HTML/OS engine.
To access the database from within your HTML documents, you use tags
such as DBADD, DBEDIT, and DBFIND. No SQL (a widespread
database query language) is needed. Unlike SQL-based solutions,
HTML/OS is a type of Fourth Generation Language (4GL). When using
4GLs, database programming is simplified, because you need to use only
one tag per database operation. HTML/OS databases are discussed in
further detail in Chapter 11, What Is a Web Database.
To set up your contact database, specify the names of the fields you want in the database
along with the field type and field length for each. In the example here, you use a database
with four fields: contact_company, contact_name, contact_phone, and
contact_email (excluding the field record, which is created automatically).
Set the field type of your fields to STR. STR stands for string, which is computer talk for
text. Specify a field length for each of your fields. Use the values shown in Figure 4.1. Set
the database name to /mycontacts, and click the Create Database button. You have
created your first database.
Open the database by clicking Open in the left column of dbConsole. Your database is in
the Home directory because you named it /mycontacts. (Names beginning with a slash
are in the Home directory.) Then click Edit in the left column to edit a new or existing
record. Add a contact to the database. Enter the contact data for your fields, and click the
Save button above the record. Click the New button above the record to start a new record.
Repeat until you have entered four or five records.
Creating Your Web Page
You're ready to build a Web database program. Just as in Chapter 3, Your First Program, you design
your program from the top down. Exit dbConsole and use the File Manager to create a new file. Fill it
with the following HTML code:
<html>
<form method=post ACTION=[Location1]>
<font size=2>Enter Search:</font><br>
<input type=text name=mysearch size=15>
<input type=submit value="Find">
</form>
Search Results:
-37-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
[Location2]
</html>
To view the Web page, click the Save/View button in your Web-based editor. You should see an
HTML page that has a search box with a Find button and a place for a search result.
Adding the Overlays
In the code you entered previously, the parameter ACTION is set to [Location1] and
[Location2] is placed where search results should be displayed. To complete your program, you
need to write code for [Location1] and [Location2]. Since you want to redisplay your Web
page when the user clicks the Find button on the page, you need to set [Location1] to the page
name by replacing ACTION=[Location1] with ACTION=<<page>> or ACTION="/mypage.html"
assuming your page is called "/mypage.html".
[Location2] is where you insert an Overlay that searches your contact database and displays a
message or a search result. What you display depends on the value of mysearch. The first time the
user accesses the page, the variable mysearch is equal to ERROR, because variables in HTML/OS
default to the value ERROR. On the other hand, if a user enters a search, mysearch is equal to the
value the user typed into the input box. In this second case, you want to conduct a search of your
database and display a table of results.
The Overlay at [Location2] needs to take into account these two cases. By using an IF-THEN
statement, you can selectively perform a search when mysearch is not equal to ERROR and display
a text message when it is not. Your Web page, with [Location1] and [Location2] swapped out
for the proper code, is shown here:
<html>
<form method=post ACTION=<<page>> >
<font size=2>Enter Search:</font><br>
<input type=text name=mysearch size=15>
<input type=submit value="Find">
</form>
Search Results:
<<IF mysearch="ERROR" THEN
DISPLAY "Please enter search above. " /DISPLAY
ELSE
sstr = 'contact_name ~ ' + '"' + mysearch + '"'
mydata = DBFIND("/mycontacts",sstr,1,20,
"contact_company,contact_name,contact_phone,contact_email")
mynewdata = LAYOUT(mydata, [1], " - ", [2], " - ",[3], " ",[4], "<BR>")
DISPLAY mynewdata /DISPLAY/IF>>
</html>
-38-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The tags that perform a search are placed between the ELSE and the /IF in the IF-THEN statement.
The tags DBFIND and LAYOUT are used. The tag DBFIND searches a database and places a subset
of the records it finds into mydata. The LAYOUT tag is used to format the data in mydata so it can
be displayed with the DISPLAY tag. The remainder of this chapter gives you a detailed explanation
of how this application works.
Understanding Your Application
The following sequence of steps occurs when a user runs the Web page presented in the preceding
section:
1. When a user enters a value in the search box and clicks the Find button, the text
that user entered into the input box is transmitted to the Web server and saved into
the variable mysearch.
2. The HTML/OS engine scans the Web page from top to bottom. As it moves through
the file, it creates an HTML page for the user. The first Overlay it encounters is
<<PAGE>>. HTML/OS adjusts the HTML form so it links back to the current page.
<<PAGE>> is a tag that contains the name of the current document.
3. The next Overlay HTML/OS encounters is below the words Search Results. The
IF-THEN statement in the Overlay runs the instructions between the ELSE and the
/IF in the event mysearch is not ERROR. In this case, the Overlay fills the variable
sstr with a query string. The instruction uses plus (+) signs to paste text together. If
mysearch is equal to george, the instruction would fill sstr with contact_name
~ "george". HTML/OS uses the variable sstr as the input of the DBFIND tag on
the following line.
4. HTML/OS runs the DBFIND Overlay tag. The tag fills mydata with a table of
information. It can do this because HTML/OS variables are two-dimensional. They
have one or more columns and one or more rows. The accompanying
"Understanding HTML/OS Variables" note discusses how HTML/OS stores variables.
A more thorough discussion is provided in Chapter 5, Working with Variables.
The five parameters in DBFIND define which database to search, which records to
retrieve, and which search results to place in mydata. Like all HTML/OS tags, parameters
are separated from each other with a comma. Parameters can be literal text, expressions,
other tags, or variables. The first parameter of DBFIND is the name of the database. The
second parameter contains the search criterion, also known as the Boolean query. If the
Boolean query is empty, it means you want to extract records from the entire database. If
that parameter contains a value like contact_name ~ "J", DBFIND extracts only
those records in the database with contact names beginning with the letter "J". Boolean
queries are discussed in greater detail in Chapter 12, Boolean Queries.
The third, fourth, and fifth parameters tell DBFIND which records to put into mydata. The
third and fourth parameters tell DBFIND to fill mydata with the first up to the tenth search
result. If it finds less than ten results, it places only what it finds in mydata. If DBFIND
finds no results, it sets mydata to an empty string—a one-row-by-one-column cell that's
empty. If it finds more than ten records, it puts only the first ten items it finds into mydata.
The last parameter is a list of field names. It tells DBFIND what to put in each column of
mydata. The columns of mydata are filled from left to right in the same order as the fields
you specified in the parameter. DBFIND is discussed in further detail in Chapter 14,
Building a Database Search Page. It is also defined in Appendix C, Major HTML/OS Tags.
-39-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
5. After running the DBFIND tag, HTML/OS runs the LAYOUT tag. The tag is used to
reorganize mydata, so when the results display on the screen, the HTML you see is
ordered the way you want. This formatting is needed, because displaying the
contents of mydata on the screen produces an ugly presentation. The LAYOUT tag is
described after Step 7, which is the last step taken when this Web page runs.
6. The line DISPLAY mynewdata /DISPLAY runs, placing the contents of
mynewdata, the table of search results, into the Web page.
7. The Web page generated by HTML/OS is transferred to the user's computer and
displayed with a browser.
Understanding HTML/OS Variables—HTML/OS variables are two-dimensional. Most of the time,
variables have one column and row; they never have less. You may want to think of HTML/OS
variables as mini-spreadsheets with a minimum size of one column by one row. For example, the
instruction myvariable = 5 creates the variable myvariable with one column and one row
containing the value 5. When HTML/OS variables are only one column by one row, you don't think
much about their two-dimensionality. You work with them without regard to their ability to have
more rows or columns; but when you need tables of well-organized values, HTML/OS' twodimensionality becomes indispensable. Searching databases is a prime example. (For a more
comprehensive discussion of databases, see Chapter 11, What Is a Web Database.) A database
search typically yields multiple records, and each record has multiple fields. Because HTML/OS
variables are two-dimensional, you can place the results of a search in a single variable. Field
values are placed in different columns of a single row. The variable is filled, one record per row. As
another example, suppose you want to work with a comma-delimited file containing login names
and passwords. (A more comprehensive discussion of delimited text files is provided in Chapter 6,
Working with Variables.) You can load the delimited file into the cells of a variable with a single
COPY tag. Or suppose you want to add up the values in a column of a variable; again, only a single
tag is needed. HTML/OS variables are discussed in further detail in Chapter 5.
Using the LAYOUT Tag
When HTML/OS displays multicell variables, it displays each cell, one after the other, left to right, top
to bottom, with nothing between each cell. Suppose mydata contained the following data:
ƒ
John
Smith
212-555-1212
Janet
Jones
310-555-1212
Jack
Chen
415-555-1212
The code DISPLAY mydata /DISPLAY would generate the following text in your HTML document:
JohnSmith212-555-1212JanetJones310-555-1212JackChen415-555-1212
This is not acceptable. You want your search results to appear as they do in Figure 4.2. The HTML
code for such a search result looks as follows:
-40-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 4.2: Your Web Database Program with Search Results Will Look like This.
John Smith - 212-555-1212<br>Janet Jones - 310-555-1212<br>Jack
Chen - 415-555-1212<br>.
Working backwards, since you know that HTML/OS displays data in cells of a variable left to right
and top to bottom, displaying the following six-column variable produces a page like that shown in
Figure 4.2.
ƒ
-
Smith
-
212-555-1212
<br>
Janet
-
Jones
-
310-555-1212
<br>
Jack
-
Chen
-
415-555-1212
<br>
AM
FL
Y
John
TE
To create this variable from mydata, use the LAYOUT tag, which is designed to take a variable with
many columns and rows and produce a new variable with just as many rows but with columns filled
in according to the parameters you specify in the tag. In this case, you want to build a six-column
variable called mynewdata from mydata. The first, third, and fifth columns in mynewdata come
from mydata. The second, fourth, and sixth columns in mynewdata are filled with a space, a dash
and a <BR> respectively. The instruction mynewdata=LAYOUT(mydata,[1]," ",[2]," ",[3],"<BR>") is what you need.
The LAYOUT tag works as follows: Its first parameter is the input variable. All other parameters
specify what to place in each column of its output. The LAYOUT tag works from left to right, creating
columns in an output variable as it reads its own parameters. It starts with its second parameter and
moves right until there are no more parameters. If a parameter contains quoted text, the entire
column in the output variable is filled with the text. If the parameter is a column number in square
brackets, the column in the output is copied from the specified column of the input variable.
In your example, the first parameter of LAYOUT is mydata, the three-column variable containing the
result of your search. The remaining parameters define what you want to appear in each of the six
columns of mynewdata. Note how you are using two kinds of parameters. To specify a column from
the starting variable, you include the column number in square brackets. On the other hand, to fill a
column with text, you specify the value explicitly.
Reformatting Your Report
The code in the "Adding the Overlays" section in this chapter produces a report with fields separated
by dashes. (Refer to Figure 4.2.) You will more likely want to format your search result in columns
using an HTML table. You may want to add column headers too. To accomplish this, you should
change the parameters in LAYOUT so the tag displays multiple HTML table rows, each beginning
-41Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
with a <tr><td> and ending with a </td></tr>. You need to place the tags, <table> and
</table> around the rows as well to delineate them. Rewrite your page as follows:
<html>
<form method=post ACTION=<<page>> >
<font size=2>Enter Search:</font><br>
<input type=text name=mysearch size=15>
<input type=submit value="Find">
</form>
Search Results:
<< IF mysearch="ERROR" THEN
DISPLAY "Please enter search above." /DISPLAY
ELSE
sstr = 'contact_name ~ ' + '"' + mysearch + '"'
mydata = DBFIND("/mycontacts",sstr,1,20,
"contact_company,contact_name,contact_phone,contact_email")
DISPLAY
"<table border=1>" +
"<tr><td>Company</td><td>Name</td><td>Phone</td></tr>"
/DISPLAY
mynewdata = LAYOUT(mydata,
"<tr><td>", [1],
"</td><td>", [2],
"</td><td>", [3],
"</td></tr>")
DISPLAY mynewdata /DISPLAY
DISPLAY "</table>" /DISPLAY
/IF
>>
</html>
This code displays a report in table format as shown in Figure 4.3. You can experiment with the
parameters in LAYOUT to vary the look of your report. Alternatively, you can control your display by
using FOR loops, which are discussed in Chapter 6, Working with Variables.
-42-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 4.3: Your Web Database Program Looks like this with Nicely Formatted Search Results.
Summary
In this chapter, you created your first Web database program. Unlike the previous chapter where you
learned the Hello World program, here you set up and accessed a database, worked with twodimensional data and did some programming. Unfortunately, you did this without any fundamental
understanding or description of how HTML/OS works.
In the next two chapters, you return to the basics. You learn more about variables, how to place
Overlays in documents, and how to use the most important Overlay tags. The next two chapters
serve as a solid foundation for those chapters that follow them dedicated to specific areas of
advanced Web development.
Exercises
The following exercises give you a feel for working with the LAYOUT tag. The last exercise builds on
what you learned in the previous chapter regarding select boxes and the ability to use them to vary
the value in a variable. Answers to all exercises are provided on this book's companion Web site as
described in the book's Preface.
Exercise 1
Extend your report so it displays contact_email. You need to modify the LAYOUT tag to
accomplish this.
Exercise 2
Add a select box to the HTML form that gives the user the ability to select a page size of 20, 40, or
100. Hint: Set the variable name of the select box to page_size and substitute the 20 in DBFIND
with the variable.
-43-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Part II:
Programming Basics
Chapter List
Chapter 5: Overlays, Inlays, and On-Click Overlays
Chapter 6: Variables, Conditionals, and Loops
Chapter 7: Dubugging Techniques
Chapter 8: Building Text Editors
Chapter 9: Building Login Pages
Chapter 10: HTML Forms Processing
-44-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 5: Underlays, Inlays, and On-Click Overlays
In this chapter, you learn the fundamentals of placing HTML/OS Overlays in ordinary HTML
documents. The title of this chapter refers to the three kinds of Overlays you can use. You start by
learning how Overlays are written. Then you will learn the different ways you can place them in
documents and when to use each kind of Overlay.
The end of this chapter discusses the Web-centric perspective—the perspective Web developers
should take when building advanced Web sites and Web-based applications.
The HTML Document
The HTML document is the center and foundation of Web-based computing. Text, images, sound,
video, data stored on the Web server (in files and databases), as well as JavaScript, Java applets,
and links to other HTML documents are controlled by the HTML document. In essence, the HTML
document is the "control" document of the Web pages. In fact, one could say HTML documents, at
least those run in browsers, are the Web.
Static Web sites are built from pure HTML documents. These HTML documents are limited in that
they have no connection to data and files stored on the server, because HTML documents are run
on the browser, not the server. Static Web sites do little more than store documents prepared in
advance and pass them to users who request them. On static Web sites, the HTML document stored
on the server is the same document that is delivered to the user's browser.
By contrast, dynamic Web sites have the ability to deliver on-the-fly HTML documents to browsers.
Although they must still deliver pure HTML documents to users, the documents stored on the server
can contain instructions that reference data and files on the server. In HTML/OS, the instructions you
place in HTML documents are called Overlays.
What is an Overlay?
Overlays are special instructions that are actually program segments. Specifically, an Overlay is a
set of << and >> brackets around programming code. Overlays can contain calculations, instructions,
and Overlay tags. They may contain only a single variable, an instruction, or hundreds of instructions.
An Overlay tag is a programming function or word used to perform a specific task. For example, the
Overlay <<mynumber=RANDOM(10)>> contains the Overlay tag RANDOM. The tag selects a random
number between one and ten and places it in the variable mynumber. Overlay tags are also referred
to as Otags, HTML/OS Tags, or Reserved Words.
There are two kinds of Overlays. Some Overlays contain one or more instructions, and some
Overlays contain a single calculation.
Overlays Containing One or More Instructions
The following Overlays are examples of Overlays containing one or more instructions:
<<DISPLAY A /DISPLAY>>
<<DISPLAY A + B /DISPLAY>>
<<A="<font color=red>" B=6 C=7>>
<<IF A > 5 THEN B=6 /IF C=7>>
Prints the contents of A.
Prints the result of A + B.
Assigns values to A, B, and C.
Executes a conditional and sets a value for C.
-45-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
An Overlay can contain as many instructions as you want. You separate instructions from each other
with one or more spaces or a new line. Like HTML, HTML/OS ignores multiple spaces and new lines,
and it isn't casesensitive, meaning that it doesn't matter whether you write a tag in upper case,lower
case, or a mix of the two. For instance, the following three instructions are the same:
1. <<DISPLAY A /DISPLAY>>
2. <<DiSpLay
a
/DiSpLaY >>
3. <<
display
A
/display
>>
Overlays Containing a Single Calculation
The following Overlays are examples of Overlays containing a single calculation:
<<A>>
<<A + B>>
<<TODAY>>
Prints the contents of a variable labeled A.
Prints the result of A + B.
Prints today's date.
An Overlay with a single calculation is a shorthand way of displaying the results of the same
calculation using a DISPLAY instruction. For example, the three previous Overlays can also be
written as follows:
<<DISPLAY A /DISPLAY>>
<<DISPLAY A + B /DISPLAY>>
<<DISPLAY TODAY /DISPLAY>>
Positioning Overlays in a Document
For a browser to be able to read an HTML document, the document should begin with an <html>
tag and end with an </html> tag. The documents stored on the server, however, do not need to
begin with a <html> and end with a </html>. They can begin or end with Overlays. In HTML/OS,
the portion of the document sitting between the <html> and </html> tags is the display portion of
the document. Outside of this area, nothing can be displayed, but instructions can be executed. For
example, when a Web page is requested, instructions above the <html> tag can direct the user to
another Web page if the user has not logged in. In general, you can place Overlays above, inside,
and below the display portion of the document. You'll see how this works in a moment.
When a user requests a Web page, HTML/OS reads the page from top to bottom. If the first thing it
finds in a Web page is an Overlay (even before it encounters the <html> tag), it runs the Overlay.
This kind of Overlay is known as an Underlay. Underlays can contain any number of instructions and
any of the Overlay tags in the HTML/OS arsenal, with the exception of those tags that change the
content of the page. Instructions, such as DISPLAY a /DISPLAY are not allowed, because
Underlays are not in the display portion of the page.
-46-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Overlays you place between the <html> and </html> tags are called Inlays, because you place
them in the display portion of the Web page. The Overlays you used in Chapters 2, Logging into
Your Copy of HTML/OS, and 3, Your First Program, were all Inlays.
Because you place these Inlays in the display portion of the page, you can use them to change the
document on the fly. Like Underlays, they too can run most of the Overlay tags in the HTML/OS
arsenal. However, although Inlays can run tags like DISPLAY, they can't run the GOTO tag, because
HTML documents are contiguous—a fancy way of saying they must remain whole. Once within the
display portion of a page, it makes no sense to tell HTML/OS to stop rendering the page and begin
rendering some other page.
The third and last kind of Overlay is called the on-click Overlay. You place on-click Overlays below
the </html> tag. They are different from Underlays and Inlays, because on-click Overlays are not
executed as HTML/OS reads a document from top to bottom. On-click Overlays run in response to a
user clicking a hypertext link or a Submit button—hence the name on-click Overlay. You also write
them differently. On-click Overlays have names since they need to be specified. Hypertext links and
HTML forms that call the on-click Overlays specify which on-click Overlay to run. For example, the
hypertext link <A HREF=dostuff>Do Some Stuff</a> calls the on-click Overlay dostuff when
it is clicked. The on-click Overlay placed below the </html> tag in the document, begins with the
word Overlay, followed by the word dostuff, the name of the on-click Overlay.
An example page with all three kinds of Overlays follows this paragraph. The Overlay at the top of
the Web page is an Underlay. It contains four instructions. The last of the four instructions contains a
GOTO tag. The instruction tells HTML/OS to stop reading the page and read home.html instead (in
the event tryno is larger than max_tries). This use of the GOTO tag is appropriate since page
rendering has not yet begun. The second Overlay displays the value of the variable tryno. It's an
Inlay. It changes the document on the fly, printing tryno where it's placed.
Just below the display portion of the page is an on-click Overlay called test_password. It runs
when a user clicks Go, which is the Submit button defined in the HTML form on the page, because
the ACTION parameter in the HTML form references test_password. Examine the following code
to pick out the various Overlays:
<< max_tries = 5
mypass = ""
IF tryno="ERROR" THEN tryno=1 /IF
IF tryno > max_tries THEN GOTO "home.html" /IF
>>
<html>
Try number <<tryno>><br>
Enter Password:
<form method=post ACTION=test_password>
<input type=password name=mypass size=10>
<input type=submit value="Go">
</form>
</html>
<<OVERLAY test_password
IF mypass="LetMeIn" THEN
-47-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
GOTO "private.html"
ELSE
tryno = tryno + 1
GOTO PAGE
/IF
>>
The on-click Overlay here was referenced from an HTML form. As an example of referencing an onclick Overlay in a hypertext link, suppose you added the following hypertext link to the document:
<A HREF=test_password name=tryno value=0>Clear Counter</a>
The link, when clicked, sets the value of tryno to 0 and runs the on-click Overlay test_password.
If you add this hyperlink to the previous document, the instruction tryno = tryno + 1 would run
(since mypass would not be equal to LetMeIn. When the page redisplays, it would print Try number
1 on the page.
Overlay-HTML Rules
Adding what you just learned about the three types of HTML/OS Overlays to a basic understanding
of HTML tags and how to write the instructions and use variables gives you everything you need to
build advanced Web sites. The combination of HTML and Overlays gives you a complete
development environment. When mixing Overlays with HTML also keep in mind the following rules:
Overlays can display HTML. For example, consider this hypertext link:
<A HREF=done.html name=abc value=0>Reset Me</a>
Its effect is the same as that of the following:
<<DISPLAY "<A HREF=done.html name=abc value=0>Reset Me</a>"
/DISPLAY>>
HTML tags can contain Overlays. For example, you can write the following:
<A HREF=<<PAGE>>>Redisplay</a>
The Overlay tag PAGE is a predefined tag that contains the name of the current document. Clicking
this link directs the user to the current page.
Overlays can not contain other Overlays. For example, you can't write the following:
<< a = a + <<b>> >>
Instead, you would write
<<a = a + b>>.
Overlay names can not contain a dot. For example, you can write the following:
-48-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<<OVERLAY myprocessing
instructions here
>>
You can't reference the same Overlay if it were written as follows:
<<OVERLAY myprocessing.data
instructions here
>>
Overlay tags are optional. An HTML document does not need to contain Overlays. It can be a
standard HTML document, created with any HTML editor, with no programming in it. For example,
the following page is valid:
<HTML>
Hello World
</HTML>
HTML tags are optional. An HTML document does not need to contain any HTML either. It can be a
single Underlay. For example, the following page is also valid:
<< myvariable = "Hello World"
GOTO "home.html"
>>
These last two rules, that Overlays and HTML tags are both optional, mean that documents can be
either pure HTML documents or pure programming. This duality is an essential component of the
Web perspective, which is discussed in the following section.
The Web-Centric Perspective
The Web is a fusion of three worlds: the publishing world, the programming world, and the database
world. HTML/OS carries this fusion to its logical conclusion, allowing the publishing (HTML) world,
the programming world, and the database world to become one. In this chapter, you've seen how the
first two worlds merge. Later, in Chapter 11, The Web Database, you see how the database world is
also "one" with the HTML document.
Recall that when using HTML/OS, the fundamental programming document is the HTML
document—as stored on the server. The document can be all programming, all HTML, or a
combination of the two.
Web-Based Computing Demands HTML Centrism—Whereas the center of the PC world is
CPUs—hardware such as video cards, and operating systems such as Linux, Windows, and
MacOS—the center of the Web computing world is the HTML document—a simple document that
doesn't care about the CPU you use, the kind of video card you have, or the operating system your
PC uses. Unfortunately, companies often apply PC programming techniques to the Webprogramming world. Whether due to habit or special interest, the techniques direct you away from
the HTML document and lead you in the direction of hyper complexity.
-49-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
If you're a programmer, the Web-centric view demands that the notion of an external API
(Application Program Interface) be dropped, because API's are defined to work with hardware and
computer screens, not the Web. The HTML document is your API. You can place most of your
programming directly in these documents. Does this mean there's no room for old-style programming?
Not at all. If necessary, you can drop compiled modules into HTML/OS as system extensions. (This
topic is not discussed in this book. It's reserved for C and C++ programmers.) You're also free to use
function libraries. (See the knowledge base on the Aestiva Web site at
http://www.aestiva.com/support/.) But these are side issues. The HTML document is now your
primary source document.
Summary
In this chapter, you learned how to place Overlays in documents. You learned they can be in the
display portion of the document (Inlays), above the <html> tag (Underlays), or—when they're
activated in response to a user clicking a button in an HTML form or a hypertext link—below the
</html> tag (on-click Overlays). You've seen that Overlays allow you to think of HTML documents
as pure HTML, pure programming, or a fusion of the two.
This chapter did not discuss how you write specific instructions, how you handle variables, or how
you write the more common Overlay tags. Those topics are discussed in depth in the next chapter,
Variables, Conditionals, and Loops.
Exercises
The two following exercises are designed to get you used to using Overlays to build highly dynamic
Web pages. Answers to all exercises are provided on this book's companion Web site as described
in the book's Preface.
Exercise 1
Create a simple calculator using an HTML form with two input boxes and a select box containing the
options Add, Subtract, Multiply, and Divide. Link the form to an on-click Overlay on the bottom of the
page. Calculate the result and return to the same page to display the result. You will only need an
IF-THEN statement.
Exercise 2
Create a guess-my-number game. Have the user click a link to start the game. The link should go to
an on-click Overlay that sets a variable between one and ten. Place an HTML form on the page.
When the user submits the form, see whether the number matches the game selection. If it does,
display a message saying the user won. If not, display a message asking the user to try again. You
will need to use a RANDOM tag and IF-THEN statements. You will need two on-click Overlays; one to
run instructions when users click a Begin Game link and one for running instructions when the users
click the Submit button in the HTML form, after entering their guess.
-50-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 6: Variables, Conditionals, and Loops
Overview
Variables are the life-blood of applications. You use them in tags, calculations, and most everything
you do. The HTML/OS development environment uses variables to carry data between HTML forms,
Overlay tags, calculations, fields in databases, and files. The same variable environment is used
throughout. In this chapter, we review the rules, requirements, limitations, and conventions used
when working with HTML/OS variables. We also discuss how to use IF-THEN conditional
statements and FOR loops, how to work with two-dimensional data, how to load comma-delimited
text files into variables, and how to work with variables in HTML forms.
The information you learn in this chapter is used throughout this book, because advanced Web
development is mostly about moving data between variables, files, and HTML documents. Along with
the fundamentals discussed in Chapter 5, Overlays, Inlays, and On-Click Overlays, these two
chapters give you most of the foundation you need. The only major area missing from this discussion
is the topic of databases, which is discussed in Part III, Database Programming.
AM
FL
Y
If you're comfortable with BASIC programming, feel free to skip over the sections that appear trivial.
But take a look at the sections on FOR loops, because they introduce you to a new kind of FOR loop
used in HTML/OS with which you may not be familiar. Also it is important that you look at how
HTML/OS works with HTML forms. You'll also see code examples that introduce you to some of the
more popular Overlay tags used in HTML/OS. You may want to take note of them.
Variable Names
TE
Regardless of your familiarity with BASIC programming, you will want to be online as you go through
this chapter so you can try the code examples provided.
Starting with how to write a variable name, you need to adhere to the following rules:
ƒ Variable names are case-insensitive.
ƒ Variable names can not begin with a number.
ƒ Variable names can contain only letters, numbers, a dot (period), and an underscore.
ƒ Other characters, including spaces, are not allowed.
ƒ Variable names can not be the same as the name of a tag or function.
The following list includes examples of valid variable names:
ƒ data14
ƒ cart_price
ƒ myfolder
ƒ acct.name.first
The last variable example in the list, acct.name.first, shows a variable containing dots. Note
that dots in variable names are treated like other characters. Object-oriented languages, such as
JavaScript, Java, and C++ use the dot as a special character. HTML/OS is a functional language
similar to those found in spreadsheet macros. The dot has no special meaning.
The following list contains examples of invalid variable names:
-51Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ 14data*
ƒ cart-price
ƒ my folder
ƒ acct$name$first
The first example in the previous list, 14data, is invalid, because it begins with a number. The
second, cart-price, is invalid, because it includes a hyphen (which, when interpreted by
HTML/OS, means subtract price from cart). The third, my folder, contains a space, which is
not allowed. The last, acct$name$first, contains dollar signs, which are special characters and
are not allowed.
When naming variables, it is best to use words with meaning. It makes your programming more
readable. Also, it's a good idea to prefix your names with the name of your application and an
underscore. This ensures your variables don't conflict with Overlay tags, because Overlay tags don't
contain underscore characters. As an example, when building a shopping-cart application, use
variable names like shop_cart, shop_taxrate, shop_shippingrate, and so on.
Default Value of Variables
The default value of variables in HTML/OS is ERROR. To be more specific, the default value of all
variables is a one-cell table with the word ERROR stored in it. For example, suppose you added the
following Overlay to your Web page:
<< DISPLAY my_error_message /DISPLAY>>
If you previously set a value for the variable my_error_message, it would display on the page
where the Overlay was positioned. If you did not previously set a value for the variable, the word
ERROR would display on the page.
Since you don't want ERROR to appear on Web pages, you should place the instruction inside an IFTHEN conditional, as follows:
<<
IF my_error_message!="ERROR" THEN
DISPLAY my_error_message /DISPLAY
/IF
my_error_message="ERROR"
>>
Two-Dimensionality
When using HTML/OS, it is important to remember that variables are two-dimensional. In other
words, all variables contain one or more cells organized in columns and rows. The simplest of all
variables, the empty string, has one row and one column with nothing in it. The instruction abc=""
sets a variable to an empty string.
A variable can never have less than one row or less than one column. The two-dimensionality of
variables does not change the way you perform simple calculations; but it gives you powerful
capabilities that come in handy when working with databases and delimited text files and when
performing operations on large sets of information. Figure 6.1 shows an assortment of variables.
-52-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 6.1: The Size of my_var Varies Depending on What You Place in It.
Specifying Variables
When working with variables, you may specify the entire variable or only a specific cell. To specify
the entire variable, write the name of the variable. To specify a specific cell, follow the variable name
with the column and row number in square brackets.
For example, cart_products refers to all the cells in the variable cart_products; whereas
cart_products[1,3] refers to the first cell in the third row of cart_products. Note also that
you don't need to specify both the column and the row. You may specify only the column number.
When you do not specify the row, the row number equals 1. For example, cart_products[3] and
cart_products[3,1] both refer to the cell in the third column of the first row of cart_products.
The Math Convention
When using HTML/OS, you specify cell positions by column and row, not by row and column. Also,
the first cell in a variable is at position 1,1, not at position 0,0. This is called the math convention. It's
intuitive, unless you were trained as a computer programmer. Many computer languages define the
upper-left column and row number of a table as position 0,0 and most often, they require that you
specify the row number ahead of the column number. If you are familiar with programming
languages, you will want to remember that HTML/OS uses the conventions from mathematics, not
computer science.
Variable Assignment
The following list contains example valid variable value assignments:
ƒ
ƒ
ƒ
ƒ
my_var = "-10"
my_var = ROW("-10","10","20","30")
my_variable [3] = "-10"
my_var[3,4] = "-10"
The first example, my_var = "-10", creates a new variable with one column and one row
containing a negative 10. Any data previously stored in my_var is lost.
The second example, my_var = ROW("-10","10","20","30"), creates a variable four
columns wide and one row in height containing the specified cells. As in the first example, any data
previously stored in my_var is lost. The example uses the tag ROW to build the one-row variable. For
additional information on ROW, see Appendix D, HTML/OS Tag Reference Guide.
-53-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The third example places a negative 10 in the cell at column 3, row 1 of my_var. Values in other
cells are preserved.
The fourth example places a negative 10 in the cell at column 3, row 4. Values in other cells are
preserved.
In the last two examples, if my_var were less than three columns wide, my_var would be
automatically expanded to have more rows and columns. Other cells would have been left
unchanged. Although the default value of a variable is a one-column, one-row cell containing the
word ERROR—when expanding the size of a variable, undefined cells are left empty. For example, if
you were to run the two instructions my_var="X" and my_var[3,3]="X", the cells in my_var at
locations other than 1,1 and 3,3 would be empty.
Data Types
Each cell of a variable can store numbers, fractions, dates, times, logical values (TRUE or FALSE),
small pieces of text, or large pieces of text (such as a text file). You don't need to write tags telling
HTML/OS how many rows or columns you need in a variable. And you don't need to declare the type
of data you put in each cell or how much data you expect to put in a cell. This feature is known as
auto-data-typing. For example, suppose you want to create a variable called stooges containing
three columns and one row with the cells containing the names Larry, Moe, and Curly. You
would write it as follows:
stooges=ROW("Larry","Moe","Curly")
Or suppose you have a text file called mydoc.txt containing 10,000 characters that you wish to
load into the variable mytext. You would write the following code:
COPY FILE="mydoc.txt" TO mytext /COPY
As you see, to create or use a variable, you simply use it.
Using Different Kinds of Variables
Although variables are auto-data-typed, you must still write them according to predefined
conventions, and they must have values within predefined limits. A summary of those limits and
conventions is shown in the following sections.
Text
You are free to write text almost any way you want. Text stored in a variable can be 0 to 100,000
characters long. It can contain non-English (eight-bit) characters. It can contain special characters
such as tabs, linefeeds, and carriage returns, and it can contain HTML codes, programming tags, email messages, XML code, and complete or partial text documents.
Integers
Integers are whole numbers—positive or negative—up to 11 characters in length. When storing
integers in a variable, do not place special characters (such as $) or commas in them. If you wish to
reformat the number, do so when you display it. For example, consider the following statement:
price = "50000"
-54-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
price_with_tax = price * (1.0825)
DISPLAY FORMAT(price_with_tax,"comma",2) /DISPLAY
The price is stored properly. Formatting is applied when the variable is displayed. The example uses
the FORMAT tag. In this case, the tag reformats the number with two decimal places, placing commas
in it if needed. For additional information on FORMAT, see Appendix D, HTML/OS Tag Reference
Guide.
Other Limits—When working with variables keep in mind the following facts about variables:
ƒ
ƒ
HTML/OS variables may have up to 1000 columns and up to 100,000 rows.
Unless otherwise specified, text searches, tags and comparisons are caseinsensitive. The one exception is when specifying files because Unix and
Linux and MacOS X systems are case-sensitive.
Fractions
Fractions are numbers—positive or negative—with one or more decimal places. When writing
fractions, make sure the total number size, including the decimal point, is less than eleven characters
in length. As with integers, do not store commas or special characters in your fraction.
Logicals
A logical is simply the word TRUE or FALSE (case-insensitive). Some instructions, such as IF-THEN
statements, use these words to perform conditional operations. Other instructions produce these
words. For example, the tag ISFILE returns a TRUE or a FALSE depending on whether it finds a
specified file. Consider the following instructions:
myimage="someimage.gif"
IF ISFILE(myimage)="TRUE" THEN
DISPLAY "<img src=" + myimage + ">" /DISPLAY
ELSE
DISPLAY "<img src=default.gif>" /DISPLAY
/IF
The IF-THEN conditional runs what's within the THEN and the ELSE if the test between the IF and
the THEN is true. If it is not true, it runs the instructions between the ELSE and the /IF. Note that the
test placed between the IF and the THEN could also be written as follows:
<< myimage="someimage.gif"
IF ISFILE(myimage) THEN
DISPLAY
"<img src=" + myimage + ">"
/DISPLAY
ELSE
DISPLAY
"<img src=default.gif>"
/DISPLAY
-55-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/IF
>>
Here the test is simply a calculation that returns the value TRUE or FALSE. You can write IF-THEN
tests both ways. Read more about Conditions in the accompanying "The Conditional" note.
The Conditional—Like standard IF-THEN conditionals, HTML/OS IF-THEN statements run
instructions based on a test appearing between the IF and the THEN. The tests can be calculations
that calculate to TRUE or FALSE, or they can be tests that compare calculations to other
calculations. The following comparisons operators can be used:
ƒ
=
Equal
<
Less than
!=
Not equal (Same as <> operator)
<=
Less than or equal
>
Greater than
~
Begins with
>=
Greater than or equal
~~
Contains
You can combine multiple tests together in a Boolean test to build more complex tests. For
example, consider the following IF-THEN statement:
IF (balance > 0 AND user_level >= 3) OR super_user="YES"
THEN
Your Instructions
/IF
It performs Your Instructions conditionally—if super_user is equal to YES or if the user's
balance is positive and have a user_level of 3 or above. When combining tests, use the words
AND, OR, or NOT, placing parentheses around each test.
Dates and Times
You must write dates and times in a specific format or they will not be interpreted correctly. In
general, they are written as: MM/DD/YYYY HH:MM. The MM stands for a two-digit month. DD stands
for a two-digit day. YYYY stands for a four-digit year. If the time HH:MM is dropped, midnight (00:00)
is assumed. If the day MM/DD/YYYY is dropped, 01/01/2000 is assumed. This date format is the only
date format used in HTML/OS. Use it when working with cookies and databases, and when
performing date and time calculations. Date calculations work on dates going back to 1970.
Overlay tags are available for displaying dates and times in multiple formats; but when stored they
should be in the format HTML/OS uses. For example, consider the following Overlay:
<< expire_date = ADDDAYS(TODAY,5)
DISPLAY
"Your demo will expire " + GETDATE(expire_date,"long")
/DISPLAY
>>
-56-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The expiration date is stored properly; but formatting is applied when you display the variable. The
GETDATE tag used here formats the number in a long format, meaning that a date, such as
05/01/2002, would display as May 1, 2002.
Y2K Convention—A Year 2000 convention in HTML/OS helps with two-digit year ambiguities.
When specifying a two-digit year less than 20, the current millennium is assumed. When greater
than 20, the millennium beginning with 1900 is assumed. Note that this Y2K convention will be
changed in future releases of the product to enable use of 2-digit years for dates after 2020. If you
are referencing dates in the distant past or distant future it is best to use 4-digit dates.
Global Variables
Variables in HTML/OS carry their values from Web page to Web page. For example, if you write the
instruction myname="Yoshi" on one Web page, five Web pages later, myname will still be equal to
Yoshi. On the Web, this is called state-persistence. In off-the-Web environments, this is called a
global variable environment. Whatever you call it, variables in HTML/OS automatically save from
page to page.
The exception to this rule is the case of local variables, which are variables that you can define
inside functions. Local variables persist while inside the function, but disappear outside of it. Local
variables give programmers the ability to build advanced algorithms. This topic is beyond the scope
of this book; but it is useful to know local variables are available to you if you need them. For
information on setting local variables in functions see the knowledge base on the Aestiva Web site
at http://www.aestiva.com/support/.
Clearing Variables
To clear a variable, set it to ERROR. Clearing variables is useful if you are concerned about the total
size of your variable environment. Since variables pass from page to page, they are saved on the
Web server as users leave and return to the site. The larger the variable environment you use and
the more users you have visiting your site, the more hard disk space you need. It is a good idea to
clear variables larger than 50,000 bytes if you do not need them. If you clear a variable before the
end of a page, HTML/OS won't store it.
State Persistence Across the Network
Variables in HTML/OS save from page to page, even as page control moves across a network. For
example, suppose you write the instruction myname="Yoshi", and the user moves from one copy
of HTML/OS to another, which is known as server jumping. Then myname will still be equal to Yoshi.
Server-jumping and other topics relating to Web-networking are discussed in Chapter 15, Database
Networking and Chapter 16, Distributed Systems.
Using Variables in Calculations
Calculations in HTML/OS can use operators to perform mathematical calculations. Operators include
the four main math operators, plus (+), minus (-), multiply (*), and divide (/) and dozens of predefined
Overlay tags for performing math, logical, and geometric calculations. These tags are listed in
Appendix D, HTML/OS Tag Reference Guide. Here are some examples of calculations:
ƒ
ƒ
ƒ
c=SQRT( SQUARE(a)+SQUARE(b))
a=c*SIN(myangle)
cart_subtotal=SUMCOL(cart_products,"4")
-57-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
In the first two examples perform mathematical calculations. The first is the calculation of the length
of a diagonal of a triangle using the Pythagorean theorem. The second is a trigonometric calculation
using the sine function. In the third example, the tag SUMCOL sums the cells in column 4 of the
variable cart_products. This is how, for example, you would sum the price of products in a
shopping cart.
Variables, Literals, and Parameters
As you write your Overlays, sometimes you want to pass data you type into an instruction. At other
times, you wish to reference data stored in a variable. This data may be used as a setting to an
Overlay tag or it may be used in a calculation. Here we review how to specify data. You learn what
variables, literals, and parameters are. Consider the following instruction:
shipping_cost = SUMCOL(cart_products,"4")
The tag SUMCOL has two parameters. The first is a variable. The second is a value, known as a
literal. It is in quotes. When writing tags, it is important to know how to differentiate between literal
values and variables, and how to place them in tags. The following rules apply:
Literal values should always be surrounded with single or double quotes except
when the value is a positive integer, in which case quotes are optional.
A literal value containing a single quote must be quoted with two double quotes. A
literal value containing a double quote must be quoted with single quotes. A literal
value containing both must be broken up into smaller sets of letters and pasted
together. Here are some examples:
ƒ
ƒ
Literal
Value stored in myvar
myvar="Can't You See?"
Can't You See?
myvar="-100"
-100 (negative one-hundred)
myvar='He said,
He said, "Can't You See?"
"' + "Can't you See?"
Variables cannot be surrounded with quotes, or else they'll look like literals.
Parameters of Overlay tags can be variables, literals, or complete calculations. Each
must be separated from the next with a comma. The ability to replace a parameter
with a calculation, which itself may contain other Overlay tags with parameters, is
called nesting. Overlay tags are nestable. For example, you use the tag
REPLACEALL to perform a search and replace on a variable. It has three parameters;
the first is the name of the variable being searched, the second is what you are
searching for, and the last is what you replace it with. To eliminate < characters from
a variable and replace them with &lt; characters (the HTML code for a less than
sign), you write the following code:
ƒ
ƒ
ƒ
myvar=REPLACEALL(myvar,"<","&lt;")
Now suppose you want to replace not only the less than sign, but also the greater than sign.
You could write it as follows:
-58-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
myvar=REPLACEALL(myvar,"<","&lt;")
myvar=REPLACEALL(myvar,">","&gt;")
Or, taking advantage of nesting, you could write the following:
myvar=REPLACEALL(REPLACEALL myvar,"<","&lt;"),">","&gt;")
Using Variables in HTML
HTML/OS variables are also called names. The nomenclature is borrowed from the use of the word
name in HTML. For example, the HTML input tag can contain name-value pairs, as in the following
example:
<input type=text name=country value="USA" size=15>
HTML/OS is an integrated environment, meaning the names in HTML forms and the variables in
HTML/OS are related. HTML/OS variables fill values in HTML forms. Names in HTML forms, when
submitted to the server, fill HTML/OS variables. It's automatic. Suppose, for example, an HTML form,
like the following one, is stored on the Web server:
<html>
<< IF firstname="ERROR" THEN
firstname="John"
Lastname="Smith"
Gender="M"
/IF
>>
<form ACTION=<<page>>>
First Name: <input type=text name=firstname size=15><br>
Last Name: <input type=text name=lastname size=15><br>
Gender: <select name=gender><option value="M">Male<option
value="F">Female</select>
</form>
</html>
Before the Web page runs, the variables have the default value ERROR, so when the page runs, they
are set to initial values as specified in the IF-THEN statement at the top of the file. The HTML form
takes on the values of those variables. When the HTML page is provided to the user, the values
appear in the HTML form. The page transmitted to the browser contains the initial values even
though you never needed to program them into your HTML form. Run the page and, using the
options in your browser to view the HTML page as seen by the browser, click View and then,
depending on your browser, Source or Page Source. Note how the name-value pairs appear in the
HTML source. The page rendered by this code and its source are shown in Figure 6.2.
-59-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 6.2: The View Source Option of Your Browser Allows You to View the HTML-Only Document
Rendered by HTML/OS.
The automatic filling of HTML form elements gives you the ability to integrate HTML forms into the
applications you write. It also allows you to span HTML forms across multiple documents. For
example, suppose you have a lengthy form that would be better off spread across multiple
documents. Divide the form into multiple forms, placing each of the smaller forms on a single page.
Then add Next and Previous hypertext links to the pages to give users the ability to navigate
between the pages containing the forms. The automatic filling of HTML forms by HTML/OS ensures
the forms still retain their information and behave like HTML forms placed on a single page.
It is a good habit to avoid setting values in HTML forms explicitly. If you need to place a value in an
HTML form component, set the value prior to displaying the form. If you do specify a value in an
HTML form, the HTML form will show the value as expected. But the HTML/OS variable will not
change to the value specified in the HTML form until it is submitted. This topic is discussed in greater
detail later in this book. See the "Further Reading" note.
Further Reading—The subject of HTML forms is such an important topic in the world of advanced
Web sites that we dedicate an entire chapter to the subject. See Chapter 10, HTML Forms
Processing, for an in-depth look at the world of HTML forms.
Working with Delimited Text Files
Delimited text files are text files containing data organized in columns and rows. Most often, the files
are exported from databases and spreadsheet programs. The character used to separate one cell
from the next is called the delimiter. End-of-line characters are used to separate one row from the
next. The following example is a delimited text file containing the state codes and names of five U.S.
states.
"CA","California"
"NY","New York"
"WA","Washington"
"VA","Virginia"
"OR","Oregon"
-60-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
As you see, the delimiter is a comma. The character surrounding each piece of data is called the
quoting character. Here the quoting character is a double- quote. Almost all database programs and
spreadsheets give you the ability to export your data to some kind of delimited text file. Most often,
the delimiter is a comma or a tab character. Sometimes the cells of each row in the text file are
surrounded with quote marks, sometimes not. To load a delimited text file into a variable use the
following instruction. (We have assumed your delimited text file is called /mystates.txt.)
COPY FILE="/mystates.txt" TS="," TO mydata /COPY
This will load the five states in /mystates.txt into mydata, creating a two-column by five-row
variable. The TS parameter (TS stands for Table Separation.) is a comma that tells HTML/OS the file
is comma-delimited. (For more information, see "The TS Parameter" note in this section.)
The TS Parameter—The TS parameter can contain between one and three characters. It can
contain the delimiter, the delimiter and a quoting character, or a delimiter, quoting character, and
an "escape" character. Most often you need to define only a delimiter, as shown in the example
used in this section. In special cases, you can add other characters. The nuances of dealing with
special cases are beyond the scope of this book. For further information, access the Knowledge
Base on the Aestiva Web site at http://www.aestiva.com/support/.
AM
FL
Y
If you do not set the TS parameter, the file will be loaded into mydata like any other text file.
Remember to specify the TS parameter. Otherwise, the COPY tag will create a single-column by
single-row variable containing the entire contents of the file.
You can also use the COPY tag to create delimited text files. For example, suppose you wish to load
the comma-delimited file shown previously and then save it to a new tab-delimited file, with the same
data. You could write the following two instructions:
COPY FILE="mystates.txt" TS="," TO mydata /COPY
TE
COPY mydata TO FILE="mystates2.txt" TS=TAB /COPY
The first COPY instruction loads the comma-delimited text file into mydata. The second takes the
variable and saves it to the new file mystates2.txt.
Working with Loops
Anyone familiar with high-school BASIC is familiar with WHILE and FOR loops. The same kinds of
loops work in HTML/OS too. WHILE loops allow you to repeat a set of instructions while a particular
condition is TRUE. FOR loops repeat sets of instructions while a predefined variable (with one row
and column) steps between two values.
When working with HTML/OS, you have the same loops; but you also have a FOR loop that repeats
sets of instructions as you move across the different rows of a variable. This last kind of loop is the
most popular loop that HTML/OS developers use.
In this section, we'll review standard WHILE and FOR loops but spend most of our time explaining the
FOR loop which is used most often.
WHILE Loops
Suppose you want to write a set of instructions that copies files in a folder to a backup folder. You
could use a WHILE loop that uses the following algorithm. First, you generate a list of files to be
-61Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
copied. Then, one at a time, you copy each to the backup folder. After each copy, you delete the file
copied from a list of files that remain to be copied. You do this until no files are left. The instructions
could be written as follows:
myfilelist = SYSLS("/folder1")
myfile_in = "/folder1/" + myfilelist[1]
myfile_out ="/folder2/" + myfilelist[1]
WHILE ISFILE(myfile_in) DO
COPY FILE=myfile_in To FILE=myfile_out /COPY
myfilelist=DELROW(myfilelist,1)
myfile_in = "/folder1/" + myfilelist[1]
myfile_out ="/folder2/" + myfilelist[1]
/WHILE
The first statement uses the Overlay tag SYSLS to place a list of filenames in /folder1 into
myfilelist. SYSLS is a predefined tag in HTML/OS that creates a variable with multiple rows,
one row for each file found in a specified folder. Filenames are placed in column one. For additional
information on SYSLS, see Appendix D, HTM/OS Tag Reference Guide.
The next two instructions build filenames for input and output files respectively. The plus (+) sign
pastes text together. The WHILE statement contains the test ISFILE(myfile_in) so it will run the
instructions between the DO and the /WHILE as long as myfile_in contains the name of a file that
exists.
In the WHILE loop, the COPY tag copies myfile_in to myfile_out. The tag DELROW removes the
first line of myfilelist; then myfile_in and myfile_out are set again using the first row of
myfilelist, and the loop is repeated. For additional information on COPY and DELROW, see
Appendix D, HTM/OS Tag Reference Guide.
As you see, WHILE loops in HTML work identically to those used in other languages. The
instructions between the DO and the /WHILE execute repeatedly, until such time that the test
between the WHILE and the DO is no longer true—at which time, execution shifts to the next
instruction following the WHILE tag.
FOR Loops
Another way to perform the operation shown in the previous code is to use a FOR loop. FOR loops,
like you find in high-school BASIC, are not popular in HTML/OS. They are clunky for the programmer.
You write them as follows:
FOR NAME=variable_name VALUE=begin_value TO end_value
STEP=step_value DO
instructions here
/FOR
These old-style FOR loops require that you determine beginning and ending points of the loop, which
is cumbersome and can lead to errors. We will not use them here. Although HTML/OS supports
these old-style loops, the more popular FOR loop is written as follows:
-62-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
FOR NAME=variable_name ROWNAME=row_name DO
instructions here
/FOR
This style loop takes advantage of the two-dimensionality of HTML/OS variables. To perform a loop,
you need to specify only the name of a variable (to loop across) and the name of a variable to fill with
a slice of that variable, as you loop across it.
You do not need to calculate beginning or ending values, because the loop knows to run once for
each row of your specified NAME. Each time the FOR loop loops, it places the next row of your
specified NAME in the specified ROWNAME.
Once again, suppose you wish to write instructions to copy files in a folder to a backup folder. Using
a FOR loop, you would write the following code:
myfilelist = SYSLS("/folder1")
FOR NAME=myfilelist ROWNAME=myfilerow DO
myfile_in = "/folder1/" + myfilerow[1]
myfile_out ="/folder2/" + myfilerow[1]
COPY FILE=myfile_in To FILE=myfile_out /COPY
/FOR
This example uses SYSLS, once again, to create a two-dimensional variable called myfilelist.
The FOR loop fills myfile with the next row in myfilelist each time through the loop. Within the
loop, you set myfile_in and myfile_out using the COPY tag; the file myfile_in is copied to the
file myfile_out.
As an example, suppose "/folder1" contained the three files albert.txt, bobby.txt, and
cindy.txt. SYSLS would create myfilelist, a variable that would look as follows:
albert.txt 22916 07/02/02
bobby.txt 12027 06/30/02
08:37:33
15:00:47
FILE
FILE
cindy.txt 34200 07/05/02 18:07:10 FILE
The first time through the FOR loop myfilerow looks like this:
albert.txt 22916 07/02/02 08:37:33 FILE
The second time through the loop myfilerow looks like this:
bobby.txt 12027 06/30/02
15:00:47
FILE
These FOR loops work by filling the variable specified by the ROWNAME parameter with a one-row
slice of the variable specified by the NAME parameter. Each time through the loop, the FOR loop fills
the variable specified in the ROWNAME parameter with the next row. The instruction
myfile_in="/folder1/"+myfilerow[1] (remember that myfilerow[1] is the same as
myfilerow[1,1]) calculates the source file, myfile_in. The instruction immediately after this
one calculates the destination file, myfile_out. That instruction is followed by the instruction that
actually copies the file.
-63-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Summary
In this chapter, you learned about variables, how they work with HTML forms and how to use IFTHEN conditionals, WHILE loops, and FOR loops. They are the fundamental building blocks of
Overlays, which you learned how to place inside HTML documents in Chapter 5, Underlays, Inlays,
and On-Click Overlays. You now have a solid foundation in HTML/OS programming.
The remaining chapters in Part II, Programming Basics, are dedicated to applied issues. The next
chapter covers debugging techniques—an important topic for the user of any development
environment. If you have been going through the code here, typing it and running it, you have
already encountered little problems and error messages along the way. The next chapter will shed
light on this extremely practical topic and provide valuable advice on how to diagnose and solve
development problems as they arise.
Exercises
The following exercises are designed to accustom you to working with two-dimensional variables,
IF-THEN statements, and FOR loops. Answers to all exercises are provided on this book's
companion Web site as described in the book's Preface.
Exercise 1
Create a file counter. When the user reaches the file counter page, see if the file contains a number.
If not, set it to 1. If it does, increment it. Display the counter value on the screen along with a
Redisplay link. You will need to use the COPY and ISINTEGER tags, and an IF-THEN statement.
Exercise 2
Create a page that displays a random message at the top of the page. To do so, each time the page
loads, copy a file containing multiple messages, one per line, into a variable. Then randomly select a
line from the variable and display it. You will need to copy from a delimited file. (Note that delimited
text files with only one column still need a delimiter. Set your delimiter to any character not in the file,
such as a vertical bar.) You will need to use the RANDOM and COPY tags. To calculate the number of
rows in a variable, use the tag ROWS.
Exercise 3
Use a FOR loop to read a list of HTML pages in a folder and display a list of HREF links. When users
click a link, they should be directed to the page.
Exercise 4
Use a FOR loop to read a list of files in a folder and display a pull-down select box with the filenames.
Add a Submit button so, when users select a file they are directed to the page.
Exercise 5
Use a FOR loop to read a two-column file containing names of states (like the example shown in this
chapter). Using a FOR loop, display a pull-down select box on the page. Hint: Use the FOR loop to
display multiple <option>..</option> lines, one for each state. Use standard HTML to display
the <select> and </select> tags.
-64-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 7: Debugging Techniques
Overview
The programs presented in this book illustrate many facets of Web development. Although we hope
you find these programs helpful, they are not a substitute for the experience gained by sitting down
and using your copy of HTML/OS to program Web pages. The experience of writing a program, line
by line, is different from looking at how programs are written. In sample programs, you see
intermediate results and the end product—programs that run without errors and behave as intended.
During development, however, programs are not likely to behave exactly as planned. You rarely
write them correctly the first time. The most experienced programmers make mistakes in the first
writing of their programs. In fact, making mistakes is an accepted part of the process of writing Web
applications—or any computer programs, for that matter. The process of building a program is
intertwined with the activity of fixing mistakes.
As you typed the example Web pages in this book into your Web-based editor, you most likely
encountered mistakes. You may have encountered parsing errors reported to you on the Web page
when you ran these example programs.
You may have used these reports to fix the problem, find the missing line, or correct the misspelled
word. In a Web page with only 10 or 20 lines, mistakes are easy to identify. But as you write longer
and longer pages, the potential to make careless typing errors and mistakes in your logic increases.
Finding the mistakes also becomes more difficult.
In general, you test and debug advanced Web pages, like those discussed in this book differently
than static or pure HTML pages. When working with pure HTML, you are mostly interested in the
"look" of the page. You need to see if the look works in the popular browsers on the Web. This is
called cross-browser compatibility testing. While you must still pay careful attention to cross-browser
issues, it is not the topic of this chapter. In this chapter, you learn about debugging the functionality
of Web pages.
As you test Web pages you want to give them a wide range of input to make sure they run as
intended. If a page comes up with an error or does not behave as intended, you must debug it.
Debugging is the process of identifying errors and correcting them. This chapter reviews commonly
used debugging techniques. You also learn about the automatic error messaging provided to you by
HTML/OS and you learn about a program called Tracer for tracking down programming bugs.
The beginning of this chapter explains the kinds of error reporting provided by HTML/OS.
Errors Reporting
The HTML/OS engine issues three kinds of error reports; full-page error reports, in-page error
reports, and HTML/OS tag-specific error results. Full-page and in-page error results are also known
as parse errors. Tag-specific error results are placed in variables by HTML/OS tags in response to
improper data. The following sections describe the different kinds of errors and reports.
Parse Errors
Parse errors occur as the HTML/OS engine parses, or interprets your document. Deciphering parse
errors reported to you on Web pages is easier if you understand how HTML/OS parses Overlay tags.
HTML/OS scans Overlay tags from left to right and immediately stops interpretation when it
-65-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
encounters an error—that is if HTML/OS is unable to continue. In some cases however, it can
continue its interpretation.
If a tag is misspelled or HTML/OS does not recognize it, it will assume you have an undefined tag or
function and stop its interpretation. If you write your tag properly but place an incorrect number of
parameters in the tag, HTML/OS reports it to you and continues its interpretation.
Full-Screen Reports
When an error occurs and HTML/OS is unable to render the HTML document, it issues a full-screen
error report. For example, consider the following page, which is a simple calculator. The first IFTHEN statement has no THEN, so HTML/OS is unable to render the page.
<html>
<title>Five-Line Math Calculator (with Error)</title>
<<
IF ISNUMBER(a+b)="FALSE"
z=1 a=0 b=0 result="?"
/IF
IF z=1 THEN result = a + b /IF
IF z=2 THEN result = a - b /IF
IF z=3 THEN result = a * b /IF
IF z=4 THEN result = a / b /IF
>>
<form method=post action=<<Page>>>
My First Calculator:<br>
<input type=text name=a size=5>
<select name=z>
<option value="1">+
<option value="2"><option value="3">*
<option value="4">/
</select>
<input type=text name=b size=5>
<input type=submit value=" = ">
<b><<result>><br>
</form>
</html>
Running this page produces the screen shown in Figure 7.1. A similar screen would be reported if,
for example, you forgot an IF or if the tag misspelled ISNUMBER
-66-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 7.1: Full-Screen Error Reports Occur When HTML/OS Cannot Render a Web Page.
The disadvantage of full-screen error reports is that they do not explicitly identify the location in the
file where the error occurs. Usually however, using the error report, you can identify the location by
the nature of the problem reported. If this does not work, you can use other methods to identify and
fix the problem. One method is called commenting out code, a topic discussed in the "Commenting
Out Code" section later in this chapter. Another method is to use a program called Tracer. The
program allows you to mark a document, so you can find the last instruction in the document that ran
before it failed. Tracer is discussed in the section "Using Tracer," located at the end of this chapter.
In-Page Reports
Not all errors cause full-screen reports. Often, when an error occurs, HTML/OS is still able to render
the HTML document. In these cases, HTML/OS issues an in-page report. In-page reports are text
messages surrounded by [[ and ]] characters. For example, suppose in our previous program you
mistakenly wrote the instruction
IF z=1 THEN result = a + b /IF
as
IF z=1 THEN result - a + b /IF
Note the second equal sign in the instruction is a minus sign. Running this instruction renders the
Web page screen shown in Figure 7.2. The page includes the in-page error report:
-67-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 7.2: HTML/OS Provides In-Page Error Reports When it Can Render the Web Pages Despite the
Error.
[[HTML/OS Error: parse error 1 a=0 b=0 result="?" /IF IF z=1
THEN result -]]
When encountering this kind of error report, read it right to left. The last character of the message is
a minus sign. That's where HTML/OS failed to interpret the instruction. The minus sign should have
been an equal sign.
Use the message inside brackets ([[ ]]) to identify the location of the error in the file. The last
characters of the message tell you exactly where HTML/OS could no longer interpret the instruction.
The left most part of the error report, HTML/OS Error parse error, provides a general
description of the error HTML/OS encountered.
Tag-Specific Results
Not all errors cause error reports. If, for example, a calculation is out of range or has an incorrect
input, the most likely result is for HTML/OS to return the value ERROR as a result for your calculation.
The following calculations, for example, place the value ERROR in myresult:
myresult = 5 / 0
myresult = 12345678 * 12345678
myresult = ROUNDUP("Two Hundred")
myresult = ADDDAYS(TODAY,"Two")
myresult = LEFT("Hello World","-9")
The first two examples produce numbers that are too large, so HTML/OS returns ERROR. The third
and fourth examples contain text parameters that should be numbers. Again HTML/OS returns
ERROR. The last example contains a parameter that's a negative number but it requires a positive
number. If you get a value of ERROR in a variable, it's likely to cause errors in other variables based
on that variable. If you find a variable with the value ERROR when it should have some other value,
work your way back in the code until you find the first HTML/OS tag that caused a value to come up
as ERROR.
-68-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
TAGRESULTS—Many HTML/OS tags place additional error messages in a special variable called
TAGRESULTS. If an HTML/OS tag does write to TAGRESULTS, it uses the following convention; it
places one of the values—TRUE, FALSE, or ERROR—in column 1, row 1 of TAGRESULTS. It also
places an English-like error message in column 1, row 2 of TAGRESULTS (or places the word OK in
this cell if it finds no error). It fills other cells of TAGRESULTS with supplementary information
depending on the specific tag. Consult your HTML/OS Reference on the specific tag to see
whether to fill TAGRESULTS and what information to place in each cell.
Debugging Your Code
The techniques used to hunt down a problem vary according to the problem you encounter and the
kind of error message you get. For example, if you run a Web page and see a blank screen, you
would follow one set of steps to debug the problem. If you run a Web page and see a full-screen
error report, you would follow a different set of steps. If you get an in-page error report, you would
follow yet another path to hunt down the error.
The best way to debug is to debug little at a time. If you find yourself encountering many bugs as you
write your code, write smaller amounts of code at a time before you test the code on the page. The
Web-based editor gives you a single-click Save/View button. Use it to test your code as you program.
If you're a beginner, it may make sense to write only four or five lines of code at a time, each time
viewing the page, fixing the problems that arise, and then returning to write additional lines of
instructions. When you encounter the bugs, you should adhere to the rules outlined in the following
sections.
Verify and Repeat
Regardless of how you hunt down a bug, you should avoid jumping to conclusions. When you
encounter a bug or problem, the first thing to do is repeat the bug so you understand the
circumstances that give rise to it. If your bug is not repeatable, edit your page until it is. Problems
that are not repeatable (also known as intermittent) are difficult to solve.
Location, Location, Location
Once you have a repeatable problem, you need to identify where the problem occurs in the code.
This step is almost always a matter of narrowing down the general location of the error until you can
isolate its exact location. Then, and only then, can you move on to find out why the problem occurred.
This is the most important step in resolving any bug. In fact, this is what good debugging is about. It
doesn't make sense to rethink your entire program. That would be too much to wrap your brain
around. It makes sense to rethink only that portion of the program where the mistake occurs.
Viewing Document Source
If you are having difficulty localizing your bug, try looking at the HTML source of the document. It
may help you tell whether the problem is HTML, code, data, or logic. Sometimes HTML/OS error
messages get stuck inside HTML brackets and are not visible until you view the source of the page
as the browser sees it. For example, suppose your Web page has a text box that is not rendering
properly. Suppose the code is as follows:
<textarea cols=<<mycols>> rows=<<myrows>> name=mytext ></textarea>
A look at the HTML document source in your browser might uncover the following line:
<textarea cols=5 rows=ERROR name=mytext ></textarea>
-69-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
This tells you that <<myrows>> rendered to the value ERROR. Now you know
to look in your instructions to see why the variable myrows is equal to
ERROR.
Viewing the actual HTML text rendered by HTML/OS gives you more clues than simply viewing the
page rendered by the browser.
The Bug Hunt
Start your hunt for the location of a bug by determining the exact page causing the problem, and
whether the problem is in your HTML or in an Overlay. If the problem is in an Overlay, you need to
determine which Overlay or HTML tag is causing the problem. Keep in mind bugs are not always
caused by an incorrect use of an Overlay tag. That's just one possible cause. Bugs are also caused
by problems in the following areas:
ƒ
Your data— If your data is the cause of a bug, narrow down the specific data that's
causing the problem. After you do this, go back and see how that particular data is
handled.
ƒ
Poorly deployed Overlay tag— If the problem is in the way an Overlay tag is used,
investigate the behavior of the Overlay tag to see exactly how it works.
ƒ
Incorrect algorithm— If the problem is in an algorithm, step through the algorithm
to verify it is doing what you believe it should be doing.
ƒ
Incorrectly launched Web page— If the problem is in the way you are running
your Web page, look at the values of the variables the Web page uses when it is
first launches to see how that compares with what you expect.
Displaying Intermediate Values
Often, you can locate a bug by viewing the changes in one or more variables used in the Web page.
To track the contents of the variable my_var, for example, sprinkle your page with debugging
instructions like dbg1=my_var, dbg2=my_var, dbg3=my_var, and so on. Then, at the bottom of
the page, just above the </html> tag, display the dbg variables with instructions like
dbg1=<<dbg1>><br> and dbg2=<<dbg2>><br>. Then rerun the page. Compare the values of
these dbg variables with what you expect. When you find values that don't match your expectations,
use the locations of the debugging instructions to narrow down the location in your code causing the
problem. Once you locate the problem delete the debugging instructions.
Commenting Out Code
In some cases, you comment out code as a way to locate a problem. For example, suppose some
variable has the value ERROR, but you do not know why. You know the variable is calculated in a
large section of code. Try commenting out this code by using the # and /# tags. Placing these tags
around instructions tells HTML/OS to ignore them. Remember that these tags only work inside
Overlays. If you need to comment out HTML tags, surround them with <!- and -> tags.
If commenting out code fixes the bug, reduce the amount of code you comment out and try again.
Repeat this process until the commented out code is a small segment that identifies the location of
the problem
-70-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
.
The Mind Is a Funny Thing—Debugging a problem is often a battle between your mind and your
program. The time you spend debugging can be frustrating, because most problems are the result
of careless mistakes and errors in logic. But writing a program without making at least a few
mistakes is a rarity.
Most often, mistakes happen because it doesn't make sense to be 100% thorough when you write
a program. It's usually faster to write a program the best you can and then go back and debug it.
Debugging a program is a mind game you must play with yourself. During the debugging process,
you must rethink some of the steps you took when writing the program. If you skip over the steps
that got you into trouble in the first place, you won't be able to solve your problem.
When you are debugging, your mind might repeatedly skip over the piece of information causing
the problem, because you think something is obvious when it is not. Do not trust yourself too much
when debugging. Use systematic approaches rather than hunches and guesses. At the very least,
you should use systematic approaches when other methods fail.
Using Tracer
AM
FL
Y
You can solve most programming problems by methodically identifying the location of the problem in
your code. In some circumstances however, you may encounter a particularly evasive bug. That's
when you may want to use the program, Tracer.
TE
Three types of bugs warrant the use of this program. First is the bug that causes an unhelpful fullpage error report. Without guidance, if your Web page contains a lot of programming, it may be
difficult to narrow down the problem. The second kind of bug results in a page crash, which gives
you no error report at all, because the bug is so severe it crashes HTML/OS, and HTML/OS is
unavailable to issue the error report. (See the accompanying "Automatic Crash Recovery" note.) The
third kind of bug is one in the logic of your programming. Here, the ability to follow the values in
variables as they change during program execution comes in handy.
Automatic Crash Recovery—It is not easy to crash HTML/OS, but it can be done. If a crash
occurs, you'll see an error report by the Web server. You won't see a full-screen message centered
in the middle of a white background (that would be an HTML/OS full-screen error report).
If this happens, HTML/OS automatically recovers. The next time you run HTML/OS, you'll
experience a 10-second delay. After that, the pages will begin working again. If after the crash, you
experience peculiar behavior, log out of your copy of HTML/OS and log in again.
In all three of these instances, you have little idea about the cause of the problem or where the
problem occurred in your page. One way to solve such a problem is to comment out code, as
explained in the "Commenting Out Code" section earlier in this chapter. Another way is to use the
Tracer program.
Tracer is a trace file analyzer. Using Tracer requires you to place TRACE tags in your Web page, run
the page, and then view the trace logs created by the TRACE tag with Tracer.
Tracer is also convenient when you wish to track down a logic problem. In such a case, you want to
follow the values in multiple variables as HTML/OS makes its way through your instructions.
Although you can display intermediate values in your code (a technique described in the "Displaying
Intermediate Values" section earlier in this chapter), you can also use Tracer, which enables you to
view a history or trace of the values in variables at the time the page was interpreted by HTML/OS.
-71Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The Trace Console
Tracer is included with your copy of HTML/OS. You'll find it in the Utilities menu on the HTML/OS
desktop. Figure 7.3 shows Tracer's main screen. The options bar, located at the top of the screen,
has links to clear the trace log (see the accompanying note, "The Trace Log"), show or hide page
code, and launch an on-line help utility. The name of the trace you are currently viewing is
immediately under the Options bar. In the upper-right corner, Forward and Back arrows enable you
to move among the different traces within the log.
Figure 7.3: The Main Tracer Screen Includes a Trace View with Many Controls.
The Trace Log—The trace log contains multiple traces. Each time you run a page with TRACE tags,
a new trace is placed in the log. The trace log itself is designed not to exceed 100,000 bytes. If the
file exceeds this limit, HTML/OS removes traces from the file to make room for new traces.
The main screen is divided into two sections. To the left is a view of a trace. Here, using the various
options provided, you can control the view of the data in the trace log file for a particular trace. To the
right is the source code of the page you traced.
The Trace view, appearing on the left side of the screen, is a grid containing all or some of the
information placed in a trace by a Web page containing TRACE tags. Trace entries appear as rows
ordered chronologically. The top row contains the first log entry made. In general, a row in the Trace
view can indicate the following problem spots:
ƒ
ƒ
ƒ
Reading from a variable
Writing to a variable
Running a TRACE tag (flag)
Tracer places reads and writes to a variable in a single column for that variable. This allows you to
follow the reads and writes to a single variable by reading down a column. Not all variables are
listed—only those you wish to follow. To select a set of variables to follow, use the Variables option
-72-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
in the Options box at the top of the Trace view. Place check marks next to the variable names you
want to follow and click the Show button.
If you wish to hide the reads or writes from Trace view, click the Hide Reads or Hide Writes links.
These links toggle back and forth between Hide and Show. Tracer also provides a Hide Flags and
Show Flags toggle link. Click this link to hide or show lines that display flags set by the TRACE tag.
Figure 7.3 displays a 16-line trace. Each row is preceded by a single-letter code. W means write. R
means read. Rows preceded by a flag indicate a TRACE flag has been run. In the next section, you
learn how to write TRACE tags to generate these kinds of logs and how to use them to debug code.
Using the TRACE Tag
Placing TRACE tags in different locations of a document instructs HTML/OS to write log entries to a
trace log file as it runs the page. The TRACE tags are placed inside Overlays and are written as
follows:
temp=TRACE(parameter)
Here parameter is either a reserved-word the TRACE tag understands, or a text message, also
known as a flag. When you execute TRACE, it does one of two things; if the parameter is a text
message (a flag), the text message is added to the trace log. If the parameter is a reserved-word,
TRACE performs the operation defined for the reserved-word. The reserved-words are as follows:
ƒ
Ignore—Ignores subsequent TRACE tags in document. The Ignore parameter is
useful if you place a number of TRACE tags in a document but want to deactivate
them without having to delete them from the document. In this event you place the
line temp=TRACE("ignore") at the top of the document.
ƒ
Clear—Deletes the trace log file. The Clear parameter is useful if you have too
many trace entries in a log. It is usually a good idea to clear your trace at the top of
the page.
On—Turns on variable tracing. The On parameter turns on the logging of reads and
writes to all variables. Note that tracing still occurs if you do not turn on logging.
TRACE tags containing text messages (flags) will still be logged; but log entries will not
be written when variables are read and written to.
Off—Turns off variable tracing. To turn off the logging of reads and writes to
variables, use the Off parameter.
ƒ
ƒ
ƒ
Buffer—Buffers logging and starts system-time recording. Set the Buffer
parameter when you want to use Tracer for an entirely different purpose—to measure
how long it takes to run different tags in your program. When you set this parameter,
HTML/OS stores log entries internally before writing them to the trace log file. It also
logs the computer system time with every entry. You buffer log entries because
otherwise, the time delays created by TRACE itself could render the time
measurements useless.
Finding the Last Point of Execution
If you encounter a crash or a full-page error report, you need to know the last point of execution in
your code before execution failed. To find this location, use TRACE with text messages (flags).
Working your way from top to bottom, place lines like temp=TRACE("Point A"), and
temp=TRACE("Point B") in your program. Run the page so it produces the same error as before.
Pull up the log report in Tracer and view the trace. You will see flags like Point A, and Point B
-73-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
appear on different rows, preceded by a graphic flag. The last flag you see is the last point of
execution reached.
As an example, consider debugging the Hello World program that follows this paragraph. We
purposely placed an error in the page. Running the page produces a full-screen error like that shown
in Figure 7.1. The trace of the page, as viewed in Tracer, is shown in Figure 7.4 on the left side of
the screen. The following code, with TRACE tags is shown on the right.
Figure 7.4: Here's a Trace of a Hello World Program Populated with TRACE Flags.
<<
# set defaults /#
IF size=0 THEN
size=RANDOM(1,5)
/IF
IF size=1 THEN color="blue"
ELSEIF size=2 THEN color="red"
ELIF size=3 THEN color="green"
ELIF size=4 THEN color="black"
ELSE color="purple"
/IF
>>
<html>
<title>Hello World</title>
<font color=<<color>> size=<<size>> face=arial>
<b>Hello World</b></font>
<form method=post ACTION=<<Page>>>
<select name=size>
-74-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<option value=0>RANDOM
<option value=1>Tiny
<option value=2>Small
<option value=3>Medium
<option value=4>Large
<option value=5>X-Large
</select>
<input type=submit value="Go">
</form>
</html>
The last flag shown is Flag #3, which appears just before the IF-THEN statement. It's no wonder—
there's a mistake immediately after this. An ELIF tag was mistakenly written as ELSEIF.
Following Changes in Variables
To follow the values of a variable as your page is executed turn on variable logging at the top of the
document. To follow the contents of variables in a specific part of the page, turn on variable logging
at the top of the section and turn it off at the end of the section. At the top of the section, write
temp=TRACE("on"). At the end, write temp=TRACE("off"). Run the page and then, in Tracer,
pull up the trace. Select the variables you wish to follow at the top of the Trace view and click the
Show button. Looking down a variable column tells you how the variable changed as the page ran.
Compare this to the code shown on the right side of the page. Using this information, check the logic
of your program until you understand why your program did not run as you intended.
For example, consider debugging the same Hello World program in the previous section. Suppose
you wish to see if the variable size changed in the first IF-THEN statement. Using your Web-based
editor, type the TRACE tags shown on the right side of Figure 7.5 into the Hello World program. Run
the page. Its trace appears on the left side of this figure. Looking down the size column, you see
two entries, a read and a write. First size was 0. Then it was changed to 4.
Figure 7.5: Here's a Trace of a Hello World Program with Variable Tracing.
-75-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Summary
In this chapter, you learned the most common techniques involved in debugging a Web page. As you
have seen, debugging is first and foremost about locating the position of the problem in the Web
page. Like they say in real estate, it's all about "location, location, location." Once you know the
location of a problem, it becomes manageable.
In this chapter, you also learned that bugs should be resolved systematically. While shot-in-the-dark
approaches are okay when you first encounter a bug, if your bug is not resolved in a few minutes, it
is best to drop such attempts and become methodical. This technique separates the amateurs from
the professionals.
Finally, with a bit of experience, you'll find debugging is like solving puzzles—it can be challenging,
stimulating, and intellectually rewarding.
Exercises
The following exercises provide buggy Web applications. Using what you learned in this chapter
identify and correct the bugs. Answers to all exercises are provided on this book's companion Web
site as described in the book's Preface.
Exercise 1
The following code has a few bugs. Test, identify, and correct them.
<html>
<title>10-Line "Average" Calculator</title>
Enter number below. We'll calculate a running average.
<form method=post ACTION=dobuttons>
Number: <input type=text name=n>
<input type=submit name=mybutton value="Add To Average">
<input type=submit namr=mybutton value="Clear">
</form>
<< IF msg != "ERROR" THEN
DISPLAY "msg" DISPLAY
/IF
>>
</html>
<<overlay dobuttons
IF mybvutton="Clear" THEN
I=0 SUM=0 AVE=0 LASTNUM=0 N=0
msg="Calculator Cleared. Start Again."
ELSE
I=I+1 SUM=SUM+N AVE=SUM/I LASTNUM=N N=0
msg="Total Entries: "+I+"<br>"+
"Total: "+SUM+"<br>"+
-76-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
"Average: "+AVE+"<br>"
/IF
>>
Exercise 2
The following code is also buggy.Test, identify, and correct the bugs in this application.
<html>
<title>25-Line RPN Calculator</title>
25-Line RPN Calculator<br><br>
Stack: <<mystack[1,1]<br>
<form method=post ACTION=calc>
<input type=text name=n size=8>
<input type=submit name=mybutton value="Enter">
<input type=submit name=mybutton value="Clear"><br>
<a href=myop name=op value=add>Add</a>
<a href=myop name=op value=sub>Subtract</a>
<a href=myop name=op value=div>Divide</a>
<a href=myop name=op value=mul>Multiply</a><br>
<a href=myop name=op value=sqr>Square</a>
<a href=myop name=op value=sqt>Square Root</a>
<a href=myop name=op value=inv>Inverse</a><br>
<a href=myop name=op value=abs>Absolute</a><br>
<a href=myop name=op value=log>Log</a><br>
<a href=myop name=op value=int>Integer</a><br>
</form>
</html>
<<overlay calc
IF mybutton="clear" THEN stack=""
ELSE APPEND stack TO n /APPEND
stack=GETCOLNOTEQ(stack,1,"")
n=""
/IF
GOTO PAGE
>>
<<overlay myop
IF ISINTEGER(n)=FALSE THEN n="" GOTO PAGE /IF
stack1=stack[1,1]
stack2=stack[1,2]
IF op="add" THEN stack2=stack1+stack2 stack1=""
ELIF op="sbt" THEN stack2=stack1-stack2 stack1=""
-77-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ELIF op="div" THEN stack2=stack1/stack2 stack1=""
ELIF op="mul" THEN stack2=stack1*stack2 stack1=""
ELIF op="sqr" THEN stack1=SQUARE(stack1)
ELIF op="sqt" THEN stack1=SQRT(stack1)
ELIF op="abs" THEN stack1=ABS(stack1)
ELIF op="inv" THEN stack1=INVERSE(stack1)
ELIF op="int" THEN stack1=FLOOR(stack1)
ELIF op="log" THEN stack1=LOG10(stack1)
/IF
stack[1,1]=stack1 stack[1,2]=stack2
stack=GETCOLNOTEQ(stack,1,"")
n=""
GOTO "PAGE"
>>
-78-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 8: Building Text Editors
Web-based text document editors are an important component of advanced Web-based applications.
They're deployed most often in the Web-based pages used to maintain a Web site. These back-end
systems give staff the ability to edit Web pages, text files containing configuration settings, and page
headers or footers. Text editors are also used in workflow systems where documents need to be
edited and passed between different parties in an organization. You also find Web-based text editors
in systems that edit templates used in e-mail auto-responders. There seems to be no end to the
situations that require a Web-based text editor.
In this chapter, you start by building a six-line text editor. Then, learn how to integrate it into a backend system and add features to it. As you go through this chapter, we recommend you use your
copy of HTML/OS—experimenting with the code provided.
The Six-Line Text Editor
This text editor is called a "six-line editor" because, well, it takes only six HTML/OS instructions to
build it. This Web-based editor is shown in Figure 8.1. The code for the editor is as follows:
Figure 8.1: You Build the Six-Line Text Editor by using Six HTML/OS Instructions and Three HTML Form
Components.
<<
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
>>
<html>
<b>Six-line Text Editor</b><br>
File Name: <<ted_file>>
<form method=post ACTION=ted_edit>
<table border=0>
<tr><td align=right>
<input type=submit name=ted_button value=Save>
-79-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=65 rows=15>
</textarea>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY
GOTO PAGE
ELSE
GOTO "menu.html"
/IF
>>
The six-line editor uses one assignment (the first instruction), two COPY tags, an IF-THEN statement,
and a GOTO tag. That's five instructions. Oh yes, it also displays the value of ted_file in an
Overlay after the text File Name:—that's a total of six HTML/OS instructions.
The editor has three HTML form elements: a text area and two submit buttons. One button is Save;
the other is Quit. When the page is launched, an Underlay at the top of the document assigns
myfile.html to the variable ted_file. We use ted_file in the six-line editor instead of hardcoding the document name in each Overlay tag, because it's good programming to place those
variables that might change in the future at the top of your document. That way, if you ever need to
use the editor to edit other documents, you can do so by simply changing the first line. Placing
definitions at the top of a document makes them easier to edit.
After this variable assignment, you COPY the contents of file ted_file into ted_text, the variable
assigned to the text area. The text area, where the file is edited, is 65 columns by 15 rows. When a
user clicks a button, the on-click Overlay ted_edit runs. It uses an IF-THEN statement to run
different Overlay tags depending on which button the user the clicks. If the user clicks Save, the text
in the box is saved to the filename stored in ted_file, which is myfile.html, and the page is
redisplayed. If the user clicks Quit, the user is sent to menu.html.
You can use this six-line editor as part of a back-end system—a system designed for staff of an
organization rather than public visitors of a Web site. A minimal back-end system requires the
addition of a login page and a menu of staff options. For example, suppose you want to build a backend system that gives staff the ability to edit a few files. Here's a password page you can use:
<html>
<title>Password Page (password.html)</title>
Staff Access - For Authorized Personnel Only
<form method=post ACTION=be_check>
Enter Password: <input type=password name=be_password size=15>
<input type=submit value="Enter">
</form>
-80-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</html>
<<overlay be_check
IF be_password = "please" THEN GOTO "menu.html" ELSE GOTO PAGE
/IF
>>
Here's a Web page that provides a menu of staff options:
<html>
<title>Staff Menu (menu.html)</title>
Staff Menu - For Authorized Personnel Only<ul>
<li> <A HREF=edit.html>Edit myfile.html</A>
<li> <A HREF=edit2.html>Edit myfile2.html</A>
<li> <A HREF=edit3.html>Edit myfile3.html</A>
</ul>
</html>
AM
FL
Y
This back-end system is composed of an entry page (password.html), a staff menu (menu.html),
and one or more editors (edit.html, edit2.html, and edit3.html). It's a complete system,
because it includes a password-protected entry page and a menu of staff options. The entry page is
accessed with a URL called a Start-link. Creating a Start-link is explained in Chapter 2, Logging into
Your Copy of HTML/OS.
TE
The password page used here is simply an HTML form that's linked to an on -lick Overlay with an
IF-THEN statement that tests whether the password entered by the user matches the word please
(or any password you hard-code there). Note that putting the word please in the page is not a
security problem. See the attached Security note in this section. The password page may be
substituted with more advanced entry pages, such as those discussed in Chapter 9, Building Login
Pages.
Security—Placing security information in Overlays, such as a password, is not a security hazard,
because Overlays are processed on the server and are not transferred to the browser. Of course,
HTML is transferred across, so you don't want to place any secure information in the HTML code
itself. Also, do not store HTML documents containing secure information on the public side of
HTML/OS. Documents that are public are marked with a red (warning) diamond in the HTML/OS
File Manager and can be moved to the private side. See Chapter 2, Logging into Your Copy of
HTML/OS, for a discussion of this topic.
The password page links to a staff menu page, which is simply an HTML document with hypertext
links that launch applications for staff use. In this case, three links were put on the page; each links
to a different copy of the six-line editor shown in Figure 8.1. Each of the editors should be modified
so it links to a different document.
In this back-end system, the six-line editor was duplicated three times. But what if you don't want to
maintain multiple copies of the editor? If you need to add features to the editor, you want to make
changes in only a single file, not multiple files. You can accomplish this by making the following two
changes: First, edit the staff menu page so the hypertext links point to the same six-line editor and
contain name-value pairs that set ted_file to the filename that needs to be edited when the user
clicks the link, as in the following example:
-81Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<A HREF=edit.html name=ted_file value="myfile6.html">Edit
myfile6.html</A>
Second, take out the first instruction of the six-line editor so it doesn't overwrite the value of
ted_file set in the link the user clicks. Now the backend system is composed of only three
documents, and more importantly, it uses only one editor. When users click a link on the staff menu
page, ted_file is set to the document name specified in the link and a single six-line editor runs; it
is ready to edit the file set in ted_file defined in the link.
Advanced Options
The series of pages described in the previous section is amazing- a complete back-end system in
only a dozen or so lines of code. We recommend you use your Web-based editor and type these
files into some pages and try them!
Of course, there's no sense on sitting on what you've learned. Completing your first text editor will
make you hungry for more features and more capabilities. A number of questions may pop into your
head, such as how to give users the ability to do the following:
ƒ Resize their text area
ƒ Select from files in a directory
ƒ Spell-check their work
ƒ Revert to prior versions of a document
ƒ Collaborate with others on a document
To allow such activities, you need advanced options. In the remainder of this chapter, you'll be
working through each, one at a time, adding code to the six-line editor for each appropriate option.
Each time you start with the original six-line editor at the beginning of this chapter and add the
necessary code. The descriptions are independent of each other, unless mentioned otherwise. Feel
free to read only those that interest you.
User-Controlled Text-Area Resizing
Perhaps the most important element of a text editor is the text area used to edit text. In the six-line
editor, the size of this text area is hard-coded in the program. This can cause problems since, as a
developer, you don't know the size or resolution of a user's screen. In fact, in some cases, neither
does the user. The users may use different computers to access the editor. In some cases, the
screen may be 800 by 600 pixels. At other times, it may be 1024 by 768. To make matters worse,
different browsers render text areas differently. Who is to say which browser the user is using? Or
how individuals configured their browsers. All of this makes it a good idea to give the users the ability
to adjust the size of their text area.
To accomplish that, you add two inputs to the screen that allow the users to enter the number of
columns and rows they want for the text area. You also add a Change button. When the user clicks
Change, you check the settings and save them to a file. When entering the page, load the settings
previously saved in the file. Use these settings when the text area is displayed. An editor with this
new option is shown in Figure 8.2. The code is shown follows the figure.
-82-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 8.2: The Six-Line Editor Now Has User-Controlled Text-Area Resizing.
<<
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
COPY FILE="configs.txt" TS="," to temp /COPY
x=temp[1] y=temp[2]
>>
<html>
<title>Six-line Text Editor (edit.html)</title>
<b>Six-line Text Editor</b><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=<<x>> rows=<<y>>>
</textarea>
</td></tr>
<tr><td>
Cols: <input type=text name=x size=4>
Rows: <input type=text name=y size=4>
<input type=submit name=ted_button value=Change>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
-83-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
COPY ted_text TO FILE=ted_file /COPY GOTO PAGE
ELIF ted_button="change" THEN
IF ISINTEGER(x)="FALSE" OR x < 5 OR x > 200 THEN x=60 /IF
IF ISINTEGER(y)="FALSE" OR y < 5 OR y > 100 THEN y=15 /IF
COPY ROW(x,y) TO FILE="configs.txt" TS="," /COPY
GOTO PAGE
ELSE
GOTO "menu.html"
/IF
>>
The code in bold is the code you add to the six-line editor to provide this resize option. Starting at the
top of the code, the instruction COPY FILE="configs.txt" TS="," TO temp /COPY reads the
file configs.txt into temp. The parameter TS="," indicates you are reading a comma-delimited
file. The file, as you will see, contains two cells (one row with two columns). The first cell contains the
number of columns for the text area. The second cell contains the height, in rows. The next line
loads these into x and y. Note that temp[1] is the same as temp[1,1], and temp[2] is the same
as temp[2,1], as explained in Chapter 6, Variables, Conditionals, and Loops.
You use the values x and y to dynamically set the number of columns and rows in the text area by
replacing the column and row values in the document with the Overlays <<x>> and <<y>>,
respectively.
Lower on the page, in bold, you see a new HTML table row. The HTML table row contains two input
boxes and Submit button that give users the ability to configure the text area. Note these HTML
elements are placed under the text area, away from the Save and Quit buttons, because smart
design dictates that you separate application buttons from those the user needs for configuration.
When a user clicks Change, the on-click Overlay, ted_edit runs. The IF-THEN statement runs the
instructions in bold. These instructions start with a test of the values in x and y. If they don't have
good values, they're set to 60 and 15, respectively. Then they are saved. Note how you use the ROW
tag to produce a two-column by one-row variable that is saved to a comma-delimited file with the
COPY tag. This is why, at the beginning of this page, it was assumed the file configs.txt was two
columns by one row, with the width of the text area in the first column and the height in the second
column.
Overall, the resize option you added to this page, required about five new instructions and a new
HTML table filled with the necessary HTML elements. The option gives the user the ability to
configure the text area as needed. Unfortunately, this solution has one minor flaw. In fact, the flaw is
so minor, most would not worry about it, but we shall.
The problem is we're saving the setting in a text file. Users using two different computers will need to
reset their text area if they need to go back and forth between two computers. This can happen if a
user, for example, edits the file from both home and work or from two different computers at work.
The solution is to save the data in a cookie rather than a file. This solution can be a bit tricky; but it's
worth it, because it attaches the setting to the computer the user is on. To accomplish this, you need
to replace the instructions that read and write to configs.txt with instructions hat read and write
to cookies. You use the tags COOKIEREAD and COOKIEWRITE. Your first attempt might produce the
following code:
-84-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<<
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
x=COOKIEREAD("ted_width")
y=COOKIEREAD("ted_height")
IF ISINTEGER(x)="FALSE" OR x < 5 OR x > 200 THEN x=60 /IF
IF ISINTEGER(y)="FALSE" OR y < 5 OR y > 100 THEN y=15 /IF
>>
<html>
<title>Six-line Text Editor (edit.html)</title>
<b>Six-line Text Editor</b><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=<<x>> rows=<<y>>>
</textarea>
</td></tr>
<tr><td>
Cols: <input type=text name=x size=4>
Rows: <input type=text name=y size=4>
<input type=submit name=ted_button value=Change>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY GOTO PAGE
ELIF ted_button="change" THEN
IF ISINTEGER(x)="FALSE" OR x < 5 OR x > 200 THEN x=60 /IF
IF ISINTEGER(y)="FALSE" OR y < 5 OR y > 100 THEN y=15 /IF
stat=COOKIEWRITE("ted_width",x,ADDAYS(today,100))
stat=COOKIEWRITE("ted_height",y, ADDAYS(today,100))
GOTO PAGE
ELSE
GOTO "menu.html"
-85-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/IF
>>
The code in bold contains the new instructions that read and write to cookies. Starting at the top, the
COPY tag, which read from configs.txt, was exchanged with two COOKIEREAD tags, which read
from cookies. The two IF-THEN statements you added check the values from the cookie. Some
people may not accept cookies, so it is reasonable to give proper values to x and y in the event they
have bad values. At the bottom of the page, the COPY tag that wrote to configs.txt was
exchanged with two COOKIEWRITE tags that write to cookies.
The COOKIEWRITE tags (see syntax of COOKIEREAD, COOKIEWRITE, and ADDAYS in Appendix D,
HTML/OS Tag Reference Guide) are set to expire 100 days in the future.
The problem however with this page is that it doesn't work—at least not completely. The tags are
correct; but when working with cookies, you can't save them and expect them to be readable in the
page displayed immediately after the write. You need to wait until the page after that. This is just how
cookies work. What this means is you don't want to read them after writing them. You can
accomplish this by setting a value in a variable when you write to cookies and checking that value
before reading them, so you can avoid reading them if you've just written them. The code to do this
is follows. The repairs to the page are shown in bold. Now you have a user-controlled, text-area
resize option that stores its data in cookies.
<<
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
IF cookie_justwrote != "TRUE" THEN
x=COOKIEREAD("ted_width")
y=COOKIEREAD("ted_height")
/IF
cookie_justwrote="FALSE"
IF ISINTEGER(x)="FALSE" OR x < 5 OR x > 200 THEN x=60 /IF
IF ISINTEGER(y)="FALSE" OR y < 5 OR y > 100 THEN y=15 /IF
>>
<html>
<title>Six-line Text Editor (edit.html)</title>
<b>Six-line Text Editor</b><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=<<x>> rows=<<y>>>
</textarea>
</td></tr>
-86-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<tr><td>
Cols: <input type=text name=x size=4>
Rows: <input type=text name=y size=4>
<input type=submit name=ted_button value=Change>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY
GOTO PAGE
ELIF ted_button="change" THEN
IF ISINTEGER(x)="FALSE" OR x < 5 OR x > 200 THEN x=60 /IF
IF ISINTEGER(y)="FALSE" OR y < 5 OR y > 100 THEN y=15 /IF
stat=COOKIEWRITE("ted_width",x,ADDDAYS(now,100))
stat=COOKIEWRITE("ted_height",y,ADDDAYS(now,100))
cookie_justwrote="TRUE"
GOTO PAGE
ELSE
GOTO "menu.html"
/IF
>>
File Selection
The six-line editor is fine if you know in advance the name of the file to edit. But what if you need to
add a way for your user to select a file? In the example code in this section, you give the user the
ability to select a file from a specific file directory. To accomplish this, you will want to add a page to
the editor that lists the files in that folder. Let's call this page select.html. When users first enter
the editor from menu.html, you want to link them to this page instead of the editor. The link in
menu.html might look as follows:
<A HREF=select.html>Edit Archived Documents</A>
In select.html, you want to list the files in some directory and generate links that look like the
following:
<A HREF=edit.html name=ted_file VALUE=[location1]>[location2]</A><br>
[location1] is where the full name of the file should go and [location2] is where the text for
the link should go. A link like this was created in the section titled "The Six-Line Text Editor" at the
beginning of this chapter. But this time, you need to create the link dynamically. To accomplish this,
you use the tag FILELIST. This tag fills ted_files with a list of files from a specified directory. It
places each file in a different row. See the Appendix D HTML/OS Tag Reference Guide for its syntax.
-87-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
You then loop across the rows of ted_files, displaying a link for each file you loop across. FOR
loops are discussed in Chapter 6, Variables, Conditionals, and Loops. The code for the Web page
follows:
<html>
<title>Document Selection (select.html)</title>
Select document to edit or <A HREF="menu.html">quit</A> to
menu.
<ul>
<<
user_folder="/archive"
ted_files=FILELIST(user_folder)
FOR NAME=ted_files ROWNAME=x DO
DISPLAY
"<A HREF=edit.html name=ted_file VALUE=" +
user_folder + "/" + x[1] + ">" + x[1] + "</A><br>"
/DISPLAY
/FOR
>>
</ul>
</html>
The DISPLAY tag inside the FOR loop displays links, one link per file. The code user_folder +
"/" + x[1] replaces [Location1] and x[1] replaces [Location2]. Now you have a page
that displays a list of files in a directory. Clicking a filename sets ted_file and launches
edit.html. Remember to delete the line ted_file.html="myfile.html" at the top of
edit.html so ted_file is not overwritten when a user clicks a link. An example select.html is
shown in Figure 8.3.
Figure 8.3: A File Selection Screen is a Useful Addition to the Six-Line Editor.
Adding Delete and Copy Buttons
-88-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Now that you know how to edit multiple files, it only makes sense to give the user the ability to delete
and copy files too. This can be done from the file selection page.
To do this, place radio buttons next to each file. Add Delete and Copy buttons to the top of the page.
Put a text box to the left of the Copy button, so the user has a place to put a destination filename.
When the user clicks the Delete button, you want to delete the selected file. When the user clicks the
Copy button, you want to copy it to the filename specified next to the Copy button. The page is
structured as follows:
<html>
<title>Document Selection (select.html)</title>
Select document to edit or <A HREF="menu.html">quit</A> to
menu.
<form method=post action=dostuff>
<font size=1>Destination</font><br>
<input type=text size=15 name=ted-dest>
<input type=submit name=button value=Copy>
<input type=submit name=button value=Delete><BR>
<<
user_folder="/archive"
ted_files=FILELIST(user_folder)
FOR NAME=ted_files ROWNAME=x DO
DISPLAY
"<input type=radio NAME=ted_selection VALUE=" + x[1] +
">" +
"<A HREF=edit.html NAME=ted_file VALUE=" +
user_folder + "/" + x[1] + ">" + x[1] + "</A><br>"
/DISPLAY
/FOR
>>
</form>
</html>
<<overlay dostuff
IF ted_selection="" THEN GOTO PAGE /IF
IF button="Copy" THEN
ted_dest=TRIM(ted_dest)
IF COUNT(ted_dest,"/")>0 OR COUNT(ted_dest," ")>0 OR
COUNT(ted_dest,"\")>0 THEN GOTO PAGE
/IF
COPY FILE="/archive/" + ted_selection TO FILE="/archive/"
+ ted_dest /COPY
ELIF button="Delete" THEN
temp=SYSRM("/archive/" + ted_selection)
-89-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/IF
GOTO PAGE
>>
When the user clicks Delete or Copy the on-click Overlay, dostuff runs. The first IF-THEN statement
redisplays the page if no file has been selected.
If the user clicks Copy, the value placed in the input box, preceding the Copy button, is trimmed
(duplicate spaces are deleted along with any leading or trailing spaces), and an IF-THEN statement
returns to the top of the page if the name contains bad characters. If the file is fine, the selected file
is copied. If the user clicks Delete, the selected file is deleted.
Adding Spell-Check
Although forms on the Web rarely include spell-check capability, if you can add such spell-check
capability, why not add it? It shows a high level of sophistication on your part as a Web developer
and it's a useful addition to any Web-based text editor.
To accomplish this, you need a spell-check kit. The kit described here is available in Aestiva's
freeware library. It's called Spell-Kit. The kit is about 7 megabytes and unpacks to about 50
megabytes.
If you're using the preinstalled copy of HTML/OS included with this book, you will have enough room
to install it. But don't install more than one copy. There's no quicker way to run out of hosting space
than to install multiple copies of this application.
Download Spell-Kit to your computer, and then, using the File Manager, upload it to your copy of
HTML/OS. Using the Install option in the Control Panel, install your copy. Install it in its default folder,
which is /apps/spellcheck. Once it's installed, you can add spell-check to any HTML form by
making a few modifications to the page. Adding spell-check capability to a page takes two steps.
First you need to add a Spell-check button to the HTML form. Second you need to add instructions in
the on-click Overlay the form uses to call a Spell-Kit page. These instructions set up a number of
parameters required by Spell-Kit and launch it. The parameters are explained in the online help
included with Spell-Kit. The six-page text editor, properly set up with a Spell-check button, is shown
in the following code. Changes appear in bold:
<<OF ted_file!="ERROR" THEN
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
/IF
>>
<html>
<b>Six-line Text Editor</b><br>
File Name: <<ted_file>>
<form method=post ACTION=ted_edit>
<table border=0>
-90-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<tr><td align=right>
<input type=submit name=ted_button value="Spell Check">
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=65 rows=15>
</textarea>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY GOTO PAGE
ELIF ted_button="Spell Check" THEN
spell.return=page
spell.textarea.var="ted_button"
AM
FL
Y
spell.form="/apps/spellcheck/template.html"
GOTO spell.engine
ELSE
GOTO "menu.html"
/IF
>>
TE
spell.engine="/apps/spellcheck/index.html"
When the user clicks the Spell Check button, the document /apps/spellcheck/spell.html
runs and displays a spell-check page, giving the user the ability to check and correct spelling
mistakes. A spell-check page is shown in Figure 8.4.
-91Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 8.4: A Spell-Check Screen is Another Useful Addition to the Six-Line Editor.
Version Control
The six-line editor does not automatically save a document's past versions. Of course, most editors
don't either. But in certain situations, the ability to archive past versions of a document automatically
is useful. Suppose you have a back-end system like the one described earlier in this chapter. The
system had a menu page with links that allowed the user to edit different documents. If one of those
documents were the home page of the Web site, version control on that page could come in handy.
To add version control to a document, you need to create a place for the archived files. You should
also save files to the archive only in special situations, not with every change. Otherwise you'll end
up with too many files. One way to do this is to add a Save To Archive button, so users can manually
save documents to the archive. You will also need a Load From Archive button, so the user can
revert to prior versions. See the following code:
<<
ted_file="myfile.html"
COPY FILE=ted_file TO ted_text /COPY
>>
<html>
<b>Six-line Text Editor with Version Control</b><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=65 rows=15>
</textarea>
</td></tr>
-92-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<tr><td>
<input type=submit name=ted_button value="Save To Archive">
<input type=submit name=ted_button value="Load From Archive">
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY GOTO PAGE
ELIF ted_button="Save To Archive" THEN
[location1]
ELIF ted_button="Load From Archive" THEN
[location2]
ELSE
GOTO "menu.html"
/IF
>>
When the user clicks Save To Archive, the instructions at [location1] need to copy the file to a
directory. The file would be given a unique name similar to the original. For example, the archive of
myfile.html might be /archive/myfile~12345.html. The number 12345 represents the
number of minutes since the beginning of the millennium. When the user clicks Load To Archive, the
instructions at [location2] would go to a page that displays a list of archived files, so the user can
select one and load it into the editor.
Save to Archive
Saving to the archive requires that you first calculate a new filename and copy the file to the archive.
To accomplish this, you use TIMEFROM to calculate the number of minutes since the last millennium,
REPLACE to do a search and replace when calculating the new archive name, and COPY to copy the
document to the archive. The instructions to do this are as follows:
num=TIMEFROM("01/01/2000","minutes")
ted_file_archive = "/archive/" + REPLACE(ted_file, ".html",
"~"+num+".html")
COPY ted_text TO FILE=ted_file /COPY
COPY ted_text TO FILE=ted_file_archive /COPY
Load From Archive
At [location2], you want to go to archive.html. So you need to replace [location2] with
the following code:
-93-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
GOTO "archive.html"
The page archive.html is similar to the page select.html discussed in the section titled "File
Selection" in this chapter. Here too, you want to display a list of files. But in this case, you want to list
only those files associated with the current file (since /archive might be used to store other archived
files). A revised select.html, one that shows only filenames beginning with the file name (without
the .html) is shown in the following code. Modifications to the original version of select.html
appear in bold.
<html>
<title>Document Selection (archive.html)</title>
Select document to load or <A HREF="edit.html">quit</A> to
editor.
<ul>
<<
user_folder="/archive"
ted_files=FILELIST(user_folder)
ted_files=GETCOLBEGIN(ted_files,1,ted_file+"~")
FOR NAME=ted_files ROWNAME=x DO
DISPLAY
"<A HREF=edit.html name=ted_file VALUE=" +
user_folder + "/" + x[1] + ">" + x[1] + "</A> (" + x[3]
+ ")<br>"
/DISPLAY
/FOR
>>
</ul>
</html>
A user who clicks the Load From Archive button is directed to this page. The page uses FILELIST,
scans the directory, filling ted_files with a list of files. It puts the filenames it finds in column 1 and
the modification dates of the files in column 3. The GETCOLBEGIN tag takes ted_files and returns
only those rows where the name in column 1 begins with the filename and a tilde. That restricts
ted_files to only those of interest. This is followed by a FOR loop that loops across the files in
ted_files. It displays a hypertext link for each file. It follows each with the file creation date in
parenthesis.
Explanations of the tags FILELIST, and GETCOLBEGIN can be found in Appendix D, HTML/OS Tag
Reference Guide. For an explanation of FOR loops, refer to Chapter 6, Variables, Conditionals, and
Loops.
Document Collaboration
Many staff environments require collaboration between multiple people on a single document. Often
the collaboration takes the form of a workflow system. For example, a writer might be responsible for
editing an HTML document for an online publication. The document is then passed to an editor for
-94-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
review, who then passes it to a Webmaster, who puts it online. This kind of collaboration is an
example of a three-person workflow system.
Other forms of collaboration include systems in which multiple parties have simultaneous access to
documents; but only one person can work on a document at a time. This is not really an example of
workflow. Rather it's an example of a file editor with file locking. One algorithm you can use to ensure
that a file is editable by only one staff member at a time is as follows: Whenever a user wishes to edit
a document, that user must check a status file. The status file should contain the user ID of the last
person who edited the document and the time of that edit. If the time in the file is too old or the status
file has nothing in it, which occurs when the user releases the document, the party requesting the file
can edit it. If not, the editor tells the user the document is not available. This kind of document-editing
system is not discussed here but is left as an exercise.
Here we discuss the construction of a workflow system. Workflow systems can be built with
databases or with files. Here we present a file-based version of workflow system. If you want to build
a database-based workflow system, keep in mind that the concepts discussed here can be
translated to the database-based situation. In that case, documents are stored as fields in a
database, rather than as separate files, and the document's work area, is actually another field in the
database, rather than a directory. Of course, this will make more sense after you learn how to build a
file-based workflow system.
Let's consider a two-person workflow system consisting of a writer and a Webmaster. The technique
is the same if you have three or more people. The system works like this. Each person in the
workflow system has a separate work area —a private directory. A document, at any point in time, is
placed in one of those directories, indicating the person to whom the document belongs. In this
example, the documents available to our writer are stored in /work1. The documents available to
our Webmaster are stored in /work2. Completed documents are stored in the /work3. We call
/work3 our archive. To create a workflow system, you need to add file selection to the six-line editor
and add buttons in the six-line editor that transfer files between the different directories. The writer
will need the Send to Webmaster button. The Webmaster will need buttons Return to Writer, Send to
Archive, and Load From Archive.
The text editor needs to offer different options to the writer and the Webmaster. In addition, each
person using the workflow system needs to see only his or her own files. In the back-end system
discussed in the beginning of this chapter, a single password was used to access the system.
Typically more advanced login pages are used. Since this is not the topic of this chapter, we'll avoid
this issue by simply changing the password page, so it can accept two passwords, one for the writer
and one for the Webmaster. The workflow system here is comprised of a password page that links
directly to a file selection page. (The menu.html page has been omitted.) The password page is as
follows:
<html>
<title>Password Page (password.html)</title>
Staff Access - For Authorized Personnel Only
<form method=post ACTION=be_check>
Enter Password: <input type=password name=be_password size=15>
<input type=submit value="Enter">
</form>
</html>
<<overlay be_check
-95-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF be_password = "hemmingway" THEN
wf_wa ="/work1" wf_user="Writer" GOTO "select.html"
ELIF be_password = "linuxdude" THEN
wf_wa ="/work2" wf_user="Webmaster" GOTO "select.html"
/IF
GOTO PAGE
>>
When a user logs in, the variable wf_wa (the WorkFlow Work Area) is set to /work1 or /work2,
depending on whether the user is a writer or a Webmaster; wf_user is set to the type of user, and
the page select.html launches.
The page select.html is almost identical to the select.html used in the section, "File
Selection," earlier this chapter. However, this time, files selected must originate in the /work1 or
/work2 directories, depending on who accesses the page. The new select.html is as follows:
<html>
<title>Document Selection (select.html)</title>
Select document to edit.
<ul>
User: <<wf_user>><br>
File Area: <<wf_wa>><br><br>
<<
user_folder=wf_wa
ted_files=FILELIST(user_folder)
FOR FILE=ted_files ROWNAME=x DO
DISPLAY
"<A HREF=edit.html NAME=ted_file VALUE=" +
user_folder + "/" + x[1] + ">" + x[1] + "</A><BR>"
/DISPLAY
/FOR
>>
</ul>
</html>
The next thing to do is change edit.html so it displays the workflow buttons we mentioned
previously, and program them to move files between the /work1, /work2, and /work3 work
areas. You want to use a single edit.html, which eliminates the need to maintain multiple versions
of the editor. You need to ensure that different buttons display depending on the value of wf_wa.
The file is structured as follows:
<<
-96-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
COPY FILE=ted_file TO ted_text /COPY
>>
<html>
<b>Workflow Editor</b><br>
User: <<wf_user>><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<<[Location1]>>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=65 rows=15>
</textarea>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY GOTO PAGE
ELIF button="Send To Webmaster" THEN
[Location2]
ELIF button="Return To Writer" THEN
[Location3]
ELIF button="Send To Archive" THEN
[Location4]
ELIF button="Load From Archive" THEN
[Location5]
ELSE
GOTO "menu.html"
/IF
>>
[Location1] is where you want to display the workflow buttons, which are different for the writer
and the Webmaster. This requires an IF-THEN statement. You can write the following code:
IF wf_wa="/work1" THEN
DISPLAY
-97-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
'<input type=submit name=ted_button value="Send To
Webmaster">'
/DISPLAY
ELIF wf_wa="/work2" THEN
DISPLAY
'<input type=submit name=ted_button value="Return to
Writer">'+
'<input type=submit name=ted_button value="Send To
Archive">'+
'<input type=submit name=ted_button value="Load From
Archive">'+
/DISPLAY
/IF
The code at locations [Location2], [Location3], and [Location4] need to move files
between /work1, /work2, and /work3. At [Location2], for example, you can use the following
instructions:
new_file=REPLACE(ted_file,"/work1","/work2")
temp=SYSMV(ted_file, new_file)
You use the REPLACE tag to perform a search and replace on the work areas, so a destination name
can be created. Then SYSMV moves the file to the new destination. See Appendix D, The HTML/OS
Reference Guide, for a description of SYSMV. Use same technique at [Location2] and
[Location3], except that the work areas are different.
[Location5] runs when the user clicks the Load From Archive button. Here you want to launch
select.html so it lists the files in /work3. To accomplish this, before launching select.html,
you set a variable loadfromarchive to TRUE. Then in select.html, you set the user_folder
to /work3, in the event loadfromarchive is TRUE. Then after you return to edit.html, you set
loadfromarchive to FALSE. These are the changes needed to complete this two-person workflow
system. The revised six-line editor, edit.html, is as follows:
<html>
<< loadfromarchive="FALSE">>
<b>Workflow Editor</b><br>
User: <<wf_user>><br>
File Name: <<ted_file>>
<form method=post action=ted_edit>
<table border=0>
<tr><td align=right>
<<
IF wf_wa="/work1" THEN
DISPLAY
-98-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
'<input type=submit name=ted_button value="Send To
Webmaster">'
/DISPLAY
ELIF wf_wa="/work2" THEN
DISPLAY
'<input type=submit name=ted_button value="Return to
Writer">'+
'<input type=submit name=ted_button value="Send To
Archive">'+
'<input type=submit name=ted_button value="Load From
Archive">'
/DISPLAY
/IF
>>
<input type=submit name=ted_button value=Save>
<input type=submit name=ted_button value=Quit></td></tr>
<tr><td>
<textarea name=ted_text cols=65 rows=15>
</textarea>
</td></tr>
</table>
</form>
</html>
<<overlay ted_edit
IF ted_button="save" THEN
COPY ted_text TO FILE=ted_file /COPY
ELIF button="Send To Webmaster" THEN
new_file=REPLACE(ted_file,"/work1","/work2")
temp=SYSMV(ted_file, new_file) GOTO PAGE
ELIF button="Return To Writer" THEN
new_file=REPLACE(ted_file,"/work2","/work1")
temp=SYSMV(ted_file, new_file) GOTO PAGE
ELIF button="Send To Archive" THEN
new_file=REPLACE(ted_file,"/work2","/work3")
temp=SYSMV(ted_file, new_file) GOTO PAGE
ELIF button="Load From Archive" THEN
loadfromarchive="TRUE"
GOTO "select.html"
ELSE
GOTO "menu.html"
-99-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/IF
>>
The new version of select.html is as follows:
<html>
<title>Document Selection (select.html)</title>
Select document to edit or <A HREF="menu.html">quit</A> to
menu.
<ul>
<<
IF loadfromarchive="TRUE" THEN
user_folder="/work3"
ELSE
user_folder=wf_wa
/IF
ted_files=FILELIST(user_folder)
FOR FILE=ted_files ROWNAME=x DO
DISPLAY
"<A HREF=edit.html NAME=ted_file VALUE=" +
user_folder + "/" + x[1] + ">" + x[1] + "</A><BR>"
/DISPLAY
/FOR
>>
Summary
In this chapter you learned how to tailor Web-based text editors to fit many needs. Unlike the legacy
world, where text editors serve a general word-processing function, Web-based text editors can be
customized and integrated into specific business and organizational systems.
Despite the sophistication of the systems built in this chapter, the construction of these text editors
required mostly COPY, DISPLAY, and GOTO tags, IF-THEN statements, and the occasional FOR
loop. Other HTML/OS tags were called upon in special situations. Specifically, COUNT, FILELIST,
REPLACE, TIMEFROM, TRIM, GETCOLBEGIN, SYSMV, and SYSRM were used. Overall, a total of
twelve different HTML/OS tags were used to build the text-editors used in this chapter.
If you found some of the tags used in this chapter difficult to understand, we recommend you refer to
Appendix D, The HTML/OS Reference Guide. Log into your copy of HTML/OS and experiment with
the tags by building simple pages that use them.
In the next chapter, you'll learn about building login pages. Login pages are used in membership
systems, in back-end systems, such as those in this chapter, and in situations where access to a
Web-application needs to be controlled.
-100-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercises
Even most Web developers can't build Web-based editors. But now you can! Do the following
exercises to refine your skills, so this kind of development becomes second nature to you. Answers
to all exercises are provided on this book's companion Web site as described in the book's Preface.
Exercise 1
Add a Reload button to the six-line editor. Hint: Reloading a page is a matter of redisplaying the page
when a user clicks a Reload button.
Exercise 2
Add a Save/View button to the six-line editor. Hint: Viewing a page is a matter of going to the page.
Use the GOTO tag to do this.
Exercise 3
An editor with a file selection provides the ability to select files in a directory. The file select.html
in this chapter does not issue any special message when the directory is empty. Change
select.html so that if it finds no files, HTML/OS displays the message "No Files Found" rather
than an empty list of files. Hint: When FILELIST finds no files, it returns an empty string.
Exercise 4
AM
FL
Y
This chapter includes code for resizing a text area and saving the values in a cookie. Modify this
example so it works when users don't have cookies. To accomplish this, modify the page so text
area heights and widths are saved and loaded from a file, in the event cookies are not readable.
TE
Exercise 5
The select.html pages in this chapter display a single column of file names. Using HTML tables,
modify select.html so it displays filenames, file sizes, and modification dates in different
columns—not unlike the File Manager used in HTML/OS.
-101Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 9: Building Login Pages
Overview
Whether you're interested in building a staff password page that's hidden from public view, a
membership login with a Lost Your Password option, or a login page that separates wholesalers from
retail customers, you'll want to read this chapter.
Off the Web, there is little flexibility in the way login pages are designed. You enter your login and
password and that's it. The entire page is dedicated to logging you in. On the Web login boxes are
merged with other information on the page. The page may include options to e-mail users'
passwords, in case they forgot them. This is not done off the Web.
Once again, as you have seen throughout this book, Web-based computing offers greater design
flexibility than legacy-style computing. In this chapter, you learn how to program login pages and how
to modify them to serve a wide variety of needs.
We start out by discussing the overall security mechanism used by HTML/OS. Then we introduce a
ten-line login page that reads its password information from a text file. Later you modify it to read its
data from a database. Then we provide code examples and explanations about how to modify the
page to perform custom tasks.
Web-Access Security
In HTML/OS you provide access to Web pages in only two ways. Either the user has a special URL
to that page that you enabled in the HTML/OS Control Panel, or the user clicks a link to that page
from an existing HTML/OS Web page. Building login pages is not about creating security
mechanisms. It's about controlling the movement of users into the site once they're on the login page.
Most often you set up a Start-link to allow users to access to the login page itself. Setting up Startlinks is described in Chapter 2, Logging into Your Copy of HTML/OS.
Login Page Functionality
All login pages share specific commonalities. Once on a login page, the user typically reaches an
HTML form containing input boxes for a login and password. When the user clicks the Submit button
on the HTML form, an on- click Overlay checks the user's access information against hard-coded
values such as data in a password file or a user database. If it matches that data, a GOTO tag gives
the user entry to a page on the site. From there, the user has access to any page linked to that page
and to any pages that page links to.
Once the user has logged in, you must set certain variables in the login page depending on what you
want to accomplish when the user logs in. For example, you may want to set a user ID, load
personal settings, define a user level, or set a home directory. These settings will be available for the
remainder of the user session, because HTML/OS variables carry across Web pages automatically.
You can use these settings to write or not write links to certain sections of the site, place the user in
a specific directory tree, or limit a user's access to data in a database.
The login pages discussed in this chapter read their user data from either comma-delimited text files
or user databases. As a rule, it's better to place user logins and passwords in a database table,
rather than a text file. But when you have few users, it is reasonable to store the logins and
passwords in a comma-delimited text file.
-102-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The Ten-Line Login Page
In the previous chapter, you saw a two-line password page that allowed access to a back-end
system. Of course, that two-line password page had its limitations. For one, it requested only a
password. It didn't request a login. While a password is often enough, in this chapter, login pages
require a login and a password. Second, the two-line password page did not have any error reporting.
If the user entered a bad login, the page was redisplayed without an error message. This ten-line
login page includes error reporting and requests both logins and passwords. The login page is
shown in Figure 9.1. The code is shown here:
<html>
<title>Login Page (login.html)</title>
<<
IF lp_msg != "ERROR" THEN
DISPLAY "<font color=red>" + lp_msg + "<font>" /DISPLAY
lp_msg = "ERROR"
/IF
>>
<form method=post ACTION=lp_check>
Login ID: <input type=text size=30 name=lp_login><br>
Password: <input type=password size=30 name=lp_pass >
<input type=submit value=Login>
</form>
<< lp_login="" lp_pass ="">>
</html>
<<OVERLAY lp_check
IF lp_login ="" OR lp_pass ="" THEN lp_msg = "Bad Login."
GOTO PAGE /IF
COPY FILE="logins.txt" TS="," TO lp_table /COPY
temp=GETCOLEQ(lp_table,1,lp_login)
IF temp[2,1] != lp_pass THEN
lp_msg="Bad Login" GOTO PAGE
ELSE
GOTO "staff.html"
/IF
>>
-103-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 9.1: The Ten-Line Login Page Uses Ten HTML/OS Instructions and an HTML Form.
You use the first Overlay in the page to display an error message, if an error exists. The IF-THEN
statement checks to see whether lp_msg is ERROR. If it isn't, it displays lp_msg and sets it to
ERROR. Remember, in HTML/OS, the default value of variables is ERROR, so the IF-THEN statement
displays lp_msg only after it receives a value.
To log in, users enter a login and password and click Login, which triggers the on-click Overlay
lp_check. The first line of the on-click Overlay determines whether the user forgot to enter a Login
or a password. If so, it sets an error message and redisplays the page. Otherwise, the logins and
passwords, which are stored in a comma-delimited text file, are loaded into lp_table. The login file
has logins in column 1 and passwords in column 2. The first three lines of the file might look as
follows:
"John","please"
"Johnny","njk100k"
"Janet","qqyw123"
When the file loads into lp_table, it becomes a variable with two columns. The TS="," parameter
of the COPY tag tells HTML/OS the file is comma-delimited. The variable lp_table that's created is
a 2-column by 3-row table that looks as follows:
John
please
Johnny
njk100k
Janet
qqyw123
To determine whether a row in lp_table has the proper login ID and password, use the tag
GETCOLEQ. GETCOLEQ is part of a family of Overlay tags designed to pull rows out of a variable
(containing multiple columns and rows) that match specific criteria. These tags return a new table
with only those rows that match. In the case of GETCOLEQ, the criterion is that a specified column
equals a specified value. For further information, see Appendix D, HTML/OS Tag Reference Guide.
The instruction, temp=GETCOLEQ(lp_table,1,lp_login), creates temp. If lp_login is not in
any of the rows of column 1, temp is an empty string (a one-column by one-row table with nothing in
it). If it finds a match, temp becomes a two-column by one-row variable containing the login in the
first column and the password in the second column.
The next instruction tests to see whether the password the user entered matches the second column
of temp. If it doesn't, HTML/OS prepares and displays an error message—denying the user access
to staff.html.
-104-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Using a Database
Our preceding login page assumes logins and passwords are stored in a text file. This is not
convenient if you have many logins or if you're building a membership system in which users are
automatically added and deleted. In these circumstances, it's much better to place login information
in a user database.
To do this, you need to set up a user database first. Setting up a database is discussed in Chapter 4,
Your First Web Database Program. We shall assume you have set up a database called
/work/users with the fields, user_login, and user_password (among others.)
Start with the ten-line login page and delete the instructions in the on-click Overlay that read from a
comma-delimited file and fill temp with rows for lp_login. Replace those instructions with a
database search. The ten-line login page that reads from a database looks as follows:
<html>
<title>Login Page (login.html)</title>
<<
IF lp_msg != "ERROR" THEN
DISPLAY "<font color=red>" + lp_msg + "</font>"/DISPLAY
lp_msg = "ERROR"
/IF
>>
<form method=post ACTION=lp_check>
Login ID: <input type=text size=30 name=lp_login><br>
Password: <input type=password size=30 name=lp_pass >
<input type=submit value="Login">
</form>
<< lp_login="" lp_pass ="">>
</html>
<<OVERLAY lp_check
sstr='user_login="'+lp_login+'" AND
user_password="'+lp_pass+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_login,user_password")
IF temp[2,1] != lp_pass THEN
lp_msg="Bad Login" GOTO PAGE
ELSE
GOTO "staff.html"
/IF
>>
-105-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The modified lines appear in bold. The first bold line sets up a Boolean search. This topic is
discussed further Chapter 12, Building Query Pages. The next instruction does the search. It uses
DBFIND, the same Overlay tag you used in Chapter 4, Your First Web Database Program. When a
user enters a login and password and clicks Login, a variable containing a search query is set up.
Then DBFIND searches the database. If no matches are found, temp is filled with the empty string. If
a record is found, temp becomes a two-column by one-row variable containing the login in the first
column and the password in the second column. Notice the similarity to the original login page. After
that an IF-THEN statement tests to see whether the second column of temp contains the password
the user entered. If it does, the user is logged in. If not, the IF-THEN statement sets an error
message and redisplays the page.
The Automatic Login
Sometimes it's a good idea to allow certain users to log in automatically by setting a cookie on a
user's browser. The cookie should contain a login and password and be set by a special page, at the
user's request. Automatically logging in a user is not a good idea when secure information and/or
credit card information is available, but it's useful for less critical situations.
One way to log in a user automatically is to add an Underlay to the login page that determines
whether a cookie has been set. If it has been set, you test the login and password. If they are set
correctly, you send the user directly to staff.html. The Underlay to do this is as follows:
<< lp_login=COOKIEREAD("auto_login")
lp_pass=COOKIEREAD("auto_pass")
IF lp_login != "ERROR" THEN
sstr='user_login="'+lp_login+'" AND
user_password="'+lp_pass+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_login,user_password")
IF temp[2,1] = lp_pass THEN
GOTO "staff.html"
/IF
/IF
>>
To test using automatic login, you will want to set up a Web page that writes cookies. You use
COOKIEWRITE to do that. See Appendix D, HTML/OS Tag Reference Guide for help using the
cookie tags. Here's a page that writes cookies. You can use this page to test the automatic login
Underlay we just discussed:
<html>
<title>Test Cookie Write</title>
<< mylogin="user001"
mypass="12345"
temp=COOKIEWRITE("auto_login",mylogin, ADDDAYS (NOW, 10)
temp=COOKIEWRITE("auto_pass",mypass, ADDDAYS (NOW, 10)
-106-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
>>
Login set to <<mylogin>>.<br>
Password set to <<mypass>>.<br>
Cookie written. Make sure login-password pair is in user
database.
</html>
Loading User-Specific Information
When a user logs into a site, you often want to set certain information about that user or perform
certain functions. Depending on the login system you are building, any of the following questions
might need to be answered with some special programming:
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
What is the user's ID?
Is the user a wholesaler or a retail customer?
What is the user's access level?
How many times has the user logged in?
Is this user staff?
Does the user have an outstanding balance due?
Are messages waiting for this user?
How should this user's Web pages be displayed?
The most common way to answer these questions is to add programming instructions to the login
page preceding the GOTO "staff.html" line. Additional user information loads from the user
database when you check the login and password.
For example, let's say you have a Web site that serves both retail and wholesale customers. You
have a field in the users database called user_level, which can have the values W or R. When a user
logs in, you could add this to the information you collect from the database. The on-click Overlay on
the login page might look as follows:
<<OVERLAY lp_check
sstr='user_login="'+lp_login+'" AND
user_password="'+lp_pass+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_login,user_password,use
r_level")
IF temp[2,1] != lp_pass THEN
lp_msg="Bad Login" GOTO PAGE
ELSE
user_level=temp[3,1]
GOTO "staff.html"
/IF
>>
-107-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Changes appear in bold. As you see, the field list in DBFIND contains an additional field. The
additional field is extracted from the database /work/users when the query runs. In this example,
it's placed in column 3 of temp. Just before logging in, the variable user_level is set to this value.
As another example, suppose the user database has a field containing an expiration date. Suppose
you want to let the user log in to the staff area only if the user's account has not expired. Here you
would want to read this field and see if it is in the past. The on-click Overlay on the login page might
look as follows:
<<OVERLAY lp_check
sstr='user_login="'+lp_login+'" AND
user_password="'+lp_pass+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_login,user_password,use
r_lastday")
IF temp[2,1] != lp_pass THEN
lp_msg="Bad Login" GOTO PAGE
ELIF ISPAST(temp[3,1])="TRUE" OR ISPAST(temp[3,1])="ERROR"
THEN
lp_msg="Access disabled. Account has expired."
GOTO PAGE
ELSE
GOTO "staff.html"
/IF >>
Here the field user_lastday is loaded from the database (in addition to the valid user login and
password). The IF-THEN statement contains an additional test that determines whether
user_lastday is in the past or invalid. You use the ISPAST tag, which returns TRUE for dates in
the past. It returns FALSE for dates not in the past. It returns ERROR for invalid dates. In this IFTHEN statement, if temp[3,1] is not a valid date or it's in the past, an error message is set and the
page redisplays, thus denying the user access to staff.html.
Lost Your Password?
Ever lose you password? What a bummer! If the Web site or system you are accessing has a way
for you to recover your password, you're in luck. There's nothing more embarrassing than having to
call the office and bother someone to recover your password.
The Lost Your Password option is a must in login systems with thousands of users. Most often, you
make the link small and unobtrusive but place it in close proximity to the login and password input
boxes.
Creating a Lost Your Password option requires the following:
ƒ Users have unique e-mail addresses assigned to them
ƒ The e-mail addresses are already on file in a users database
-108-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ You feel comfortable e-mailing the login and password to the address on file
If these conditions are satisfied, you can create a Lost Your Password option. To do this, you build a
page that e-mails users a message with their login and password. A two-page set is shown here;
users access the first page of the two-page set from a link next to the login form. It might look as
follows:
<font size=1 face=arial,helvetica>
<A HREF="forgot.html">Forgot your password?</a>
</font>
On forgot.html, you place an HTML form that requests an e-mail address. When a user clicks the
button on the form to request login and password information, you check the database for that user's
login information. If you find it, you compose the e-mail message, send it out, and go to
forgot_thanks.html to instruct the user to pick up his or her e-mail information. If you can't find
the e-mail address, you issue an error message.
The page forgot.html looks as follows:
<html>
<title>Forgot Your Password </title>
<b>Forgot Your Password?</b>
Enter email address you have on-file with us
and we'll email you your login and password.
<<
IF lp_msg != "ERROR" THEN
DISPLAY "<font color=red>" + lp_msg + "<font>" /DISPLAY
lp_msg = "ERROR"
/IF
>>
<form method=post action=lp_emailcheck>
Email Address: <input type=text size=30 name=lp_email>
<input type=submit value="E-mail Me Now">
</form>
</html>
<<OVERLAY lp_emailcheck
sstr='user_email="'+lp_email+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_email,user_login,user_p
assword")
IF temp[1,1] != lp_email THEN
lp_msg="E-mail Address Not Found. Try again. " GOTO
PAGE
ELSE
-109-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
user_login = temp[2, 1]user_password = temp [3,1]
msg='Dear Sir/Madam,' + LF +
' Here is your password. Please keep on file.' +
' Login ID: ' + user_login + LF +
' Password: ' + user_password + LF +
'Thank you for using our Lost Your Password
feature.' + LF +
'- Staff' + LF
MAIL msg TO ADDRESS=lp_email
SUBJECT='Recovered Login and Password'
/MAIL
GOTO "forgot_thanks.html"
/IF
>>
The page forgot_thanks.html looks as follows:
<html>
<title>Forgot Your Password - Thanks</title>
<b>Your login and password has been emailed to you.</b><br><br>
Check your e-mail box and return to the login page to try
again.<br><br>
Happy to be of service.<br>
-Staff
</html>
The page forgot.html uses an HTML form that requests the e-mail address. When a user clicks
E-mail Me Now, the on-click Overlay lp_emailcheck runs. lp_emailcheck places a message in
msg. The MAIL tag sends the message. Then forgot_thanks.html is displayed.
Note how msg is composed. An LF appears at the end of each line. LF stands for linefeed. Unlike
Web pages, each line of an e-mail message should end with this special character. The line break in
Web pages is <br>. The line break in e-mail messages is LF.
Those Darn Line Breaks—Line breaks are tricky little characters. First of all, you can't see them.
Second, the characters you use as line breaks are different on Unix, Windows, and Macintosh
systems. Here are the kinds of characters each system uses:
ƒ
ƒ
ƒ
The line break for Macintosh files is CR (the carriage-return character).
The line break for Unix files is LF (the linefeed character).
The line break for Windows files is CR+LF (carriage-return followed by a linefeed).
Here are some guidelines that may help you:
ƒ
ƒ
Web pages ignore the differences in line breaks. They're translated into spaces.
Don't worry about them when working with Web pages.
When composing e-mail, terminate each line with LF.
-110-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
When working with text files, sometimes (not often, but often enough) it is best to
know the kind of line breaks used. In these cases, before working with the file,
replace CR+LFs to LFs and CRs to LFs. Use REPLACEALL. The tag does a search
and replace, for example,
mytxt=REPLACEALL(REPLACEALL(mytxt,CR+LF,LF),CR,LF).
Using an E-Mail Template
It is often a good idea to make automated e-mails customizable. The way to do that is to store the email message you'll be sending out in a separate file (the template) and make the file editable with a
text-editor similar to those discussed in Chapter 8. Building Text Editors.
To change this code so it uses a template, place the e-mail message in forgot_password.txt.
Replace any variables in the message with placeholders (unique sequences of characters you can
replace later with the actual variable values). In the following example, we use the placeholders
[login] and [password]. The template file for the message is as follows:
Dear Sir/Madam,
Here is your password. Please keep on file.
Login ID: [login]
Password: [password]
AM
FL
Y
Thank you for using our Lost Your Password feature.
- Staff
You place the placeholders, [login] and [password], where the actual variables should go.
In the programming, after the file is read into a variable, the actual variables replace the placeholders.
You use REPLACEALL to do this. The on-click Overlay for the template page shown previously,
modified to read its message from a file and perform the necessary search-and-replace procedures,
is as follows:
TE
<<OVERLAY lp_emailcheck
sstr='user_email="'+lp_email+'"'
temp=DBFIND("/work/users",sstr,1,1,"user_email,user_login,user_p
assword")
IF temp[1,1] != lp_email THEN
lp_msg="E-mail Address Not Found. Try again. " GOTO
PAGE
ELSE
user_login = temp[2,1] user_password = temp[3,1]
COPY FILE="forgot_password.txt" TO msg /COPY
msg=REPLACEALL(msg,"[login]",user_login)
msg=REPLACEALL(msg,"[password]",user_password)
MAIL msg TO ADDRESS=lp_email
SUBJECT='Recovered Login and Password'
/MAIL
GOTO "forgot_thanks.html"
/IF
-111Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
>>
Changes appear in bold. The COPY tag copies forgot_password.txt into msg. Two search-andreplace actions are performed. Each time REPLACEALL looks for a placeholder and replaces it with
the contents of a variable. The rest of the on-click Overlay is the same as before.
Summary
In this chapter, we've reviewed how to set up login pages. You discovered that you can tailor them in
an infinite number of ways to serve just about any need; but the construction process is the same.
Give the user access to a page using a Start-link or a link from another HTML/OS page. Then, on
that page, add an HTML form. When the user tries to log in, check a file or database. If login is
successful, set some variables, perform some tasks, and let the user into the private side of your site.
Exercises
In the following exercises, you modify the ten-line login page. In each case, start with the version of
the ten-line login page that reads from a database found in the "Using a Database" section earlier in
this chapter. Answers to all exercises are provided on this book's companion Web site as described
in the book's Preface.
Exercise 1
Modify the ten-line login page so it sets a wholesale or retail flag when the user logs in. If the user is
a wholesaler, send the user to wholesale.html. If not, send the user to retail.html. Create a
Start-link for your login page.
Exercise 2
Modify the ten-line login page so it checks the users' account balances before letting them in. Hint:
To do this, use a users database with a user_balance field. To test your page, have the login page
deduct some amount from users' balances every time they log in.
-112-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 10: HTML Forms Processing
Overview
It's difficult to imagine an advanced Web site without HTML forms. The most sophisticated HTML
forms work in concert with other HTML forms placed on additional Web pages. You see them used in
multipage questionnaires, product checkout pages, and sophisticated user input forms like the Webbased tax filing service shown in Figure 10.1.
Figure 10.1: This U.S. Tax Filing Service is an Application of HTML Forms Processing in an HTML/OSDriven Web Site.
More often HTML forms are confined to a single Web page. They are used in everyday Web
components such as login pages, guest books, database record editors, text editors, calculators,
data-entry pages, setting pages, and almost everywhere data needs to be captured from users.
The HTML form is the primary mechanism for collecting data from users. It plays a central role in
advanced Web site construction, because advanced Web sites require user interaction.
Unfortunately, books on HTML cannot fully explain the capabilities of HTML forms. Without a
complimentary engine such as HTML/OS, HTML tags are limited. That is why, although this book
assumes you have knowledge of HTML, it does not assume you have a complete knowledge of
HTML forms. This chapter fills this gap.
This chapter begins with a discussion of 10 HTML form components. You learn how each
component is initialized and how each passes data to variables when the user submits an HTML
form. Later in the chapter, the topic of how to validate user input is discussed. Then, at the end of
this chapter, you explore the construction of a fifteen-line spreadsheet editor and a six-line upload
page—two useful applications you can build with HTML form components.
Chapter Conventions—This chapter does not organize HTML form tags by their formal names.
Instead, combinations of HTML form tags, along with their defining attributes, have been isolated
-113-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
and called components. Unnecessary HTML attributes have been dropped. This chapter is a
practical guide, so we avoid those HTML tags that play a minimal role in advanced cross-browser
Web-site construction. Consult an official HTML reference if you need more complete description of
the form tags and attributes available in HTML.
The Components
You build almost all HTML forms using 10 main components, which you place between a set of
<form ACTION=destination> and </form> tags. The destination here refers to another HTML
document or an on-click Overlay. See the accompanied "The HTML/OS Destination" note.
The HTML/OS Destination—Hypertext links, GOTO tags, and ACTION parameters in HTML forms
require that you specify a destination. The destination may be a document name specified as full or
relative path, such as cart.html or /work/two.html. If no dot appears in the destination,
HTML/OS assumes the destination is an on-click Overlay. In these first two cases, the user never
sees the actual document name specified since it is encoded and hidden from the user. These links
are said to be inside HTML/OS. If the destination is written as a full URL, such as
https://www.securepost.com/capture.html, the link is said to be outside HTML/OS. The
users do see the URL in their browsers, and when they click it, they are sent directly to the remote
location without any passing of data to the server hosting the copy of HTML/OS. The data is
processed on a remote server, outside HTML/OS. This last kind of destination is not discussed
here. Here only destinations that stay inside of the HTML/OS environment are discussed.
Unlike many HTML tags, HTML forms cannot be nested (placed within each other) and neither can
their components. Furthermore, it is usually best to place only one HTML form on a Web page.
When a user clicks a Submit button in an HTML form, only the data in the HTML form passes to the
Web server. Data in other HTML forms on the page is lost. As a result, putting more than one HTML
form on a page can cause confusion for users and is not advised.
These 10 HTML form components, which are sufficient for building an advanced Web site or a Webbased application, are shown in Figure 10.2:
ƒ
Textarea
ƒ
Radio
ƒ
Select
ƒ
Checkbox
ƒ
Text box
ƒ
Upload
ƒ
Password box
ƒ
Submit
ƒ
Hidden
ƒ
Image
-114-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 10.2: These 10 HTML Form Components are all You Need to Build an Advanced Web Site.
Setting up an HTML form component requires three pieces of information. First, you need to know
the attributes of the HTML form tag that gives the component the look you want. For example, the
single-row input box uses the attribute size to set the length of the input box. As another example,
the input text area uses the attributes cols and rows to set the number of rows and columns of the
input text area.
Second, you need to know the content of the component when it's displayed. For example, if a pulldown menu appears on the screen, you may wish to set the default selection. Or if a text box
appears on the screen, you will want to determine the text in the box. In HTML/OS the HTML form
components are initialized with the values in the variable names specified in the HTML form at the
time of page rendering. For example, writing <input type=text name=abc size=20>, builds a
single-row input box on the screen that's 20 characters wide. The name=abc in the component tells
HTML/OS to initialize the box with the contents of the variable abc. Setting up the content of an
HTML form component is about making sure the content of the variable name in the HTML form is
what you wish to appear on the screen when the component is displayed.
Last, you need to know what information passes back to HTML/OS when the user submits the HTML
form. Typically, this is just the variable name defined in the form. However, in the special case when
multiple HTML form components, each with the same variable name, are placed in an HTML form,
the submitted data is placed in different rows of the same variable. This is discussed when relevant,
as each component is discussed.
A brief description, or capsule, of each HTML form component is provided here. Each includes the
component's name, its syntax, a brief description of how it is initialized, and how it passes data to
HTML/OS when it's submitted. An example is also provided for each component as well. Use these
capsules as a convenient reference.
Using Textarea
<textarea cols=no_of_columns rows=no_of_rows name=name></textarea>
This component displays an editable text area no_of_columns wide and no_of_rows high filled with
the contents of the variable name. One component is used per name. This component requires a
closing </textarea> tag. When submitted, the content of the text area is placed in the variable
name. When writing the tag you don't need anything between the <textarea> and </textarea>
tags. Placing text between them overrides the contents of the text area with that text.
Example:
<textarea cols=65 rows=15 name=mydata></textarea>
-115-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Using Select
<select name=name size=no_items>
<option value=v1>item1<option value=v2>item2<option value=v3>...
</select>
This component displays a pull-down menu. It uses the two HTML tags select and option. One
set of components is used per name. The first item in the pull-down menu on the screen is item1.
The second is item2 and so on. The pull-down menu will show no_items items at a time. A scroll bar
will appear on the screen if the total number of menu items exceeds the size specified. A menu item
will be highlighted on the screen if its value matches the contents of the variable name. When
submitted, the value associated with the item selected (v1, v2, v3, etc.) is saved to the variable name.
Example:
<select name=gender><option=M>Male<option=F>Female</select>
Placing the word multiple in the <select> header enables select boxes to accept more than one
selection. Menu items with values that match a row in the specified name are highlighted on the
screen. When submitted, the values associated with the items selected are placed in the specified
name, one per row.
Example:
<select name=addons size=2 multiple>
<option value=1>Onions<option value=2>Extra Fries<option
value=3>Coke
</select>
If addons is a one-column, two-row variable containing a 1 in one cell and a 3 in the other, the menu
items Onions and Coke will be highlighted. If the user selects both Extra Fries and Coke, addons
will become a one-column, two-row variable containing a 2 in one cell and a 3 in the other.
Using Text Box
<input type=text name=name size=no_cols>
This component displays a single-row input box, no_cols characters wide. Most often only one
component is used per name specified. The box is initialized by HTML/OS with the contents of the
variable name. When submitted, the values the user entered in the box are placed in the specified
name.
When more than one input box has the same name, the boxes load with different rows of the
specified name. When submitted, the values in each box are placed in the specified name, one per
row.
Example:
<input type=text name=mytext size=60>
Using Password Box
<input type=password name=name size=no_cols>
This component is similar to the text box (type=text). Here too a single-row text box, no_cols
characters wide, is displayed. Use one component per name and then initialize the box with the
contents of the variable name. But here, characters in the box are displayed as stars (*). Here too,
when the user submits the HTML form, the values entered in the box are placed in the specified
name.
Example:
-116-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<input type=password name=mypassword size=30>
Using Hidden
<input type=password name=name>
This component is an invisible HTML form component. You initialize the box with the contents of the
variable name. But it is not visible. You use one component per name. When the user submits the
HTML form, the value stored in the input box is placed in the specified name.
Example:
<input type=hidden name=myrecord>
Using Radio
<input type=radio name=name value=v1>
<input type=radio name=name value=v2>
<input type=radio name=name value=v3>
This component is a radio button. Unlike all the previous HTML form components, here the
component is repeated. Typically one component is specified for each possible value of the same
variable name. A radio button will appear selected on the screen if its value matches the contents of
the variable name. Radio buttons function so that the user can select only one button at a time (for a
given variable name). When the user submits the HTML form, the value in the component (v1, v2,
or v3) that he or she selected is saved to the variable name.
Example:
<input type=radio name=color value=red>Red<br>
<input type=radio name=color value=black>Black<br>
<input type=radio name=color value=green>Green<br>
Using Checkbox
<input type=checkbox name=name value=v1>
<input type=checkbox name=name value=v2>
<input type=checkbox name=name value=v3>
This component is a check box button. It is similar to the Radio component in that the component is
repeated for each possible value of the same variable name. Unlike the Radio component, the user
can select multiple check boxes for a given variable name. A check box will appear clicked on the
screen if its value matches the contents of any of the rows in column one of the variable name.
When the user submits the HTML form, the values in the selected components (v1, v2, or v3) are
saved to different rows of the variable name. This component behaves similarly to the Select
component (with the Multiple flag) discussed earlier in the "Using Select" section.
Example:
<input type=text name=extra value="SK150">Carrying Case<br>
<input type=text name=extra value="SK567">Extended Warranty<br>
Using Upload
<input type=file size=no_cols name=filename>
This component is a special file upload button. It looks like a text box with an extra Browse button
used for browsing the user's computer for a file. To use an Upload component, you need to add
enctype="multipart/form-data" to the HTML form header. An upload page is presented at
the end of this chapter. This component does not give you the ability to set the default contents of
the Upload box. When the user submits the HTML form, the file is uploaded and the name of the file
-117-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
in the Upload box is saved to the variable filename. The actual file is placed in the /upload
directory with the name filename.
Example:
<input type=file size=50 name=myfile>
Using Submit
<input type=submit name=name value=value>
This component is a submit button. When the user clicks it, the data in all components in the HTML
form are submitted. The submit button itself displays as the value specified. When the user submits
the HTML form, this value is placed in the specified name.
Example:
<input type=submit name=mybutton value="Continue">
Using Image
<input type=image border=width name=name src=image_name>
This component is another kind of submit button. Here too, when a user clicks the button, the data in
all HTML form components are submitted. The submit button itself is the image_name specified in
the component. The image will have a border, width pixels wide. Unlike other submit buttons, when
the user submits the HTML form, a value is not placed in the specified name. Instead the value TRUE
or FALSE is placed in the name. Your code can use this value to detect which button is clicked.
When a user clicks an Image button, HTML/OS captures the pixel coordinates where the button was
clicked and places those values in name.x and name.y where name is the variable specified in the
component.
Data Validation
Data validation involves checking whether the information placed in a component makes sense. For
example, you need to check a box asking for an e-mail address to ensure the data the user places in
the box is a valid e-mail address. In some cases, the input is required. At other times, it is not. In
addition, data validation involves presenting the information in a helpful way to users so they are less
likely to make mistakes. When users type entries, they often introduce errors. That's why it's better to
provide pull-down menus, radio buttons, or check boxes, if you can use them, to capture user input.
In many situations, however, users have no choice but to type their entries. In this section, you look
at the code necessary to detect user errors, and to alert them of their mistakes as well as ensure that
the data users enter in HTML forms is validated.
Avoid Browser-Side Data Validation—Browser-side data validation involves adding JavaScript
functions to HTML forms. The functions are activated before the HTML form is submitted, alerting
the user of invalid data. These JavaScripts however have problems. First, some browsers will
crash or complain when they run JavaScripts. Second, technically speaking, JavaScript does not
protect your site from bad data, because JavaScript can be turned off or bypassed. These reliability
problems are unsolvable. Browser-side data validation should be avoided.
Writing the Code
You add data validation to HTML forms in the same manner regardless of the HTML form you are
working with. When the user submits the HTML form, an on-click Overlay runs. At the top of this
Overlay, you add IF-THEN statements to check the data entered in the form. If an error is found an
error message is set and the page is redisplayed. If no problems are found, the instructions already
in the on-click Overlay run. The HTML form is typically structured as follows:
-118-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<html>
<title>Sample HTML Form</title>
<form ACTION=process_me>
<<[Location1]>>
html form components go here
</form>
<<OVERLAY process_me
[Location2]
other instructions go here
>>
[Location1] is where you place the error message. The code you place at [Location1]
displays the error message if it exists. For example, the following code can be used:
IF fp_error != "ERROR" THEN
DISPLAY fp_error /DISPLAY
/IF
fp_error="ERROR"
When this Overlay first runs, fp_error is ERROR so the Overlay does nothing. If, on the other hand,
a value is placed in fp_error, the contents of fp_error are placed in the document, alerting the
users of the error they made.
[Location2] is where you place the data validation instructions. For example, an HTML form that
requests the e-mail address my_var, might have an on-click Overlay with the following instructions:
fp_error=""
IF COUNT(my_var, "@")!=1 OR
LENGTH(my_var)<7 OR
COUNT(my_var," ")>0 OR
COUNT(my_var,".")=0 THEN
fp_error=fp_error + "Bad E-mail Address.<br>"
/IF
IF fp_error != "" THEN
fp_error="<font color=red>Error: "+fp_error+"<br>Please Try
Again.</font>"
GOTO PAGE
/IF
First, fp_error is initialized to the empty string. Then an IF-THEN statement runs. You would use
this particular IF-THEN statement to validate an e-mail entry. If the entry doesn't have a single at
sign (@), if it doesn't contain a dot, if it has a space, or if it's less than seven characters in length, the
e-mail address is invalid. You can place additional IF-THEN statements after this one if you need to
validate data entered into other input boxes.
-119-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
After that an IF-THEN statement sees whether fp_error is still the empty string. If it is not an
empty string, an error occurred. In this case, fp_error is beautified a bit and the page is
redisplayed. If fp_error is still an empty string, no validation error has been detected so
instructions you place after the code can run.
The Thirty-Line Guest Book
As an example, consider the task of building a guest book with data validation. Names users enter in
the guest book are appended to a comma-delimited file.
Here you place an input form in an HTML form, requesting a user's full name, e-mail address, and
phone number. You validate all three inputs. If successful, you append the user data to a commadelimited text file and redisplay the page with a Thank You message. If not, you redisplay the page
with an error message. The page requires about 30 lines of HTML/OS instructions and 4 HTML form
components. You use the seven HTML/OS tags IF-THEN, DISPLAY, COUNT, TRIM, LENGTH,
REPLACEALL, and GOTO. For details on these tags see Appendix D, HTML/OS Tag Reference Guide.
Here's the code:
<html>
<title>Guest Book Page</title>
<b>Guest Book</b><br>
Enter personal information below so we can update you from time
to time.
<form ACTION=process_it>
<<IF fp_error != "ERROR" THEN
DISPLAY fp_error /DISPLAY
/IF
fp_error="ERROR"
>>
Full Name: <input type=text size=60 name=guest_fullname><br>
E-mail Address: <input type=text size=60 name=guest_email><br>
Day time Phone: <input type=text size=60 name=guest_phone><br>
<input type=submit value="Add To Guest Book">
</form>
</html>
<<OVERLAY process_it
fp_error=""
my_var=guest_fullname
IF COUNT(TRIM(guest_fullname)," ")=0 OR
LENGTH(guest_fullname)<4 THEN
fp_error=fp_error + "Enter Full Name.<br>"
/IF
my_var=guest_email
IF COUNT(my_var, "@")!=1 OR
-120-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
COUNT(my_var," ")>0 OR
LENGTH(my_var)<7 OR
COUNT(my_var,".")=0 THEN
fp_error=fp_error + "Bad E-mail Address.<br>"
/IF
my_var=guest_phone
my_var=REPLACEALL(my_var,")", "")
my_var=REPLACEALL(my_var,"(", "")
my_var=REPLACEALL(my_var,"-", "")
my_var=REPLACEALL(my_var," ", "")
IF ISINTEGER(LEFT(my_var,7))="FALSE" OR LENGTH(my_var)<7
THEN
fp_error=fp_error + "Bad Phone Number.<br>"
ELIF ISINTEGER(LEFT(my_var,7))="TRUE" AND LENGTH(my_var)<10
THEN
fp_error=fp_error + "No Area Code in Phone Number.<br>"
IF fp_error != "" THEN
AM
FL
Y
/IF
fp_error="<table border=0 width=700><tr><td
bgcolor=red>"+
"<font face=arial,helvetica size=2
color=white>"+
Again.</font></td>"+
"</tr></table>"
TE
"Error: "+fp_error+"Please Try
GOTO PAGE
ELSE
fp_error="ERROR"
/IF
APPEND ROW(guest_fullname,guest_email,guest_phone) TO
FILE="myguests.txt" TS="," /APPEND
guest_fullname=""
guest_email=""
guest_phone=""
fp_error="<table border=0 width=700><tr><td
bgcolor=green>"+
"<font face=arial,helvetica size=4 color=white>"+
"Thank You.</font></td></tr></table>"
GOTO PAGE
>>
-121Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
[Location1] and [Location2] in the earlier code in the "Writing the Code" section are replaced
with instructions appearing in bold. [Location1] displays fp_error if fp_error contains a value
other than ERROR. [Location2] runs three validation statements, appends a row of data to the
file myguests.txt, clears the variables used in the guest book, sets fp_error to a Thank You
message, and redisplays the page.
In the previous example, you validated an e-mail address. You also need to validate other kinds of
data. Furthermore, sometimes you require the data and sometimes you do not.
As a rule, data that's not required should still be validated. If the data is not required, use the same
instructions to test the data's validity; but also accept an empty string as a valid input. Here are
instructions for validating e-mail, integers, a single word, a phone number, a full name, a zip code, a
credit card number, and a currency. In each case, the variable you're testing is called str and the
variable req is preset to Y in the event the input is required. In each example, if an error is detected,
an error message is placed in the variable fp_error.
# validating e-mail /#
IF req !="Y" AND str != "" THEN
IF (COUNT(str,"@")!=1 OR LOCATE(str,".")=0 OR
LOCATE(str," ")!=0 OR LENGTH(str)<7) THEN
fp_error="Invalid e-mail.<br>" /RETURN
/IF
/IF
# validating integers /#
IF req !="Y" AND str != "" THEN
IF ISINTEGER(str)!="TRUE" THEN
fp_error="Invalid number.<br>" /RETURN
/IF
/IF
# validating single word /#
IF req !="Y" AND str != "" THEN
IF LOCATE(TRIM(str)," ") != 0 THEN
fp_error="Invalid word.<br>"
/IF
/IF
# validating phone number /#
IF req !="Y" AND str != "" THEN
str=REPLACEALL(str,")", "")
str=REPLACEALL(str,"(", "")
str=REPLACEALL(str,"-", "")
str=REPLACEALL(str," ", "")
IF ISINTEGER(LEFT(str,7))= "FALSE" OR LENGTH(str)<7 THEN
fp_error="Invalid phone number.<br>"
-122-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ELIF ISINTEGER(LEFT(str,10))= "FALSE" OR LENGTH(str)<10
THEN
fp_error="Invalid phone number. No area code?<br>"
/IF
/IF
# validating full name /#
IF req !="Y" AND str != "" THEN
IF (COUNT(TRIM(guest_fullname)," ")=0 OR
LENGTH(guest_fullname)<4) THEN
fp_error="Invalid full name.<br>"
/IF
/IF
# validating U.S. zipcode /#
IF req !="Y" AND str != "" THEN
IF (LENGTH(str)<5 OR
ISINTEGER(LEFT(str,5))="FALSE" OR
ISINTEGER(REPLACEALL(str,"-",""))="FALSE") THEN
fp_error="Invalid zip code.<br>"
/IF
/IF
# validating credit card /#
IF req !="Y" AND str != "" THEN
str=REPLACEALL(REPLACEALL(REPLACEALL(str,"-",""),"
",""),",","")
IF ISMOD10(str)= "FALSE" THEN
fp_error="Invalid credit card number."
/IF
/IF
# validating currency /#
IF req !="Y" AND str != "" THEN
str=TRIM(REPLACEALL(REPLACEALL(str,",",""),"$",""))
IF str != FORMAT(str,"normal",2) THEN
fp_error="Invalid Data Type"
/IF
/IF
You can use these instructions in your code to validate various kinds of data. You can also use them
to build a VALIDATE tag using HTML/OS functions. An exercise at the end of this chapter, for those
readers familiar with the concept of defining functions, is provided.
-123-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The Fifteen-Line Spreadsheet Editor
Suppose you want to give users the ability to edit the contents of a table of information. For example,
a site may include a table of products currently on sale—information that needs to be edited by staff.
As another example, you may need to provide a screen for filling out the x-y coordinates of a chart.
In both these cases, a spreadsheet editor is necessary for editing a delimited text file containing the
table of information. A sample spreadsheet editor is shown in Figure 10.3.
Figure 10.3: This Spreadsheet Editor Gives Users the Ability to Edit Two-Dimensional Data.
To build the editor, you must place multiple input boxes on the screen in a two-dimensional grid.
Each input box must have its own name so it can be initialized and pass data back to HTML/OS. You
accomplish this by specifying the exact cell in the specified name for each component. For example,
to build a grid with two columns and two rows, you can write the following code:
<input type=text name=my_var[1,1] size=5>
<input type=text name=my_var[2,1] size=5><br>
<input type=text name=my_var[1,2] size=5>
<input type=text name=my_var[2,2] size=5>
When HTML/OS renders these four components, it initializes them with the values in the cells
specified. When a user submits the HTML form containing these components, the values the user
enters in the text boxes replace the values in these cells.
To build a spreadsheet editor, you can type multiple components, as shown previously, into a Web
page. However, this method lacks flexibility. You may want to make the editor adjust to the full size
of a table that has been read from a file. You may want the editor to be adjustable by the user. You
cannot provide these options if you simply type the components into a document. Instead, you
should place the components dynamically in the page. An editor built this way follows:
<<
csv_file="edit_csv.txt" # data file name /#
csv_x=5
# width /#
csv_y=5
# height /#
COPY FILE=csv_file TS="," TO csv_array /COPY
>>
<html>
-124-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<title>Simple Spreadsheet Editor</title>
<form method=post action=do_csv>
<center>
<table border=1 cellpadding=3 cellspacing=0>
<tr><td align=right bgcolor=#777777>
<input type=submit name=mybutton value="Save">
<input type=submit name=mybutton value="Load">
</td></tr>
<tr><td>
<table border=0 cellpadding=0 cellspacing=0>
<<
y=1
WHILE y <= csv_y DO
DISPLAY "<tr>" /DISPLAY
x=1
WHILE x <= csv_x DO
DISPLAY "<td><input type=text size=10 "+
"name=csv_array["+x+","+y+"]></td>"
/DISPLAY
x=x+1
/WHILE
DISPLAY "</tr>" /DISPLAY
y=y+1
/WHILE
>>
</form>
</table>
</td></tr></table>
</html>
<<OVERLAY do_csv
IF mybutton="Save" THEN
COPY csv_array TO FILE=csv_file TS="," /COPY
ELIF mybutton="Load" THEN
COPY FILE=csv_file TS="," TO csv_array /COPY
/IF
GOTO PAGE
>>
The Underlay on the page initializes the settings for the page and loads the table of information from
a delimited text file. Inside the page, an Inlay uses a WHILE loop to display the HTML form
-125-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
components. Note that the HTML form includes two submit buttons; one to save and one to load the
file. Also see how the name in the input box is composed. The code "name=csv_array[" + x +
"," + y + "]></td>" builds the name used in the component. It is important to remember that
the column and row numbers need to be dynamically generated since they are different for each
component.
This spreadsheet editor was built with the four HTML/OS tags COPY, IF-THEN, WHILE, and
DISPLAY. It required fifteen HTML/OS instructions and two submit buttons. More advanced
spreadsheet editors are available in the Clip library in the Aestiva Web site's User Center at
http://www.aestiva.com/support/.
The Six-Line Upload Page
The ability to upload a file is a feature that's needed in many Web-based systems. Whether you want
to build a page that uploads comma-delimited files, spreadsheets, documents, or images, you first
need to know how to build an upload page.
The Upload component described in the beginning of this chapter is the component you use to build
this page. First you place the Upload component in your HTML form. Then you add a special
attribute to the HTML form header so it accepts file uploads (you see that shortly). When the user
submits the HTML form, the file that's uploaded is saved to a file. The file needs to be unique for
each user, because multiple uploads may be occurring simultaneously. The file is then moved to its
desired destination name.
Here's an upload page that uses six HTML/OS instructions, an Upload component, and a submit
button. The page allows users to upload a company logo, mylogo.gif:
<html>
<title>Six-Line Upload Page</title>
<b>Upload Company Logo</b>
<form enctype="multipart/form-data" ACTION=upload_me><br>
<font size=1>FILE</font><br>
<input type=file name=file<<usernum>> size=20>
<input type=submit value="Upload Company Logo Now">
</form>
</html>
<<OVERLAY upload_me
temp=GETCOLEQ(FILELIST("/upload"),1, "file"+usernum)
IF temp[1]= "file"+usernum AND temp[2] > 500 THEN
temp=SYSMV("/upload/file"+usernum, "mylogo.gif")
GOTO "upload_succeeded.html"
ELSE
GOTO "upload_failed.html"
/IF
>>
-126-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The HTML form header includes the enctype="multipart/form-data" attribute, which you
need to include whenever you use Upload components in HTML pages. Technically speaking the
attribute tells the HTML browser how to encode the uploaded file. Without it file uploads fail.
The Upload component saves uploaded files in the /upload folder with the name specified in the
name-value pair in the tag. Here the name-value pair is name=file<<usernum>>. USERNUM is an
Overlay tag that contains the user's session number. As a result, this Upload component always
places the uploaded data in a unique file for that user. This way of specifying the name is necessary
to allow two users to upload a file at the same time without overwriting each other.
When a user submits the HTML form, the contents of the file on the browser computer is uploaded
and saved to /upload/file356, for example (for a user with a USERNUM of 356), and then the onclick Overlay upload_me runs.
The first instruction in the on-click Overlay lists the files in the /upload folder, extracting from the
file list only the row that matches the filename of the uploaded file. You use the FILELIST and
GETCOLEQ tags to accomplish this. The tags are nested, meaning you place one inside the other. To
understand how nested tags work, always start from the inside and work out. The inner-most
HTML/OS tag is FILELIST, which produces a multicolumn variable. Column 1 contains filenames.
Column 2 contains file sizes. This table is used as the first parameter of GETCOLEQ, an HTML/OS
tag that extracts from a table only those rows in which a particular column (in this case column 1)
matches a specified value (in this case the file name). See Appendix D, HTML/OS Tag Reference
Guide for a detailed description of these tags.
The matching row is placed in the variable temp. When a user successfully uploads a file, the first
cell of temp contains the filename, and the second cell contains the number of bytes uploaded. As a
precaution, both of these cells are checked before moving the uploaded file to its intended
destination—mylogo.gif.
Summary
In this chapter, you reviewed the 10 HTML form components used in advanced Web-site
construction. Interestingly enough, although these 10 components are powerful, as a user of the
Web, you are already familiar with them. They are the building blocks of all Web sites. No matter
how advanced you get, these components stay the same.
You also learned some data-validation techniques and reviewed two applications: a fifteen-line
spreadsheet editor and a five-line upload page. We recommend you copy the code in this chapter
into your copy of HTML/OS and modify the code until you are comfortable using all 10 HTML form
components.
Exercises
The following exercises require heavy use of HTML forms and a thorough understanding of
HTML/OS fundamentals. Some of the exercises are fairly involved. Complete these exercises to
reinforce what you've learned in this chapter and in Part II, Programming Basics. Answers to all
exercises are provided on this book's companion Web site as described in the book's Preface.
Exercise 1
Build a calculator that can add, subtract, divide, and multiply two numbers. Build it using the following
sets of HTML components:
-127-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
ƒ
ƒ
Build the calculator using two text boxes with a pull-down menu (with the operators
add, subtract, multiply, and divide) placed between the two and followed by a
Calculate button. Show the calculation result when the page is redisplayed.
Build the same calculator using only pull-down menus. Replace each text box with
five pull-down menus each containing the numbers 0 through 9.
Build the calculator using a long text area with buttons underneath it. Make it
resemble a calculator touch pad on your keyboard. In other words, the buttons
should include a three-by-three grid of numbers between 1 and 9. Then to the right of
these, place the buttons 0, 00, and Enter. Above these, place the four operators
Add, Subtract, Multiply, and Divide.
Exercise 2
Build a set of instructions that validate a date. Hint: Use the ISDATE tag, which is described in
Appendix D, HTML/OS Tag Reference Guide.
Exercise 3
Build a VALIDATE tag that takes three parameters; the data being validated, a data-type code (such
as email, word, creditcard, etc.), and a req flag that is set to y if the field is required. Use the
instructions provided earlier in this chapter to build the function. Use the FUNCTION tag to do this. It
is described in Appendix D, HTML/OS Tag Reference Guide.
Exercise 4
The spreadsheet editor presented in "The Fifteen-Line Spreadsheet Editor" section of this chapter
reads values organized in a two-dimensional grid. Replace the text boxes with check boxes, and add
a Clear button and a Next Generation button to create the classic Game of Life. Hint: The Game of
Life is played on a two-dimensional grid. Each cell is either alive/on or dead/off. The new state of
each cell (The Next Generation) is determined by its old state and the sum of the alive cells among
its surrounding eight nearest neighboring cells. The rules are as follows: A cell in the next generation
is alive if a living cell is surrounded by either 2 or 3 alive cells or if a dead cell flips into the alive state
by being surrounded by exactly 3 living cells. Otherwise it dies. These special rules were invented by
the mathematician J.H. Conway. In the actual game the grid is infinite. In this case, you can use a
15-by-15 set of cells.
Exercise 5
Build a color selector bar—a bar with 16 colors on it. When the user clicks a color the selected color
displays on the screen. Hint: First create an image bar, 160 pixels wide with one color for every 10
pixels. Then note that Image components can tell you the pixel location on the image clicked when
the user submits the image. Use this to determine the color selection.
-128-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Part III:
Database Programming
Chapter List
Chapter 11: The Web Database
Chapter 12: Building Query Pages
Chapter 13: Building Database Reports
Chapter 14: Building Database Editors
Chapter 15: Database Networking
-129-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 11: The Web Database
This chapter introduces you to the Web database architecture used by HTML/OS. You learn about
the kinds of database fields available to you, how to pass data to and from databases, how to merge
data from a database into a Web page, and how to build an eight-line database record editor. By the
end of the chapter you'll have a basic understanding of how to work with database records. The
chapter starts with a brief discussion of how the Web database used by HTML/OS differs from
conventional database architectures.
The main thing to note about the HTML/OS database architecture is that it's very reliable and easy to
use since it is not built from multiple products and components designed prior to the advent of the
Web. The architecture purposefully lacks the complexities of legacy systems.
Web versus Legacy Databases
Before diving into how HTML/OS databases work, it is important to note that the Web databases
used by HTML/OS have architectural features distinct from legacy databases—databases
engineered prior to the advent of the Web. The differences are widespread, spanning how data is
displayed, maintained, accessed, and stored. If you are unfamiliar with legacy database
architectures, that's okay. But you may find this section's discussion helpful to more fully appreciate
the advantages of HTML/OS databases.
First and foremost, legacy databases run as stand-alone products without the Web. These products
were originally designed for delivering information to standard computer screens, so the products are
often poor or incapable of rendering Web pages. Most leave the task of Web page rendering to
integration tools, also known as middleware products.
These legacy databases connect with middleware products via database servers used to serve
database information to other programs. On the Web however, you already have servers that can
serve database information—the same Web servers that serve Web pages. The HTML/OS database
architecture takes advantage of this. With HTML/OS, database servers and middleware products are
not needed. For you this means no database server to set up and maintain. This also increases site
reliability since, if the database server fails, the Web site fails.
Second, the Web database used by HTML/OS provides a seamless connection between data in
databases and Web pages. Single instructions are used to move data back and forth among Web
pages, HTML form components, and database tables. Legacy databases hooked up to the Web do
not do this. They must channel data through database gateway languages such as SQL (structured
query language). In HTML/OS there is no such limitation.
Last, the HTML/OS database is Web-based. Administration, programming, and data maintenance is
done through a browser. Legacy databases, at best, are partially administered through a browser.
Changes to HTML/OS databases can be performed from the nearest computer hooked up to the
Web.
The HTML/OS Database Architecture
Although HTML/OS databases have many differences with legacy databases, their primary purpose,
the storage of large amounts of well-organized information, is the same. Like other databases, the
HTML/OS database organizes information in tables subdivided into multiple records, which
themselves are subdivided into one or more fields.
In HTML/OS, when setting up a database, you decide on the database tables and fields needed for
-130-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
your project and create them with a point-and-click application called dbConsole. There are no
schemas to set up, as discussed in the accompanying note, "No Schemas to Define."
No Schemas to Define—In HTML/OS a database and a table are the same. In legacy systems a
database has multiple tables, each related to the next with schemas—sets of definitions that link
fields together. In HTML/OS these schemas are not needed. This makes it easier at a later date to
change or add fields to your tables to accommodate design changes or add new features. Note
that although the HTML/OS database is relational, these relations do not need to be set up in
advance. Instead, they are set up within Overlay tags. This topic is discussed in Chapter 12,
Building Query Pages.
The Database Table
The HTML/OS database table is composed of data and indexes (cross-referencing data used to
speed up access to data) stored in multiple files, each placed in the same directory. These files
share the same name except for their extensions. Data is placed in files with the extension db. Index
information is placed in files with the extension idx. The name of a database is the name of these
files without their extension. For example, the database /work/customers uses the files
/work/customers.db and /work/customers.idx.
AM
FL
Y
Like HTML files, databases can be specified by full path or relative to the current directory. A Web
page in the /work directory, for example, can reference the /work/customers database as
/work/customers or simply customers.
Field Types
TE
Each record in a database is composed of multiple fields. The first field of all HTML/OS records is
called RECORD. It contains a unique number, called the record ID, which is determined automatically
when a record is created. All other fields are definable. The naming convention used to name fields
is the same as that used for variables, as explained in Chapter 6, Variables, Conditionals, and
Loops.
Each field can store one type of information. The requirement that fields store only one kind of
information is used as a convenience to the HTML/OS engine. Without predefined field types,
computer searches run slower. The field types used by HTML/OS, along with their 3-letter codes,
are; String or STR, Integer or INT, Floating Point or FLT, Currency or DOL and Date or Time or DAT.
Strings
String is computer talk for text. In HTML/OS, strings can be 1 to 100,000 characters in length, like
cells in HTML/OS variables. They can contain text, HTML, names of images and binary files, logical
values (TRUE or FALSE), or entire text documents.
Integers
These are whole numbers, negative or positive. They can be between 1 and 11 characters, like
integers in variables. Note that credit card numbers and social security numbers are not integers.
They are too long and often contain other characters such as spaces and hyphens. Use STR for
those situations.
Floating Points
Floating point is computer talk for fraction. These are numbers, positive or negative, up to 11
-131Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
characters in length, inclusive of a decimal point. They have the same limits placed on fractions used
in HTML/OS variables.
Currencies
Currencies are fractions with two digits after the decimal point. Currencies are limited to 999,999,999
in whatever currency you are using.
Date and Time
Date and Time fields can contain dates, times, or a combination of dates and times. These fields
must obey the same rules that dates and times obey when stored as variables. See the "Using
Different Kinds of Variables" section of Chapter 6, Variables, Conditionals, and Loops for specific
rules and limits.
Working with Databases
Using HTML/OS databases usually requires two steps. First you set up the databases you need in
dbConsole. See the accompanying "On-the-fly Database Construction" note regarding the use of the
word "usually."
On-the-fly Database Construction—Database setup is "usually" performed within the dbConsole
application. But it doesn't have to be. You can use Overlay tags. In circumstances where you want
your application to set up databases or provide database maintenance functionality, Overlay tags
such as DBCREATE, DBPURGE, DBIMPORT can be used. Most often however, you want to use
dbConsole since it is point-and-click. See the knowledge base on the Aestiva Web site at
http://www.aestiva.com/support/ for information about building databases on the fly.
Second, once the databases are set up, you read and write to them using Overlay tags placed in
your document. The Overlay tags transfer data between databases and HTML/OS variables. These
Overlay tags, called DB tags for short, begin with the characters DB. They have names like DBADD,
DBDELETE, and DBEDIT.
You learn how to use these tags in the section "Using DB Tags," later in this chapter. Here you look
at the options provided to you by dbConsole.
Using dbConsole
You first encountered dbConsole in Chapter 4, Your First Web Database Program. There you used
dbConsole to build a database and populate it. Here you look at the other options available to you.
To run this application, click the dbConsole icon on your HTML/OS desktop. Click Open to select a
database. You should see a screen like that shown in Figure 11.1. An Options menu appears on the
left side of the screen. You use the first option, Open, to select a database table. See the attached
note, "Open Means Selected."
-132-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 11.1: The Database Selection Screen Highlights Databases Making Them Easy to Recognize.
Open Means Selected—The term "open" in dbConsole is used to indicate the selection of
database tables. It does not mean "open" in the computer networking sense. Web databases do
not keep connections to databases open or persistent as you work with them, since the Web pages
that access databases are themselves not persistent. Web databases have open connections to
them only in the split second they are accessed.
Once a database has been selected, you can use the options Edit, Purge, Index, Change, Import,
Export, Copy, Delete, and Info. You can use Open and Create used selecting a database. These
options are discussed separately in the sections that follow.
Edit
Click Edit to pull up the first record of your selected database table. The name of that database is
shown at the top of the right pane of dbConsole's window. You use Edit to edit records in the
database manually. Options to add, delete, and modify the contents of fields in the record are
provided. An Autosave option is also provided. When you place a check mark next Autosave,
dbConsole automatically saves the current record as you move forward and back among the records.
Purge
Click Purge to expunge the database of deleted records—records that you have internally marked
as deleted (see the "Copy and Delete" section later in this chapter) but still occupy space in the
database. Purge rewrites the database files, eliminating these empty spaces in the table. Purging
eliminates empty space and makes databases smaller, faster, and more manageable. Once a
database has been purged, the deleted records become unrecoverable. Until then, you can recover
deleted records with the DBSEARCHX and DBGETX tags. These tags are not described in this book
but are described in the Aestiva knowledge base at http://www.aestiva.com/support/.
Index
Use this option to rebuild cross-referencing files used by HTML/OS to access the database. Indexing
is advised if you have more than a two thousand records that have not been indexed. Indexing
improves the performance of database searches and record retrieval.
In HTML/OS you can build these cross-referencing files or indexes in steps. This ability to index
incrementally is provided so large databases, otherwise requiring a lot of time to index (more time
-133-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
than the browser has available to it before it times out), can be indexed in stages. Click Index, set
your index increment, and click the Index button. The index increment is the number of records to be
indexed per step. A database smaller than 100 megabytes can be indexed in a single step. For
databases with record sizes of about 2 kilobytes, an increment value of 50,000 is recommended.
Repeat until no remaining records need to be indexed.
Create
This option builds a new database table. When a database is selected the structure of the currently
selected database is used as a starting point. When no database is selected, an empty form is
provided. Click Create to view an input form used for defining the fields in a table. Each row of the
form has a place for you to specify a field name, a field type, whether the field should be indexed,
and a field length. Click the Add Fields button above the input form to enlarge your input form to
accommodate additional fields. A Create screen is shown in Figure 11.2. As you specify the fields in
your database keep in mind the following:
Figure 11.2: You Can Expand the Database Create Screen to Accommodate Additional Fields by
Clicking the Add Fields Button.
ƒ
ƒ
ƒ
ƒ
Field names are specified like variable names. The names cannot begin with a
number and must contain only numbers, letters, and underscores, but no special
characters such as a space or a dollar sign.
When specifying a field type make sure you do not confuse integers with strings.
Credit card numbers, Social Security numbers, and large IDs are strings. Logical
values such as TRUE and FALSE are also stored in the database as STR. If you
plan on referencing images, sound or video files, or any binary files, place the
names of those files in a STR field and the files in a directory of your liking. Note
that the entries for the field do not need to contain the full URL to the image or
binary file. It does not need to contain the full path to the file either. Just the name
of the file is needed. The path to the file can be specified later in the HTML
document.
Leave indexing set to Y so you can search your fields. Setting a field index to N
disables the index for that field. That will save some hard disk space but these
days, hard disk space is not a premium.
If you specify a field type of STR, it can be 1 to 100,000 characters in length. If you
specify a field type of INT, FLT or DOL, specify a length of 1 to 11. Anything
more is a waste. If you specify a field type of DAT, specify a field length of 16 to
leave room for both dates and times.
-134-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Once you have defined the fields you want for your database, enter a name for the database, and
click Create Database. After creating a database, you can select it with Open, import data into it,
manually add data to it with Edit, and access it with Overlay tags you place in HTML documents.
Change
Use the Change option to change the fields defined for a database that has already been created
and populated with data. Figure 11.3 shows a Change database screen. It includes options to add,
delete, and change fields.
Figure 11.3: The Change Database Screen Includes an Add Group Option for Adding Group Fields.
To change a field, select it and click Change. Then make the changes you want, return to make
changes to any other fields you want, and click Save Changes when done. The database structure
will be changed, preserving the data in the database. Note that reductions in field sizes are not
advised unless you are sure none of your fields exceeds the newer smaller field size. Otherwise data
may be truncated, which means that some of the data may be lost.
Group Fields
The Change database section also includes a button called Add Group, which gives you the ability
to add a special type of field, called a group field, to the record definition. Any number of group fields
may be added. Group fields are virtual fields built from one or more regular fields—like those defined
on your Create screen. When a group field is searched, HTML/OS searches through a special index
defined for all the fields belonging to it.
Group fields have another property too; they are word-indexed. This means that all words within the
fields defined for the group field are internally cross-referenced. Multiword searches of group fields
are much faster than searching within standard fields. The indexes used with standard fields are not
indexed by each word in the field. Group fields are discussed in greater detail in Chapter 12, Building
Query Pages. The Group field is useful when building high-speed searches where, for example,
users enter one or more words in a single Find box on a Web page.
Import
Use this option to load many records at one time from a text file into a database. Suppose, for
example, you have data in a legacy database or spreadsheet you wish to load into your Web
-135-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
database. Your first step is to export the data from the other database or spreadsheet into a
delimited text file. You use the Save As or Export option of your other database or spreadsheet
program to do this. Save the file as a tab-delimited, CSV, or comma-delimited file. Almost all
database and spreadsheet programs have an option to export data as some kind of delimited text file.
Once your data has been exported as a file, upload it using the File Manager to your HTML/OS site.
Now you're ready to import it.
Click Import on your dbConsole menu and select your delimited text file. You will see a screen like
that shown in Figure 11.4.
Figure 11.4: The Import Option in dbConsole Allows You Load Various Kinds of Delimited Files Into
HTML/OS Databases.
Select the delimiter of your file and click Continue. The Import option will read in the first line of the
file and list it on the left side of the page as shown in Figure 11.5. To the right of each column entry is
a pull-down menu allowing you to select the field the column of the delimited text file should be
loaded into. To ignore a column of text in the file, set the pull-down menu to Skip. When done, click
Import and all the records in the file will be loaded into the database. After importing your data you
will want to check it. Click Edit to browse the imported records to ensure data has been placed in
their proper fields.
Figure 11.5: When Importing From a Delimited File, For Every Column in the File, You Specify the Field
in the Database it Should Fill.
-136-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Export
The Export option exports a database to a comma-delimited text file. Use this option if you need to
transfer data from an HTML/OS database to a legacy database or spreadsheet.
Copy and Delete
The Copy and Delete options allow you to copy or delete a database table. These options copy or
delete all the files associated with the specified database. These files include db and may include
idx if the database is indexed. Note that databases can also be moved and deleted with the
HTML/OS File Manager.
Info
Click Info to view the structure of your database along with general information about the database.
This is a good screen to print for a handy reference of the fields in your database. See Figure 11.6.
Figure 11.6: The Info Screen in dbConsole Provides a Snapshot of Your Database.
Using DB Tags
Once you've created a database and added some records to it, or perhaps loaded it with data
imported from another database, you are ready to add Overlay tags to your document to access it.
DB tags, the Overlay tags that work with databases, are available for performing, among other tasks,
basic operations such as reading, editing, or adding records to a database. They are discussed here.
More advanced tags for indexing, purging, and copying databases are also available but not
discussed here.
DB tags are described in Appendix D, HTML/OS Tag Reference Guide and the Aestiva knowledge
base at http://www.aestiva.com/support/.
The Eight-Line Database Editor
Let's look at an eight-line database editor that lists records in a guest book database and allows
users to specify a record to edit, delete, or copy to a new record. We introduce this Web page to
illustrate the basic DB tags; DBGETREC, DBEDIT, DBADD, and DBDELETE. The eight-line database
editor is shown in Figure 11.7.
-137-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 11.7: This Eight-Line Database Editor is Great at Editing Small Databases.
The code for the database editor in Figure 11.7 is as follows:
<html>
<title>Eight-Line Database Editor</title>
<font size=5><b>8-Line Database Editor</b></font><br>
Records in Database:
<< myrecords=DBFIND("tinydb","",1,100,"record")
DISPLAY COLTOLIST(myrecords,1," ") /DISPLAY
>>
<form method=post action=dbopts>
Record ID: <input type=text name=record size=10>
<input type=submit name=mybutton
value="Load">&nbsp;&nbsp;&nbsp;
<input type=submit name=mybutton value="Save">
<input type=submit name=mybutton value="Delete">
<input type=submit name=mybutton value="Add"><hr>
Full Name: <input type=text name=f1><br>
E-mail Address: <input type=text name=f2><br>
Phone Number: <input type=text name=f3><br>
</form>
</html>
<<OVERLAY dbopts
IF mybutton="Load" THEN
temp=DBGETREC("tinydb",record)
ELIF mybutton="Save" THEN
temp=DBEDIT("tinydb",record)
ELIF mybutton="Delete" THEN
temp=DBDELETE("tinydb",record)
ELIF mybutton="Add" THEN
temp=DBADD("tinydb",record)
/IF
GOTO PAGE
>>
-138-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The eight-line database editor uses the tinydb database with the fields f1, f2, and f3. The editor first
lists the records in the database. This is followed by a row that contains an input box to enter a
record number followed by Load, Save, Add, and Delete buttons. Below that row are three input
boxes for entering a full name, e-mail address, and phone number.
The first Overlay in the eight-line database editor reads the database tinydb and places a column
of record numbers in the variable myrecords. The DBFIND tag, as you may recall, is the same
Overlay tag used and described in Chapter 4, Your First Web Database Program.
This instruction is followed by one that displays the value returned by COLTOLIST(myrecords,1,"
"). The COLTOLIST tag converts a column of numbers in a variable to a list, with each list entry
separated by a special character known as a delimiter. The first parameter of COLTOLIST is the
variable from which a column is extracted. The second is the column number, and the third
parameter is the delimiter. Here column 1 was specified and the delimiter is a space.
The tag specified here, for example, would convert a variable with the word Larry in column 1, row
1 and [email protected] in column 1, row 2 and 415-555-1234 in column 1, row 3, into the
text Larry [email protected] 415-555-1234. Here COLTOLIST is used to convert the
column of records returned by the database search into a list of record numbers, each separated
from the next with a space.
Below this instruction is an HTML form linked to the on-click Overlay dbopts. When a user clicks
any of the buttons in the HTML form, the on-click Overlay dbopts runs. The dbopts Overlay
contains an IF-THEN statement. The Overlay that runs depends on which button the user presses.
In HTML/OS, only a single DB tag is required to perform most database operations, because most
DB tags work with all the fields in a record at one time.
If the user clicks the Save or Add button, the tags DBEDIT or DBADD run. These tags fill the fields of
a database record using the values in variables named the same as the fields used by the database.
For example, suppose you run the following instructions:
f1="Larry" f2="[email protected]" f3="415-555-1234"
temp=DBADD("tinydb")
This creates a new record in the database tinydb, with the values Larry,
[email protected], and 415-555-1234 in the fields f1, f2, and f3 respectively. The DBADD
tag, after adding the record to the database, places the value TRUE in column 1, row 1 of temp, and
the value OK in column 1, row 2. It also places the record ID of the record added in column 2, row 1
of temp. temp is the value returned by the DBADD tag.
Note that the DBADD and DBEDIT tags, like other tags in HTML/OS, return a value, which is why you
write temp=DBADD and not just DBADD. The value returned is called a status result. The value abides
by an HTML/OS convention used for all status results (assuming the Overlay tag returns a status
result). A TRUE or a FALSE is placed in column 1, row 1. In column 1, row 2, the word OK is placed,
or in the event the tag encounters a problem, an English-like error message. (Supplemental data is
sometimes placed in other cells of a status result, depending on the particular Overlay tag.) Status
results are described for each Overlay tag, if created, in Appendix D, HTML/OS Tag Reference
Guide.
-139-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
In the eight-line database editor, options to delete and load data are also provided. Clicking Delete
runs the DBDELETE tag. This tag deletes a specified record from the database and returns a status
result.
When a user clicks the Load button, the DBGETREC tag runs. This tag loads data from a record into
variables. You might say it does the opposite of DBEDIT. Whereas DBEDIT loads data from
variables into a record, DBGETREC takes data from a specified record and loads it into variables. For
example, the reverse of the two-line DBADD instruction previously described would be:
temp=DBGETREC("tinydb",record)
record is the record ID of a record you with to load. The DBGETREC tag takes the values in the
fields of the specified record and loads them into variables of the same name. As you may recall
from Chapter 10, HTML Form Processing, once a variable is loaded it appears in the HTML form
automatically. In the eight-line database editor, the DBGETREC instruction is followed with a GOTO
PAGE and no other instructions. That's all that is needed to load the data from the record into the
form on the page. Once the page is reloaded it will display the data from the record since HTML
forms display the values in the variable environment.
Whenever the eight-line database editor page is reloaded, the current set of records in the database
displays, the HTML form displays with the current values of the fields in the database record, and the
user is provided, once again, with options to edit a database record.
Summary
In this chapter you obtained a basic understanding of Web databases. You've seen that working with
databases requires one tag per operation and that all the fields in a record are typically read or
written once.
The eight-line database editor used in this chapter, while serving to explain the Overlay tags DBADD,
DBDELETE, DBEDIT, and DBGETREC, had three obvious limitations:
ƒ While the editor was suited to editing a small number of records, it is not suited to larger
databases. Here we display all the record numbers on the screen. You would not want
to do that when you have many records.
ƒ Only record numbers were displayed on the screen. A good record editor would list
additional information for each record, and to load a record, you would simply click an
item in the list.
ƒ The editor allows the user to edit any record. You do not want to do that if you wish to limit
the records the user is able to edit.
These issues and more are left for Chapter 14, Building Database Editors. There you'll learn how to
build more sophisticated database record editing systems.
Next, in Chapter 12, Building Query Pages, you explore how to retrieve, not one or all of the records
in a database, but a selection of records. You move on to learning the details and variations of how
one queries and extracts information from databases.
Exercises
The following exercises give you a feel for working with the basic DB tags described in this chapter.
You are asked to take the eight-line database editor and modify it in a number of ways. Answers to
-140-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
the exercises here are provided on this book's companion Web site as described in the book's
Preface.
Exercise 1
The eight-line editor used in this chapter does not display messages as records are edited, added, or
deleted. Add messages to the application, so when a database operation is performed, its success
or failure is reported at the top of the page.
Exercise 2
Convert the eight-line editor into a guest book page. To do so, eliminate all buttons on the page
except the Add button. Rename the button to Save To Guest Book. And, when the record is added,
display a Thank-you message at the top of the page.
Exercise 3
TE
AM
FL
Y
Take the guest book you constructed in the previous exercise and add data validation to it so the
user is not able to add a record to the database if a valid e-mail address has not been entered or if
the e-mail address entered already exists. To accomplish this, use the same e-mail validation
methods used in Chapter 10, HTML Form Processing. To see whether a user is already in the
database, use DBFIND.
-141Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 12: Building Query Pages
Overview
Most database searches start with a user entering values in one or more text boxes in an HTML form,
clicking a Find button, and getting a listing of results. In this chapter you learn how to build these
kinds of pages. You start with a page consisting of a single input. Then you explore, in general, how
queries are constructed and how to set up and use group fields—special fields that accept multiword
queries. Later you explore search pages with multiple input boxes, and at the end of this chapter you
learn how to build query pages that search more than one database.
Later, in Chapter 13, Building Database Reports, you learn how the queries are displayed, sorted, or
broken up so search result pages can be printed. Here, you focus on the construction of queries,
using a simple report page to display your search results. The page, /work/report.html,
displays results using a LAYOUT tag, the same tag used and described in Chapter 4, Your First Web
Database Program. The page /work/report.html is as follows:
<html>
<title>Two-Line Database Report</title>
Search Results:<ul>
<table border=1>
<< DISPLAY
LAYOUT(myresults,"<tr>",
"<td>",[1],"</td>",
"<td>",[2],"</td>",
"<td>",[3],"</td>",
"</tr>")
/DISPLAY
>>
</table>
</ul>
</html>
This page displays the first three columns of the variable myresults, the variable we shall load with
search results throughout this chapter. Unless stated otherwise, we shall also assume the database
/work/contacts with the five fields record (the first field of all HTML/OS databases),
contact_company, contact_name, contact_phone, and contact_email.
The Four-Line Query Page
You start by looking at a simple search page consisting of a text box and a Find button. The code for
the page is as follows:
<html>
<title>Four-Line Query Page</title>
-142-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<form method=post ACTION=findit>
Enter Search: <input type=text name=mysearch size=10>
<input type=submit value="Find" ></form>
</html>
<<overlay findit
mysearch=CUTALL(mysearch, '"')
myquery= 'contact_company ~ ' + '"' + mysearch + '"'
myresults=DBFIND("/work/contacts",myquery,1,50,
"contact_company,contact_name, contact_phone")
GOTO "/work/results.html"
>>
When a user enters a value in the HTML form and clicks Find the on-click Overlay findit runs.
There, mysearch, the variable submitted by the form, is stripped of any double-quotes using the
CUTALL tag. Then a special instruction called a query string, is created and saved to the variable
myquery. This special instruction is one of the parameters of the DBFIND tag—telling the tag which
records to retrieve from the database and place in myresults. After DBFIND runs, the user is
directed to the page /work/results, where the variable myresults is displayed. The bold
instructions of this on-click Overlay set up the query string used in the DBFIND tag. The two
instructions convert the input typed into the HTML form by a user into a query string the DBFIND tag
can understand.
Creating query pages is really about placing an HTML form with one or more inputs on a page and
then, after the HTML form is submitted, converting the values entered by the user into a query string.
In the next section you dive into the specifics of how these query strings are composed.
The Query String
When an HTML form is submitted a set of variables entered by a user is available for your use. To
convert these variables to a query string, you first need to know how query strings are formatted.
In HTML/OS these query strings, also known as Boolean expressions, are built from one or more
statements, the smallest of which is composed of three parts: a field name, an operator, and a value.
The statement, contact_name ~ "Bill", is such an example. The field name is contact_name.
The operator is a tilde (~), which means begins with. The value is Bill.
Multiple statements can be combined with the words AND or OR. You use AND to find the overlap
between two search results. You use OR to combine two search results. You can also precede these
statements with the word NOT, indicating all records other than the search results. For example, let's
say you have the statement RECORD > 5 and another statement RECORD < 10. Writing (RECORD
> 5 AND RECORD < 10) gives you the overlap of these two statements—all numbers greater than
5 but less than 10. Writing (RECORD > 5 OR RECORD < 10) gives you all possible records.
Writing NOT(RECORD > 5) gives you all records less than or equal to 5. It's the same as writing
RECORD <= 5.
In general, it is a good idea to surround pairs of statements combined with the AND or OR words with
parentheses, so the whole pair of statements can be thought of as a single statement and combined
with others, using the NOT, AND, or OR tags. Parentheses eliminate ambiguity. When DBFIND
encounters a set of parentheses, it calculates the searches in the inner parentheses first. As an
example, consider the following query string:
-143-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
(RECORD > 5 AND RECORD < 10) OR (RECORD > 15)
The use of parentheses tells DBFIND how to combine the individual statements. In this query string
DBFIND will take the first set of statements in parenthesis and combine it with the last statement,
RECORD > 15. The result is all records between 5 and 10 plus records greater than 15.
Using Quote Marks
The query string contact_name ~ "Bill" has quote marks around the value Bill. The query
string RECORD > 5 does not. So when should you use quote marks? The rule is that you can always
use quote marks around a value; but they are optional when the value is a number The query string
RECORD > 5 can be written as RECORD > "5" as well.
Quote marks are needed since values in query strings often contain spaces. Without the quote
marks, queries will fail when they encounter values with spaces, because DBFIND will get confused
as to where values begin and end.
In the four-line query page, you used the CUTALL tag to extract the quote mark from the query. Then
a query was composed with double quotes around the value. This was done so, regardless of the
user input, the DBFIND tag would properly interpret the query string.
The best advice is to use quote marks whenever you're in doubt, because as long as you use them
correctly, they never cause problems—and in fact often avoid subtle problems that otherwise may
creep into your code.
Field Operators
When building query strings, each value is compared against a field using a special operator
consisting of one or two characters. This operator will depend on the kind of field being compared
against. The Integer (INT), Floating Point (FLT), Currency (DOL), and Date and Time (DAT) fields
use the same sets of operators. String (STR) uses a different set of operators. Note also that
comparisons on strings are case insensitive. For example, Bill will match bill or biLL in a string
search. The operators available to each are as follows:
Operators for Integers, Floating Points, Currencies, and Date and Time
ƒ
<
Less than
<=
Less than or equal
>
Greater than
>=
Great than or equal
<>
Not equal (Same as !=)
=
Equal
Operators for Strings
ƒ
=
Equal
~
Begins
-144-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
with
~~
Contains
These operators give you flexibility in setting up queries on a database. The two bold lines of the
four-line query page described in the "Four-Line Query Page" section at the beginning of this chapter
search only the contact_name field. But what if you want the search to include other fields? There
are two ways to accomplish this. First, you can set up a group field that includes the fields you want
to search and have your query string search that field instead of contact_name. Group fields are
described in the next section of this chapter. Or you can compose a more complex query. For
example, the two bold lines of the four-line query page could be substituted with the following lines:
mysearch=CUTALL(mysearch, '"')
myquery= '(contact_company ~ ' + '"' + mysearch + '") OR' +
'(contact_name
~ ' + '"' + mysearch + '") OR' +
'(contact_phone ~ ' + '"' + mysearch + '") OR' +
'(contact_email ~ ' + '"' + mysearch + '")
Note the mixed use of quote marks. Since you want to place a double quote before and after the
value in mysearch, you need to paste these together carefully, keeping in mind double quotes need
to be surrounded by single quotes.
Also note that the use of the tilde (~) operator. This operator, which means begins with, finds all
records where the specified field begins with the value specified. The operator tilde-tilde (~~), which
means contained in, finds all records where the specified field contains the value specified. See the
accompanying "Using the Tilde-Tilde Operator" note.
Using the Tilde-Tilde Operator—Searches using the tilde-tilde (~~) operator are slower than
other searches, because DBFIND needs to locate a specific pattern within a field and is not able to
take advantage of its internal cross-referencing (indices). While most queries take about one-fiftieth
of a second, the tilde-tilde search can take much longer. In databases containing more than 10,000
records, for example, the search can take many seconds. Unless your database is small, it is better
to set up group fields than use a tilde-tilde search.
Using Group Fields
In the previous section query strings contained standard fields defined as Integers, Floating Points,
Currencies, Date and Time, and Strings. Query strings can also search group fields. You set up
group fields in dbConsole. They are virtual fields composed of one or more string (STR) fields. Like
searches on strings, searches of group fields are case insensitive.
Let's set up a group field for the contact database discussed at the start of this chapter and in
Chapter 4, Your First Web Database Program. To do this, access your HTML/OS desktop and click
the dbConsole icon.
1. Then click Open and select the database /mycontacts.
2. Click Change.
3. Click Add Group to add a group field to the database. You should see a screen
like that shown in Figure 12.1.
-145-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 12.1: Use the Change Section in dbConsole to Add Group Fields to Databases.
Suppose you want searches to span the fields contact_company, contact_name,
contact_email, and contact_phone.
1. Enter the field name contact_key in the Group Name text box.
2. Select a field from the pull-down menu, and click Add To Group repeatedly until
the Group Members list contains all of the fields you want in the group.
3. Click Save when done.
4. Then click Complete Changes.
Now /work/contacts has a group field called contact_key.
The code to search a group field is like the code used to search other kinds of fields except that the
operators that work with group fields differ, because group fields are word-indexed, meaning each
field is searched according to the words in the field, not just the field as a whole.
If a user enters multiple words into a search, DBFIND, when searching group fields, checks those
words against all the words in fields that are members of the group field. When searching string (STR)
fields, DBFIND looks at the entire field as one value. For example, a search for Blue Box searches
fields with the word Blue followed with a space, followed by Box. If the search is done on a group
field, the words Blue and Box are considered two different searches. Each word is matched against
the words in the fields belonging to the group field.
Searches of group fields are done in one of two ways. On one hand, you may want any of the words
in the search to match any of the words in the fields belonging to the member fields. On the other
hand, you may want all of the words in the search to match at least one word in the fields belonging
to the member fields. The first kind of search is called an OR search. The second is called an AND
search. In computer science the AND is sometimes represented by the ampersand (&) character and
the OR is sometimes represented by the pipe (|) character. Following this convention, the operators
used with group fields are followed by one of these two characters. They are as follows:
Operators for Group Fields
ƒ
=|
Equal (on any word specified, same as =)
~|
Begins with (on any word in any field, same as ~)
=&
Equal (on all words specified)
~&
Begins with (on all words specified)
-146-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
When setting up general searches on a Web site tilde-OR and tilde-AND searches are used most
often. Tilde-OR searches are typically used for smaller databases and tilde-AND searches for larger
databases. You will need to experiment with your database to see which works better for you.
To improve the search used in the four-line query page, you search a group field instead of the
contact_company field. Instead of writing the following
myquery= 'contact_company ~ ' + '"' + mysearch + '"'
you write
myquery= 'contact_key ~| ' + '"' + mysearch + '"'
Now users entering multiple words will get matches to all of their words. And all the fields defined for
contact_key will be searched instead of only the contact_name field. The new and improved
four-line query page, with changes in bold, is as follows:
<html>
<title>Four-Line Query Page</title>
<form method=post ACTION=findit>
Enter Search: <input type=text name=mysearch size=10>
<input type=submit value="Find" ></form>
</html>
<<overlay findit
mysearch=CUTALL(mysearch, '"')
myquery= 'contact_key ~| ' + '"' + mysearch + '"'
myresults=DBFIND("/work/contacts",myquery,1,50,
"contact_company,contact_name, contact_phone")
GOTO "/work/results.html"
>>
Working with Multiple Inputs
In the four-line query page the user is presented with only a single input box. On many query pages,
however, you will want to present the user with multiple query selections. Searches of automotive,
real-estate, and dating databases are a few cases that require multiple search criteria.
Like all query pages, query pages with multiple user selections are about converting the variables set
in the HTML form presented to the user into a query string. It's just that the query is more complex.
In general, HTML forms contain the following kinds of query requests:
ƒ Free text— Inputs asking the user to type in one or more words
ƒ Selection— Inputs asking the user select from a list of possible choices
ƒ Check all that apply— Inputs asking the user to place a check mark in zero or more
boxes
ƒ Query pairs— Inputs asking the user to specify a pair of inputs such as a begin and end
date
-147-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
For example, an automotive database may contain a search for a car make, model, year, and
whether the user wants a new or used car. The car make, model and year can be selected from a list
of possible values. Multiple-selection queries involving pull-down menus or radio buttons are used. A
real-estate property search may query the user for a price range, a geographic area, and ask that
users check off a list of desired amenities such as skylights, fireplace, and a Jacuzzi tub. The price
range is a kind of query pair requiring the use of two pull-down menus or text boxes, one for
minimum price and the other for maximum price. The geographic area may be a Zip code typed into
a box. That would be a free text query. The desired amenities would be a check-all-that-apply type of
query. Checkboxes are used there.
To build the HTML form that requests the queries, you use the components discussed in Chapter 10,
HTML Forms Processing. When the user submits the HTML form, you validate the data; also
described in Chapter 10. If the query is valid, you build your query string. This last step is what we
describe here.
In general, when building a complex query, you start by building each individual query. Then, when
you are done, you paste them together. Consider, for example, a hypothetical Web page for a
singles dating club. The HTML form used on the page is as follows:
<form method=post ACTION=matefinder>
Find a Mate:<ul>
I'm looking for a <select name=gender><option
value=M>Male<option value=F>Female</select><ul>
Age must be between <input type=text name=min_age size=5>
and <input type=text name=max_age size=5><br>
He/she may be
<input type=checkbox name=height[1,1] value=short>short,
<input type=checkbox name=height[1,2] value=ave>average,
<input type=checkbox name=height[1,3] value=tall>tall<br>
</ul>
<input type=submit value="List Eligible Mates">
</form>
When the HTML form is submitted the four variables; gender, min_age, max_age,
height[1,1], height[1,2], and height[1,3] are set. To create a query string, you compose
three statements. You build the first with the variable gender. Code the second query statement by
using min_age and max_age. In the last, you use the values in the variable height. Once this is
done, paste together the three statements with ANDs to create the final query string.
Let's assume the dating database is called /work/singles and contains the fields s_gender,
s_age, and s_height. Let's also assume the s_gender fields always contains the letter M or F;
s_age always contains an integer, and s_height always contains the word tall, ave, or short.
The first query statement is written as follows:
q1='s_gender = "' + gender + '"'
The second is written as follows:
q2='(s_age >= ' + min_age + ' AND s_age <= ' + max_age + ')'
-148-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The last query statement needs to build a statement from all of the height selections. To do this, we
can use these IF-THEN statements as follows:
q3=""
IF height[1,1] != "" THEN q3=q3 + OR
s_height="'+height[1,1]+'"' /IF
IF height[1,2] != "" THEN q3=q3 + 'OR s_height="'+height[1,2]
+'"' /IF
IF height[1,3] != "" THEN q3=q3 + 'OR s_height="'+height[1,3]
+'"' /IF
IF q3 != "" THEN q3 = " AND " + REPLACE(q3,"(OR","(") /IF
Combine these with the following instruction:
myquery="(" + q1 + " AND (" + q2 + "AND" + q3 + "))"
The result is the query string needed by DBFIND to perform the search. Note that this query was built
by first creating the smaller queries q1, q2, and q3. They were pasted together at the end. The
composition of q1 and q2 is straightforward. The composition of q3 is tricky. There, the statement
q3=q3 + "OR field ="+value+")" was used to build q3 by successively combining q3 with
new statements every time a nonempty selection was found. Note that the technique can be used
whether you have three items that need to be checked off, or ten items. Here we used three IFTHEN statements; but alternatively, you could use a loop (as an exercise at the end of this chapter
demonstrates). In any event, the problem with the statements is that q3 starts out an empty string,
so you get an extra OR in the first statement created. For example, if a user were to select short
and average the value of q3 immediately following the third IF-THEN statement would be
((OR s_height="short")AND s_height="average")
To correct the extra AND, you follow the IF-THEN statement with an instruction that replaces the four
character sequence (AND with the ( character. That gives you the correct q3 query statement.
In general, using the techniques discussed here, you can build complex queries that span many
fields. The field names and selections may change, but the concepts are the same.
On-the-Fly Joins
The query pages thus far have searched a single database. Sometimes however, querying a single
database is not enough. Multidatabase searches are required when multiple results need to be
simultaneously extracted from more than one database. A real-estate database report that lists the
names of listing agents for each property, although the names of the listing agents are stored in a
database different than the property database, is such an example. This situation requires on-the-fly
joins of multiple databases. The concept of on-the-fly joins is explained in this section.
A multidatabase search should not be confused with the concept of drill-down. A drill-down search
occurs when an initial search yields a result that should lead to further information (by the user
clicking hyperlinks to that information, for example), some of which may be stored in other databases.
A listing of real-estate properties with hypertext links of sales people's names is an example.
Although multiple databases are accessed, they are not accessed at one time. On-the-fly joins are
not needed.
-149-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Most often on-the-fly joins are needed when you wish to augment the search results from one
database with data stored in another table. First you perform a search on a database. Then you
combine the search result obtained from the first database with a search of a second database.
In HTML/OS you perform searches across multiple databases by searching each database, one after
the other, taking the result of one search and feeding it into the next search, until all databases have
been searched. The first search uses the DBFIND tag. Successive searches use the DBFINDJ tag.
For example, suppose you have a database called /work/orders containing, among others, the
fields order_company, order_amount, and order_date. If you want all the orders between
two dates, date1 and date2, you need to search the /work/orders database. The query string
and search might look as follows:
myquery='order_date <= "' + date1 + '" AND order_date > "' +
date2 + '"'
myresults=DBFIND("/work/orders",myquery,
1,50,"order_company,order_amount,order_date")
The DBFIND tag, using this query string, fills myresults with fields from /work/orders. If you
recall from Chapter 4, Your First Web Database Program, DBFIND has five parameters. The first is
the database name. The second is the query string. The third and fourth parameters specify the
range of search results to return, and the last parameter is the list of fields in the specified database
to return.
The /work/orders database, however, doesn't include contact names and phone numbers. That
data is stored in /work/contacts. To extend the search with data from the /work/contacts
database you use the DBFINDJ tag. See Appendix D, HTML/OS Tag Reference Guide for further
information. The J stands for Join. The following code is added after the instruction containing the
DBFIND tag:
myresults=DBFINDJ("/work/contacts","",1,50,
"contact_name,contact_phone",myresults,1,"contact_company")
The DBFINDJ tag has the same parameters as DBFIND plus three additional: the search result of
the prior search, the column number in the search result that matches a field in this next database
and the name of that field. These three parameters define a join of the first search result table
(myresults) with a database (/work/contacts). In general, when DBFINDJ does its search, it
starts with a table of results from a prior search and builds a new table using that, extending that
table horizontally. The variable myresults starts out with the three columns. The table returned by
DBFINDJ is five columns, the last two being the contact name and phone. An as example, suppose
the result of DBFIND was as follows:
ABC Rentals
1550.00
Babott Insurance 2370.50
04/02/2001
07/20/2001
ABC Rentals
3025.75
07/28/2001
Delphi Leasing
1900.00
12/17/2001
The DBFINDJ tag would match each row of this result with records in /work/contacts having the
same company name. If it finds a match, a contact name and phone are added to columns four and
five. If no match is found the columns are left blank. If duplicate matches are found, the row is
duplicated, and columns 4 and 5 are filled with contact names and phone numbers.
-150-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Let's say, for example, /work/contacts contained the following five records:
Record #1: ABC Rentals
John Smith 415-555-1234 jabcrental.com
Record #2: Babott Insur Max Marcs
619-515-1234 [email protected]
Record #3: Bella Music
Bella Brown 714-555-1212 [email protected]
Record #4: Brass Tax
Jacki Prince 212-555-1234 [email protected]
Record #5: Delphi Leasing Janet Adams 303-555-1234 [email protected]
The join of this database with myresults, the three-column value returned by DBFIND, yield the
following five-column result:
ABC Rentals
Babott Insur
1550.00
2370.50
04/02/2001
07/20/2001
John Smith
Max Marcs
ABC Rentals
3025.75
07/28/2001
John Smith
Delphi Leasing
1900.00
12/17/2001
Janet Adams
415-555-1234
619-555-1234
415-555-1234
303-555-1234
Summary
AM
FL
Y
The DBFIND search yielded a three-column result. This table was extended to five columns with a
DBFINDJ tag, drawing two fields from a second database. If more information needs to be added to
the result, you can use additional DBFINDJ tags.
TE
In this chapter you obtained a basic understanding of how to build query pages. You learned that all
search pages, whether simple or complex, are about taking the variables submitted in an HTML form
and using them to build Boolean query strings.
Next, in Chapter 13, Building Database Reports, you learn what to do with these search results and
how to format them on a Web page. The next chapter, along with this one, gives you the ability to
build a wide variety of database search and result pages—an important component of advanced
Web sites and Web-based applications.
Exercises
In the following exercises you extend and build additional query pages using what you learned in this
chapter. Answers to these exercises are provided on this book's companion Web site as described in
the book's Preface.
Exercise 1
Expand the four-line query page to accept both equals and begins with searches. Do this by adding
a radio button to the page allowing users to select the kind of search they want. Then add the
necessary IF-THEN statement to the on-click Overlay to set the correct operator to use in the query
string.
Exercise 2
In the "Working with Multiple Inputs" section of this chapter, you built an on-click Overlay that
processed a complex query for searching a singles dating database. In the on-click Overlay you
used a series of IF-THEN statements to build query q3. Replace this section of code with code that
-151Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
uses a FOR loop across the variable height so additional selections may be added to the HTML
form without having to rewrite code in the on-click Overlay.
Exercise 3
Group searches are typically used to search words within a description; but they can also be used to
search lists contained within a single field. As an example, a database for real-estate properties often
needs a place to store property attributes. In real estate the possible attributes of a house may
exceed 50. Instead of adding a field for each attribute, it makes more sense to have a single
Attributes field containing a list of attribute codes. Using this approach, build a sample real-estate
database and a query page that allows a user to select one or more property attributes (high ceilings,
dishwasher, dryer, fireplace, ocean view, etc.)
Exercise 4
The on-the-fly joins discussed in this chapter did not contain a query in the DBFINDJ, because they
were used to extend the results of an existing search result. In some situations however a search
string is needed as input to theDBFINDJ tag. For example, consider the case of an orders
database and an itemsordered database. Every time an order is placed a record is added to the
orders database, which may contain a customer type (wholesale or retail, for example). At the
same time records are added to the itemsordered database; one new record for each item
ordered. The record contains a product ID and selling price. Knowing this, build a sample twodatabase system and a query page that allows a user to query the database for wholesale or retail
orders above or below specific price amounts.
-152-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 13: Building Database Reports
Overview
A database report is worth a thousand fields. Building highly functional, easy-to-read database
reports is time well spent. What's more, custom Web-based reports offer more functionality than
conventional database reports. Web-based reports can be dynamic. Click a column and the report
can redraw itself—sorted by that column. Click an entry and the report can give you detail on an item.
In this chapter you learn how to build these kinds of reports.
In the last chapter, you built query pages that converted one or more user inputs into a query string
that was then fed into a DBFIND tag to produce a search result. In this chapter, you look at the
variety of ways the search result produced by DBFIND can be customized and presented on the
page.
You start with a simple seven-line database report and add features to it, one by one. Changes are
described and explained. By the end of the chapter, you'll know how to build multipage, printable,
sortable, multipurpose reports.
Seven-Line Database Report
In Chapter 12, Building Query Pages, you built query strings used by DBFIND tags. Users were
directed to the page /work/report.html where search results were displayed. The page was a
two-line database report that took the variable myresults and displayed it using a LAYOUT tag.
In this chapter you wish to do more on the report page. Instead of assuming users access the report
page after composing a query and performing a search, here you assume the user has composed a
query string but has not done the search. You let the report page do the search. The page has seven
HTML/OS instructions, which is why it is called a seven-line database report. It uses a DBFIND tag to
search the database, a FOR loop, three DISPLAY tags for displaying three columns from myresults,
and two more DISPLAY tags for displaying beginning and ending HTML table row tags (<tr> and
</tr>). You place the report within an HTML table to ensure columns line up. The page is as
follows:
<html>
<title>Seven-line Database Report </title>
<< myresults=DBFIND("/work/contacts",myquery,1,100,
"contact_company,contact_name,contact_phone")
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088 colspan=4 align=center>
<b>My Report</b></td></tr>
<tr>
<td>Column One</td>
<td>Column Two</td>
<td>Column Three</td>
</tr>
-153-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<<
FOR NAME=myresults ROWNAME=myrow DO
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
</table>
</html>
Unlike the two-line report described in Chapter 12, Building Query Pages, this report uses a FOR
loop to display rows from myresults, one at a time. The LAYOUT tag is switched to a FOR loop to
gain control over the lines displayed. The added control is not necessary at this point; but later, as
you add sophisticated features to the report, it will be. As an example, suppose you want to change
the seven-line report to one that displays every other row in the report in light green to emulate green
bar computer paper. To do so you add an IF-THEN statement inside the FOR loop that toggles the
contents of a variable back and forth between lightgreen and white. The code is shown here
with changes marked in bold.
<html>
<title>Seven-line Green Bar Report </title>
<< myresults=DBFIND("/work/contacts",myquery,1,100,
"contact_company,contact_name,contact_phone")
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088 colspan=3 align=center>
<b>My Report</b></td></tr>
<tr>
<td>Column One</td>
<td>Column Two</td>
<td>Column Three</td>
</tr>
<<
FOR NAME=myresults ROWNAME=myrow DO
IF x="white" THEN x="lightgreen" ELSE x="white" /IF
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td><font color="+x+">"+myrow[1]+"</font></td>"
/DISPLAY
DISPLAY "<td><font color="+x+">"+myrow[2]+"</font></td>"
/DISPLAY
-154-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
DISPLAY "<td><font color="+x+">"+myrow[3]+"</font></td>"
/DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
</table>
</html>
Using the FOR loop gives you the flexibility to perform different tasks as different rows are displayed.
This is not possible with the LAYOUT tag. In this green bar report, an HTML font tag is placed
around each cell of myrow. The color in the font tag is set to the variable x and given values that
toggle between lightgreen and white.
Adding Page Up and Page Down
The seven-line database report of the previous section displayed the variable myresults. The
DBFIND tag returned the first 100 rows. In some circumstances however, you will want to display
more rows. Of course, you can increase the parameter used in DBFIND; but that is not always
enough. You may not want the user to scroll up and down the page. You may prefer providing links
to page up and down through the results.
At least two modifications to the prior report are needed for this. First, the two parameters in DBFIND
that define the range of search results to return need to be changed to variables. Second, page up
and page down links need to be added to the page. When a user clicks a page up or page down link,
you need Overlay tags to modify the variables used in the DBFIND tag that define the range of
search results to return.
A third, optional modification is to display for the user the range of search results on the page and
the total number of search results available. That data can be extracted from the TAGRESULTS
variable that's set when the DBFIND tag runs. We do that here as well. Here's a page with these
modifications.
<html>
<title>Database Report With Page Up and Page Down</title>
<< pagesize=20
IF ISINTEGER(i)="FALSE" THEN i=1 /IF
j=i+pagesize-1
myresults=DBFIND("/work/contacts",myquery,i,j,
"contact_company,contact_name,contact_phone")
j=TAGRESULTS[4,1]
t=TAGRESULTS[5,1]
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088><b>My Report</b></td>
<td align=right colspan=2>Results <<i>> to <<j>> of <<t>></td></tr>
<tr><td colspan=2><a href=pgup>Page Up</a></td>
-155-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<td align=right><a href=pgdn>Page Down</a></td></tr>
<tr>
<td>Column One</td>
<td>Column Two</td>
<td>Column Three</td>
</tr>
<<
FOR NAME=myresults ROWNAME=myrow DO
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
</table>
</html>
<<overlay pgdn
i=i+pagesize IF i > t THEN i=i-pagesize /IF GOTO PAGE
>>
<<overlay pgup
i=i-pagesize IF i < 1 THEN i=1 /IF GOTO PAGE
>>
Changes are shown in bold. In the first Overlay you start by setting the pagesize variable. Then,
you set the two parameters in DBFIND that set the range of results to return to i and j. To ensure i
and j are initialized, you place an IF-THEN statement prior to the search. Just below the DBFIND
instruction, values for j and t are assigned using cells in the TAGRESULTS variable filled by DBFIND.
Note that although j is set before running the DBFIND tag, it is also set after the DBFIND tag runs.
This is necessary since j does not always equal i+pagesize—at least not on the last set of search
results a user might view. The contents of TAGRESULTS are described in the description of DBFIND
in Appendix D, HTML/OS Tag Reference Guide.
Below this Overlay, an HTML table filled with search results is displayed. The first row displays the
range of results on the page. The row below this contains Page Up and Page Down links. The first
runs the pgup on-click Overlay and the other runs the pgdn on-click Overlay. When a user clicks
these links, the i and j variables are adjusted causing a page up and page down when the page is
redisplayed.
Building Printable Reports
Printable reports are formatted so they split nicely across different pages of a printout. Printable
reports also have page numbers and headers on each page. In many ways, printable report pages
are constructed like the report pages with page up and page down discussed in the previous section,
-156-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
except that the pages in the report are displayed all at once on the same Web page and split only
when printed.
The challenge when building printable reports is fitting the lines to the printed page. Since the
number of lines that fit on a page can vary from browser to browser, and since the HTML language
does not include a page-break tag, you need to improvise. The trick comes from knowing that
browsers, when they print Web pages, try to fit HTML tables on a single printed page. If you split the
report into distinct HTML tables, each large enough to fill the page, but not too large so that it carries
over to the next printed page, then you will get one HTML table on each printed page. You use that
fact when building printable reports.
To convert the seven-line database report to a printable format, you need to make two modifications.
First, groups of rows in the search result need to be split among different HTML tables. Second, at
the top of each of these HTML tables you need to add a title line with a page number and a total
number of pages. The page to do this is as follows:
<html>
<title>Printable Database Report</title>
<< myresults=DBFIND("/work/contacts",myquery,1,1000,
"contact_company,contact_name,contact_phone")
t=TAGRESULTS[5,1]
rowsperpage=40
page_no=1
page_last=ROUNDUP(t/rowsperpage)
i=1
FOR NAME=myresults ROWNAME=myrow DO
IF i=1 THEN
IF i+rowsperpage > t THEN j=t ELSE j=i+rowsperpage /IF
DISPLAY
'<table border=1 cellspacing=0 cellpadding=2>'+
'<tr><td align=right colspan=3>'+
'Page '+page_no+' of '+page_last+'</td></tr>'+
'<tr>'+
'<td>Column One</td>'+
'<td>Column Two</td>'+
'<td>Column Three</td>'+
'</tr>'
/DISPLAY
/IF
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
-157-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
DISPLAY "</tr>" /DISPLAY
IF i=rowsperpage THEN
DISPLAY "</table>" /DISPLAY
i=0
page_no=page_no+1
/IF
i=i+1
/FOR
>>
</table>
</html>
Changes are shown in bold. First you set a rowsperpage variable to 40. This value should be
adjusted for your printer. Then page_no is initialized and you calculate total_pages by using the
ROUNDUP tag. This tag rounds up fractions to their next largest integer. The remaining modifications
are made in the FOR loop. At the top of each page, when i is equal to 1, a header is displayed. The
header opens an HTML table and then displays an HTML table row that displays page_no and
page_last, the current and last page (or table, since one HTML table is printed per page). When i
is equal to the rowsperpage, the end HTML table tag is displayed. After every row the variable i is
incremented, but reset to 1 at the end of each HTML table. This gives you the desired effect—a Web
page with multiple HTML tables, each with rowsperpage rows (except on the last page) and each
with its own header. When the page is printed, each HTML table will appear on a different printed
page giving you the affect you want.
Building Reports with Sortable Columns
Reports with sortable columns give users maximum flexibility in how they view their reports. Ideally,
you want users to click the top of any column to redisplay their report sorted by the column they
clicked. If they click again, the report should be redisplayed again, this time in a reverse sort.
A subtlety to this is that clicking a column doesn't always mean you should change the order in which
the column is sorted. When changing the sort column, you want to sort by the last way that column
was sorted.
Adding sortable columns to a report requires three modifications. First you add links to the top of
each column for users to click; then you add a variable that stores how each column is sorted along
with the code needed to change the sort when a user clicks the links, and last, you replace the
DBFIND tag with DBFINDSORT. The DBFINDSORT tag is similar to DBFIND except it has two
additional parameters for returning search results sorted by a specific column and sort order (reverse
or default). The code for a database report page with sortable columns is as follows:
<html>
<title>Database Report with Sortable Columns</title>
<<
IF ISINTEGER(resort_col)="FALSE" THEN
sort_fields=ROW("contact_company","contact_name","contact_phone"
-158-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
)
sort_types=ROW("Y","Y","Y")
sort_col=1
sort_col_last=1
/IF
IF resort_col != "ERROR" THEN
sort_col=resort_col
resort_col="ERROR"
IF sort_col=sort_col_last THEN
IF sort_types[sort_col]="Y" THEN
sort_types[sort_col]="N"
ELSE sort_types[sort_col]="y"
/IF
/IF
/IF
sort_col_last=sort_col
myresults=DBFINDSORT("/work/contacts",myquery,1,100,
sort_fields[sort_col],sort_types[sort_col],
"contact_company,contact_name,contact_phone")
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088 colspan=4 align=center>
<b>My Report</b></td></tr>
<tr>
<td><a href=<<page>> NAME=resort_col value=1>Column
One</a></td>
<td><a href=<<page>> NAME=resort_col value=2>Column
Two</a></td>
<td><a href=<<page>> NAME=resort_col value=3>Column
Three</a></td>
</tr>
<<
FOR NAME=myresults ROWNAME=myrow DO
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
-159-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</table>
</html>
The first IF-THEN statement in the page initializes variables. The code inside it runs only the first
time the page runs—when sort_col has not yet been initialized. It initializes sort_fields, a
three-column by one-row variable containing the field names displayed in each column. Then it sets
sort_types, a three-column by one-row variable that stores the sort order of each column. After
that sort_col, a variable containing the current sorted column, and sort_col_last, a value for
the last sorted column, are set to 1.
The next IF-THEN statement responds to clicks of a column link. It makes sure sort_col is
modified only if a column is clicked, in other words, only when resort_col has an integer value.
Once a click has been detected, the value of sort_col is set to resort_col and resort_col is
changed back to a noninteger value. Inside this IF-THEN statement, the cell in sort_types for the
clicked column number is toggled if the column that was clicked changes. Otherwise the cells in
sort_types are left alone.
Finally, in the HTML table header, the column headers are changed to links that set the
resort_col variable when clicked and return to redisplay the page.
Linking Reports to Detail Pages
Database reports on the Web can easily be linked to additional detail information or related reports.
You do this by changing entries in the report to hypertext links. When a user clicks an item, a
variable is set, and the user is sent to a detail page or another report page.
Suppose for example you have a report that displays the sales orders for the day. Each line of the
report includes an order ID, an amount for the order, and the salesperson responsible for the order.
The report can be extended. The order ID can be converted into a link that, when clicked, directs the
user to a detail page for the order. Or the name of the salesperson can be converted into a link that,
when clicked, directs the user to a page that displays the orders over the last month for that
salesperson.
In general any item in a report can be converted into a link—providing an additional user selection.
Click the item and more detail is provided. The kind of detail provided is up to you.
In the seven-line database report, column 1 contains the name of a company. As an example, you
can convert the company name into a link. When a user clicks the link, the user is directed to a page
that accesses a company database and displays information on the company. To do this, take the
following line:
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
and replace it with:
DISPLAY '<td><a href=detail.html '+
'NAME=cname value="'+myrow[1]+'">' +
myrow[1]+"</a></td>"
/DISPLAY
This converts the entries in the first row into hypertext links. The other two columns of links do not
change.
-160-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
You write the detail page (assuming a company database with the fields company_name,
company_country, company_state, and company_startdate) as follows:
<HTML>
<title>Company Detail Page></title>
<< temp=DBGET('companydb",'company_name="'cname+'"',1)>>
Company Detail:<ul>
Name: <<company_name>><br>
Country: <<company_country>><br>
State: <<company_state>><br>
Start Date: <<company_startdate>><br>
</ul>
</html>
Summary
AM
FL
Y
In this chapter you obtained a basic understanding of how to build database report pages. The green
bar report showed you that reports can be displayed with horizontal guides to make the report easier
to read. You learned that despite the lack of printing controls in the HTML tag language, printable
reports are possible. You also learned how to build reports with sortable columns—a fantastic way to
empower users of your reports. Finally, at the end of the chapter you learned that you can link
database reports to other reports and detail information—giving you an endless number of ways to
define database reports.
TE
Together with Chapter 12, Building Query Pages, this chapter gives you the ability to build complete
reporting systems as well as those parts of Web-based systems requiring you to select a set of
records from a database and present them to the user.
The concepts learned in these two chapters apply to more than reports. Document databases are
accessed via lists of items selected from a database. Although not reports, query pages and report
pages are needed. Indeed, these two chapters give you the background necessary to build a wide
variety of systems involving database record selection and presentation.
Next, in Chapter 14, Building Database Editors, you learn how to build pages for editing database
records. That chapter moves you beyond the eight-line database record editor discussed in Chapter
11, The Web Database, and using what you learned in the these last two chapters, gives you
powerful solutions useful in advanced Web development.
Exercises
In the exercises here you combine and extend the database report features discussed in this chapter.
Answers to these exercises are provided on this book's companion Web site as described in the
book's Preface.
Exercise 1
In this chapter a page up and page down report was described. Alter the report so the page up
option is not displayed on the first page and the page down option is not displayed on the last page
of the report.
-161Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 2
The printable report used in this chapter used a hard-coded value for the number of rows per page.
Make the page size adjustable using two hypertext links. When the links are clicked, have the rows
per page setting increase and decrease respectively. Store the rows per page setting in a cookie and
read it at the top of the page—giving you a printable report with a user-adjustable page size.
Exercise 3
Modify the report page with page up and page down so it uses page numbers rather than page up
and page down links. To do this display links for each page in the report, placing in each a namevalue pair that sets the value of the variable i (the first row of that page) when clicked.
Exercise 4
Build a database report page that has both sortable columns and a page up and page down feature.
Exercise 5
Build a report that also serves as a query system. This can be done with databases containing
repeated data. For example, consider a product-order database that stores product names,
salespeople assigned to the order, a product category, and an order date. Clicking a salesperson
should limit the report to that salesperson. Clicking a product name should limit the report to that
product. Clicking a category should limit the report to products in that category. At the top of the
report, provide a query box that limits the report to a specific date range and, when clicked, also
resets the report.
-162-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 14: Building Database Editors
Overview
In Chapter 11, The Web Database you first encountered a database editor. The eight-line database
editor allowed users and staff to add, edit, or delete records in a database. The simple editor
however, was suited only to situations involving a single user and a small database. In this chapter
you build general-purpose database editors worthy of the most advanced Web sites. The database
editors you learn to build here can be adapted to a wide variety of development situations including
the construction of user and membership databases, personal settings databases, real-estate,
automotive, and support databases, knowledge bases, image and video databases, accounting and
finance systems, and manufacturing and inventory databases. Indeed, database editors are useful
everywhere a database needs to be edited.
Technically, the database editors discussed in this chapter resemble those discussed in Chapter 8,
Building Text Editors, only here data is stored in database structures rather than text files, and
instead of editing a single document or element, you edit one or more elements. The other difference
is that here, when you search for an item to edit, you search a database. In text editors, when you
search for an item to edit, you search for a file in a directory.
In this chapter you build all of the components that make up a database-record editing system. Your
starting point is a fifteen-line editor. You add features to it, build a Find page that works along with it,
and at the end of the chapter, you learn how to expand the editor to work with multiple related
databases.
Accessing Your Database Editor
Most often your database editor will be password protected. To access it you need a login page. The
login page can be identical to the database login page discussed in the section "Using A Database"
in Chapter 9, Building Login Pages. No need to reinvent the wheel, you can use that one or any
other login page discussed in that chapter. Just make sure you know the variable that contains the
user's login ID. In that one it's lp_login.
If the database you wish to edit includes records that can only be edited by specific users, make sure
the login ID is placed in some field of each record. If your restrictions are more general, based on
groups of users for example, you will want to place a group code in each record and find the group
code of the user who logged in. Most often the user's group code is stored in the login database. In
these cases, when the user logs in, you retrieve the user's group code.
Once a user has logged in, you direct the user to a menu of options. The page might look as follows:
<html>
<title>Staff Menu (dbmenu.html)</title>
Staff Menu - For Authorized Personnel Only<ul>
<li> <A HREF=editdb.html>Edit My Database </a>
<li> <A HREF=opts.html>Other Options</a>
<li> <A HREF=opts2.html>Still Other Options </a>
-163-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</ul>
</html>
Nothing much here. Just a few hypertext links. But then, that's all you need. The first hypertext link
goes to editdb.html, your database-record editor.
The Fifteen-Line Database Editor
Consider the fifteen-line database editor shown in Figure 14.1. We assume a database with six fields:
record, record_user, record_title, f1, f2, and f3. The record_user field is used to
store the login ID of the owner of the record. The record_title field stores the record title. And
the other fields can store whatever you want. When the database editor is first accessed, an empty
record is displayed. Save, Copy, Delete, and Find buttons are provided. The buttons are fully
functional except Find, which is a link to a page you build later in this chapter.
Figure 14.1: The Fifteen-Line Database Editor Works with User-Assigned Records.
This editor blocks writes to records unless they belong to the user. This is an important consideration
when building database editors, since databases often store information for multiple parties requiring
access to only their data. The code for this editor is as follows:
<html>
<title>Fifteen-Line Database Editor (editdb.html)</title>
<font size=5><b>Fifteen-Line Database Editor</b></font><br>
<form method=post action=de_opts>
<input type=hidden name=record>
<input type=submit name=de_button value="Save">
<input type=submit name=de_button value="Copy">
<input type=submit name=de_button
value="Delete">&nbsp;&nbsp;&nbsp;
-164-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<input type=submit name=de_button value="Find"><br>
Title: <input type=text name=record_title><br>
Data #1: <input type=text name=f1><br>
Data #2: <input type=text name=f2><br>
Data #3: <input type=text name=f3><br>
</form>
</html>
<<OVERLAY de_opts
de_query= 'record="'+record+'"'+
'AND record_user="'+lp_login+'"'
IF de_button="Save" THEN
de_temp=DBFIND("bigdb",de_query,1,1,"record")
IF de_temp=record THEN
de_temp=DBEDIT("bigdb",record)
/IF
ELIF de_button="Copy" THEN
record_title=record_title+"(New)"
record_user=lp_login
de_temp=DBADD("bigdb",record)
record=temp[2,1]
ELIF de_button="Delete" THEN
de_temp=DBFIND("bigdb",de_query,1,1,"record")
IF de_temp=record THEN
de_temp=DBDELETE("bigdb",record)
/IF
ELIF de_button="Find" THEN
GOTO "finddb.html"
/IF
GOTO PAGE
>>
Like the eight-line editor presented at the end of Chapter 11, The Web Database, the fifteen-line
database editor uses few Overlay tags. The editor is built with IF-THEN, DBFIND, DBADD,
DBEDIT, DBDELETE, and GOTO tags. The HTML form consists of nine components consisting of
four submit buttons, four text boxes, and a hidden input component.
The hidden HTML form component is used to guarantee the editor works even if the page was
arrived at using a browser's Back button. See the accompanied "Back Button Protection" note.
Back Button Protection—The Back button of HTML browsers allows users to access pages in
their browser's history. This can create situations in which the data stored on the server is different
than that displayed on the screen, which can cause problems if you are not careful. For example,
suppose you are editing one record and then, later on, start editing another. By clicking the Back
-165-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
button of your browser you can recover the first edit screen. Submitting the page resubmits the
data in that page to HTML/OS, which thinks you are on the second page. The remedy to this is to
place in the HTML form a hidden input that stores the record ID. That way, even if the page is
submitted, the record is resubmitted so your application knows which record to work with.
When a user clicks a fifteen-line editor button the on-click Overlay de_opts runs. If the user clicks
the Save button, you do a security check using the DBFIND tag. You check to see whether the
record you are about to write is owned by that user, in other words, whether the record_user field
is equal to the user's login ID. The check ensures that you write to records only when the user owns
them. If the check is successful, de_temp will be equal to the record ID. If unsuccessful, de_query
will be equal to the empty string. The record is saved only if the test is successful. Note that
de_query is composed by writing the following code:
de_query= 'record="'+record+'"'+
'AND record_user="'+lp_login+'"'
The 'AND record_user="'+lp_login+'"' part of the query restricts access to only records
belonging to the user. In general, security on a database is set up this way. See the accompanied
note, Adding Group-Based Security.
Adding Group-Based Security—Group-based security is when only those users belonging to a
specific group can read or write to a record. To convert this page to one that uses group-based
security, place a group code in the field record_user. Determine the group code for users when
they log in, and in de_query, replace the lp_login variable with the variable for the group code.
When the user clicks the Copy button, you create a new record using the data in the HTML form.
This is the user's new record option. Here we call it Copy. You can name it whatever you want. No
security check needs to be done here; but you do need to make sure the user's login ID is saved to
record_user so future accesses are protected.
When the user clicks the Delete button, you want to delete the record. As with the Save button, you
do a security check before changing the record. You use the same technique you used in the Save
option which deletes the record only if the user owns it.
Adding User Feedback
The fifteen-line editor can be improved by adding user feedback to it. You do this by displaying a
message at the top of the page after a user clicks a button, providing feedback to actions taken. This
is a highly recommended modification, since without this feedback users may become uncomfortable
regarding the success or failure of the actions they take. The editor with user feedback is as follows:
<html>
<title>Database Editor With Feedback</title>
<font size=5><b>Database Editor With Feedback</b></font><br>
<< IF de_msg != "ERROR" THEN
DISPLAY "<br>"+de_msg /DISPLAY
/IF
de_msg="ERROR"
-166-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
>>
<form method=post action=de_opts>
<input type=hidden name=record>
<input type=submit name=de_button value="Save">
<input type=submit name=de_button value="Copy">
<input type=submit name=de_button
value="Delete">&nbsp;&nbsp;&nbsp;
<input type=submit name=de_button value="Find">
Title: <input type=text name=record_title><br>
Data #1: <input type=text name=f1><br>
Data #2: <input type=text name=f2><br>
Data #3: <input type=text name=f3><br>
</form>
</html>
<<OVERLAY de_opts
de_query='record_user="'+lp_login+'" AND record="'+record+'"'
IF de_button="Save" THEN
de_temp=DBFIND("bigdb",de_query,1,1,"record")
IF de_temp=record THEN
de_temp=DBEDIT("bigdb",record)
de_msg="Record saved."
ELSE
de_msg="<font color=red>Cannot save. Use Copy
button.</font>"
/IF
ELIF de_button="Copy" THEN
record_title=record_title+"(New)"
record_user=lp_login
de_temp=DBADD("bigdb",record)
record=temp[2,1]
de_msg="Record added."
ELIF de_button="Delete" THEN
de_temp=DBFIND("bigdb",de_query,1,1,"record")
IF de_temp=record THEN
de_temp=DBDELETE("bigdb",record)
de_msg="Record deleted."
ELSE
de_msg="<font color=red>Cannot delete.</font>"
/IF
ELIF de_button="Find" THEN
-167-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
GOTO "finddb.html"
/IF
GOTO PAGE
>>
The first Overlay on the page displays the message. The other modifications, all shown in bold, set
values for de_msg. Note that IF-THEN statements are extended with ELSE tags to ensure all
situations are captured.
The modified code shown here does a good job at providing feedback to the user. However, it can
be improved even further. The Save and Copy buttons do not validate the data. Are any of the fields
required? Should they be of a specific data type? These issues have not been addressed in the code.
Such code will depend on the type of data you wish to validate. To add data validation to this page,
apply the code provided in Chapter 10, HTML Forms Processing.
The Six-Line Find Page
The Find button in the fifteen-line database editor directs the user to the Find page, finddb.html.
This page needs to give the user the ability to select a record, load the record, and direct the user
back to the editor.
The design of this page will vary depending on the nature of your application. The Find page of a
real-estate database may include a search by MLS number. The Find page of a classifieds system
may simply be a list of the ads placed, showing the user which ads are active and which are not. The
Find page for an ordering system may show the user the last fifty orders placed.
When the number of records belonging to a user is high, you provide the user a search option. If the
number of records per user is limited, it is best to list the titles of those records on the screen. Here's
the code for a Find page that does that:
<<
de_query='record_user="' + lp_login + '"'
de_choices=DBFIND("bigdb",de_query,1,100,"record,record_title")
>>
<html>
<title>Six-Line Find Page</title>
<b>Select Record to Edit:</b><br>
<a href=editdb.html>Back to Editor</a><ul>
<<
FOR NAME=de_choices ROWNAME=x DO
DISPLAY
"<a href=selectit name=record value=" + x[1] + ">" + x[2]
+"</a>" + "<br>"
/DISPLAY
/FOR
>>
</html>
-168-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<<overlay selectit
de_temp=DBGETREC("bigdb",record)
GOTO "editdb.html"
>>
At the top of the page you use DBFIND to fill de_choices with the first 100 records belonging to the
user. Then in the page, using a FOR loop, you display the titles of each. When the user clicks a title,
record is set, and the Overlay selectit runs. There, the record loads, and the user returns to
editdb.html to edit the record.
Other Features
The database record editor described here can be modified and expanded in a variety of ways.
Three common changes involve switching a text box for a text area, replacing the text boxes with
pull-down menus or other HTML components, and adding an image upload option to the Edit page.
Adding a Text Area
You can apply all of the advanced editor options discussed in Chapter 8, Building Text Editors, here.
The HTML form element consisted of a text area.So, if you replace one or more of the elements of
the HTML form used here with a text area you can use the features discussed in that chapter. For
example, you can add text area resizing to it. You can set a cookie to the size of the text area. You
can add spell-check too. Refer to Chapter 8 for details.
Varying the HTML Form
Furthermore, using what you learned in Chapter 10, HTML Forms Processing, you can replace the
text box components with other HTML form components. For example, suppose the field f2 was
restricted to the values A, B, and C. Then you would want to replace the following line:
Data #1: <input type=text name=f2><br>
with the HTML form component:
Data #1: <select name=f2>
<option value=A>A
<option value=B>B
<option value=C>C
</select>
Depending on the nature of the application, you may want to use any number of HTML form
components. Refer to Chapter 10 for descriptions of each HTML form component.
Adding File Upload
As another example, suppose the field f2 contained the name of an image. You might want to
display the image next to the input box and place an Upload hypertext link next to that. When a user
clicks the link, you would direct the user to a more sophisticated version of the six-line upload page
presented in Chapter 10. The page would look as follows:
<html>
<title>Ten-line Image Upload Page</title>
-169-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<b>Ten-line Upload Page</b>
<form enctype="multipart/form-data" ACTION=upload_me><br>
<font size=1>FILE</font><br>
<input type=file name=file<<usernum>> size=20>
<input type=submit value="Upload Image Now">
</form>
</html>
<<OVERLAY upload_me
upfile="file"+usernum
de_ext=LOWER(RIGHT(@upfile,4))
IF de_ext != ".gif" AND de_ext != ".jpg" THEN
de_msg="Unsupported image type" GOTO "editdb.html"
/IF
temp=GETCOLEQ(FILELIST("/upload"),1, "file"+usernum)
IF temp[1]= "file"+usernum AND temp[2] > 500 THEN
f2=record+de_ext
temp=SYSMV("/upload/file"+usernum, "/myimages/"+f2)
de_msg="Upload Successful."
ELSE
de_msg="Upload Failed."
/IF
GOTO "editdb.html"
>>
Like the six-line upload page described in the "Upload File" section of Chapter 10, here too, the file
needs to be captured. But here you have provided a name for the upload file. It is record.jpg or
record.gif depending on the type of image uploaded.
You also add file-type validation here, accepting only images with .gif and .jpg extensions in their
names. To determine the kind of image file uploaded, you capture the last four characters of the
filename on the browser computer. Use the two following lines:
upfile="file"+usernum
de_ext=LOWER(RIGHT(@upfile,4))
The first of these two lines places the variable name used in the HTML form's file component into the
upfile variable. The name of the variable itself varies so this is tricky. When a file is uploaded, the
filename of the uploaded file (on the browser computer) is placed in that variable name set in the
HTML form upload component. For example, if the uploaded file is C:\myscanner\ABCD.GIF and
the user session is 45, the contents of the variable file45 would be C:\myscanner\ABCD.GIF.
Then, the first of these two lines would place file45 in upfile. The second line would place in
de_ext the lower-case value of the last four characters of C:\myscanner\ABCD.GIF (the value
stored in the variable name stored in upfile). In HTML/OS, preceding the name of a variable with
the @ character tells HTML/OS to use the contents of the variable specified in the variable instead of
the contents of the variable itself.
-170-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
After these two lines run, an IF-THEN statement validates the file type. If the file type is okay, the file
moves to its proper destination in the /myimages directory, f2 is set, and the user is returned to the
editdb.html page.
Working with Multiple Databases
The database editors discussed in this chapter have until now worked with a single database. In
some circumstances this is not enough. Sometimes a database will be related to secondary detail
information that needs to be edited along with the database. That information might also reside in a
database. The first database is called the primary database. The second is called the secondary
database.
Product orders, for example, are organized this way. When orders are placed, each is assigned an
order number and stored in a primary order database. The items in each order are stored in a
secondary items database. Each record in the items database includes the order number, so it can
be related to an order record. An order placed for three items requires a record in an order database
and three records in an items database.
You also need a multidatabase system when working with multiple-choice quizzes. You can use a
three-database system. You place each quiz in a quiz database. You store the questions used in a
quiz in a questions database. You store the possible answers to each question in an answers
database.
TE
AM
FL
Y
As another example, consider a project database consisting of projects and action items. Assume
each project has zero or more action items associated with it. This needs two databases: one
containing project descriptions and the other containing action items, start dates, and due dates. It is
best to edit these two databases at the same time-using the same database edit page. Such a page
is shown Figure 14.2.
Figure 14.2: The Project Database Edit Page Works with Two Databases.
The page allows the user to enter a project description and a list of action items. Each action item
has a title line, a start date, and a due date. Text boxes to add a new action item are also provided.
To build this page, you make some assumptions about the databases being used. You call the
primary database projects and give it the fields record, project_num, project_name, and
project_description. The secondary database, you call todos with the fields record,
-171Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
project_num, todo_line, todo_startdate, and todo_duedate. Using these assumptions,
the code for the page is as follows:
<html>
<title>Twenty-line Project Edit Page</title>
<b>Project Edit Page</b>
<form ACTION=project_opts>
<input type=hidden name=record>
<input type=hidden name=project_num>
<table border=0><tr><td align=right colspan=4>
Project: <input type=text name=project_name length=50><br>
<textarea name=project_description cols=65
rows=7></textarea></td></tr>
<tr><td>Action Item</td><td>Start Date</td><td>Due
Date</td></tr>
<< temp=DBFIND("todos",'project_num="' + project_num +
'"',1,100,
"todo_line,todo_startdate,todo_duedate,record")
IF temp[1,1] != "" THEN
FOR NAME=temp ROWNAME=x DO
trashlink='<font size=1>'+
'<a href=rmx name=x value='+x[4]+'>DELETE</a></font>'
DISPLAY '<tr><td>'+x[1]+'</td><td>'+x[2]+'</td>'+
'<td>'+x[3]+'</td><td>'+trashlink+'</td></tr>'
/DISPLAY
/FOR
/IF
>>
<tr><td colspan=4><hr size=1></td></tr>
<tr><td>New Action Item</td>
<td>Start Date</td><td>Due Date</td></tr>
<tr>
<td><input type=text name=new_todo size=25></td>
<td><input type=text name=new_startdate size=15></td>
<td><input type=text name=new_duedate size=15></td>
</tr></table><br>
<input type=submit name=mybutton value="Save Changes">
<input type=submit name=mybutton value="Delete Project">
<input type=submit name=mybutton value="Menu">
</form>
-172-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</html>
<<overlay project_opts
IF mybutton="Menu" THEN GOTO "project_finder.html"
ELIF mybutton="Delete Project" THEN
temp=DBDELETE("projects",record)
temp=DBREMOVE("todos","project_num="+project_num)
GOTO "project_finder.html"
/IF
# Save Changes /#
temp=DBFIND("projects","project_num="+project_num,1,1,"record")
IF temp[1,1]=record THEN
# edit record /#
temp=DBEDIT("projects",record)
ELSE
# create new record /#
project_num=DBFINDSORT("projects","",1,1,"project_num","y","project
_num")+1
IF ISINTEGER(project_num)="FALSE" THEN project_num=1 /IF
temp=DBADD("projects")
/IF
IF new_todo != "" THEN
todo_line=new_todo
todo_startdate=new_startdate
todo_duedate=new_duedate
new_todo=""
temp=DBADD("todos")
/IF
GOTO PAGE
>>
<<overlay rmx
temp=DBDELETE("todos",x)
GOTO PAGE
>>
The page includes the elements you find in a standard edit page plus three additional features: a
listing of related line items from the secondary database, a delete option next to each item, and at
the bottom of the page, an option to add a new line item.
-173-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The listing of related items uses a DBFIND tag to find the related items and a FOR loop to list them
on the page. The list of items is placed in an IF-THEN statement to ensure the list is displayed only
when items are found. Also note that a Delete link is placed at the end of each row. The link, when
clicked, runs the Overlay rmx, which uses a DBDELETE tag to remove the item before redisplaying
the page.
The bottom of the HTML form includes Save Changes, Delete Project, and Menu buttons.
When a user clicks the Save Changes button, the DBFIND tag sees whether the project submitted
exists in the projects database. If it does, the record is updated with DBEDIT. If it does not, a new
projects record is created. This requires the creation of a new project number. It does this by
taking the largest project number in the projects database (using DBFINDSORT) and incrementing
it.
The line below the DBFINDSORT line sets project_num to 1 in the event DBFINDSORT yields a bad
result—a situation that occurs when the project database is empty.
When a user clicks the Delete Project button the following two lines run:
temp=DBDELETE("projects",record)
temp=DBREMOVE("todos","project_num="+project_num)
The first deletes the current record in the projects database. The second uses a DBREMOVE tag,
an Overlay tag for deleting multiple records concurrently similar to DBDELETE except that its
parameter is a query string rather than a record number. See Appendix D, HTML/OS Tag Reference
Guide, for a description.
When a user clicks the Menu button the user is directed to the project_finder.html page. The
page is a fictitious query page left as an exercise at the end of this chapter.
Summary
In this chapter you learned about database editors, an important component of sophisticated
database-driven Web sites and Web-based applications. You learned how to add user or group
security to database records and how to combine multiple databases in a single page. The concepts
learned here, along with those in the previous chapters of Part III, Database Programming, combine
to give you awesome Web development capabilities.
In the next chapter you learn about database networking, a topic based on the concept of a Web
network—a feature built into the HTML/OS engine that allows you to access databases and files
residing at remote locations across the Web.
Exercises
In the following exercises you build database edit pages using what you learned in this chapter and
the previous three chapters. Answers to these exercises are provided on this book's companion Web
site as described in the book's Preface.
Exercise 1
The database editor described in the "Fifteen-line Database Editor" section of this chapter has an
HTML form with three text boxes. Change one of the text boxes to a text area. Make the text area
adjustable using what you learned in Chapter 8, Building Text Editors.
-174-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 2
Expand the six-line find page described in this chapter to include a search box that limits the display
of results. Also provide a Clear button, that when clicked, goes to an on-click Overlay that resets the
search. Place the search box at the top of the page, just above the report. Use what you learned in
Chapter 12, Building Query Pages, to complete this exercise.
Exercise 3
Using what you learned in Chapter 13, Building Database Reports, add page up and page down
links to the find page you built in Exercise 2.
Exercise 4
The project database described in the "Working with Multiple Databases" section of this chapter
does not include user security. It is designed for a single user. Using the same technique used in the
"Fifteen-line Database Editor" section of this chapter for extending the page so it can work with
multiple users.
Exercise 5
Using what you learned in Chapter 9, Building Login Pages, build a login page for the project
database application described in the "Working with Multiple Databases" section of this chapter and
connect the page to the page you created in Exercise 4. Then, using what you learned in Chapter 12,
Building Query Pages, build a project_finder.html page for it. You now have a complete
multiuser, Web-based project-management application.
-175-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 15: Database Networking
Web-based database networking is about accessing databases that reside at one location on the
Web with Web-based programs located at other locations. Surprisingly, although the Web excels at
networking, performing these tasks has not been straightforward within HTML documents and Webbased applications. In 1998, to make network access easier, Aestiva invented the Web network—a
network that sits on top of the Web. The database networking used in HTML/OS takes advantage of
this technology.
In this chapter you first look at the basics of Web networking and learn why it is such a powerful tool
for developing advanced Web sites and Web-based applications. Then you learn how to set up a
Web network. Later you learn how to define different levels of access, how to read and write to
network databases, and you review development topics of interest to developers of Web sites that
use networked databases.
The Web Network
The World Wide Web is the ultimate network. Its success stems from its underlying TCP/IP network
protocol. In recognition of this, Aestiva developed a network architecture that takes advantage of the
TCP/IP network. To ensure that it takes advantage of the TCP/IP network, the HTML/OS Web
network architecture uses a direct interface to the underlying TCP/IP network of the Web. It uses no
intermediate protocols. As a result, it shares many amazing features of the TCP/IP architecture,
while adding a few of its own—the most striking of which, is that it's easy to use.
An HTML/OS Web network comprises of two or more copies of HTML/OS placed at different Web
locations. Each copy of HTML/OS has a node name in much the same way computers off the Web
have network drive names. Give each copy of HTML/OS a name such as A, B, C, or D. Once that
is done and the network connections are set up, databases in your network can access the Web
network by preceding them with their node name. If you are on Node B and want to access the
/work/projects database on Node A you write A:/work/projects.
The Web network works for files too. If you are on Node A and want to access the file /hello.html
on Node B, you write B:/hello.html. The Web network also works for HTML documents. Need to
link a Web page on the A computer to B:/hello.html? Then write your hypertext link as follows:
<A HREF=B:/hello.html>Welcome Message From The President</a>
The HTML/OS Web network frees you from the intricacies of TCP/IP internals—allowing you to add
sophisticated networking within the programs you build. The HTML/OS Web network has the
following features:
ƒ World-wide— Different copies of HTML/OS in a Web network can reside across the world.
ƒ Hardware-irrelevant— Copies of HTML/OS can run on Macintosh, Windows, Unix, or
Linux systems. The network can be heterogeneous, consisting of files and databases
residing on different kinds of hardware.
ƒ Distributed— If a server running a copy of HTML/OS fails, others on the Web network will
continue to operate as long as they don't need to access the failed server. There is no
central point of failure.
ƒ Scalable— If a single server can't support the traffic going to it, access can be spread
across multiple computers by spreading the HTML documents across multiple servers.
-176-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ Tolerant— If a network connection fails, when it resumes, so too will the copies in the
Web network dependent on that connection.
ƒ Easy-to-program— You access files and databases by preceding file or database names
with their location in the Web network.
ƒ Easy Setup— Network setup is point-and-click, using the Control Panel that comes with
each copy of HTML/OS.
The HTML/OS Web network architecture shares the cross-platform, scalability, and stability features
of the Web. But unlike TCP/IP network programming, Web networking is accessible without special
programming. Instead you simply follow file and database naming conventions whereby database
names are preceded with node names. In the next few sections you learn what a network node is
and how Web networks are set up.
The Network Node
The Web network in HTML/OS consists of two or more copies of HTML/OS. Each copy is called a
node. Nodes can reside at different physical locations. Most often these nodes reside on the public
Web, however Web networks can also be set up on Web servers placed within private intranets.
They can even span equipment placed on both sides of a firewall (spanning an intranet and the Web)
See the accompanying "Working on Both Sides of the Firewall" note if your network spans both the
Web and an intranet.
Working on Both Sides of the Firewall—The Web network used in HTML/OS has the ability to span
both sides of a firewall. A firewall is a network gateway that blocks users on the public Internet from
accessing sites within a private intranet while still allowing users in the intranet to access the Web.
Note that HTML/OS does not adversely affect the firewall. The firewall still blocks access from the
Web to copies of HTML/OS within the intranet, but copies of HTML/OS within the intranet will be
able to read and write to databases and files located on copies of HTML/OS across the firewall, on
the Web.
When any two nodes in a Web network communicate with each other, they do so based on the
settings in the two nodes. Each node determines which other nodes can access its files and
databases. A node cannot access another node unless that node gives it access. Security is the
responsibility of each node in the network. This is known as a peering architecture. To set up a Web
network, you must set up each HTML/OS copy in the Web network. To set up a node, you use the
Network and Security menu options in the HTML/OS Control Panel.
Network Setup
Consider setting up a Web network consisting of three copies of HTML/OS. To set up the network,
you need to know the name of each node, the URL to each copy of HTML/OS, and (for security
reasons) the first part or the entire IP address of each node. In this section assume the following for
each copy of HTML/OS:
ƒ
Node
1
Node
2
Node
3
Network Identity: A
IP Address: 128.50.123.2
URL: http://www.d1.com/cgi/htmlos
Network Identity: B
IP Address: 128.50.50.135
URL: http://www.d2.com/cgi-bin/htmlos.cgi
Network Identity: C
IP Address: 128.50.?.?
URL: http://www.d3.com/scripts/htmlos.exe
-177-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Setting up a network involves two steps. First you need to define connections between the nodes. In
Node 1 you define connections to Node 2 and Node 3.
In Node 2 you define connections to Nodes 1 and 3, and in Node 3 you define connections to Nodes
1 and 2. Second, in each of the nodes, you need to enable access to the other nodes. The following
sections give you details for setting up the nodes.
Setting up Node 1
To set up Node 1, access that copy of HTML/OS and follow these steps:
1. Select the Control Panel from the desktop menu bar.
2. Click Network in the Control Panel to expand the left menu so you can see your
Network options. Next you set up the copy of HTML/OS. The name of the node
is called the node identity.
3. Click Identity and enter A for the node name.
4. Click the Save Settings button when done. Now the node identity is set for this
copy of HTML/OS. The node identity precedes file and database names
specified in HTML documents at other nodes to access files and databases on
this node. For example, to access the file /hello.html from another copy of
HTML/OS you can write A:/hello.html. Note that files and databases are
specified with their full path, starting with a required slash (/).
5. Next set up access to nodes B and C. Click Configuration on the left Control
Panel menu. You'll see a screen like that shown in Figure 15.1. This screen
allows you to specify multiple nodes. Each node line requires the name of the
node, the URL to the node and the IP address of the remote node.
Figure 15.1: The HTML/OS Control Panel Network Configuration Screen Contains the
Nodes You Peer With.
The name of the node is the node identity of the particular node. Node identities can be single
characters or multiple characters (no spaces). In general, it is best to keep node names short since
they precede file and database names. Nobody likes to type long names! Single character names
are preferred. Or you can use short descriptive names like web, host1, or s25. The URL to the
node is the URL you would enter in the Location text box of your browser to access the remote node.
The last part of the URL, which is typically htmlos, htmlos.cgi, or htmlos.exe, is the name of
-178-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
the HTML/OS engine on the remote server. This URL is the URL to the login page of the remote
server less the /login.html part.
The IP address is a security setting. When incoming network requests from a remote nodes are
received, this IP address is checked to ensure the remote node is what it says it is. If the IP address
is omitted, no check is done. If a partial IP address is provided, the beginning of the incoming IP
address must match the setting. If a remote node comes from a location where IP addresses are set
dynamically, use the first two or three parts of that IP address—the part of the IP address that does
not change dynamically. An example is provided in a moment. Continue with the following steps to
network Nodes 2 and 3:
6. So to network with Node 2 type B in the Node box, http://www.d2.com/cgibin/htmlos.cgi in the URL box, and 128.50.50.135 in the IP box and click
Save.
7. For Node 3 type C in the Node box, http://www.d3.com/scripts/htmlos.exe, in the
URL box and 128.50 in the IP box; then click Save again.
Now you have links set up to both of these nodes. Note that only the first two parts of the IP address
are specified since the rest of the IP address is unknown. This is what happens when the remote
server has a dynamically generated IP address. If you are unsure of the IP address needed, use the
Get IP Address tool at the bottom of the page. See the accompanying "The IP Address Tool" note.
The IP Address Tool—The bottom of the Configuration screen includes a Get IP Address button.
Use this if you do not know the IP address of a remote domain. Enter the domain used in the URL
in the URL box and click Get IP Address. This will give you the IP address of that domain. Note
that this IP address is the IP address as seen by that server. Actually, the IP address needed in
this configuration is the IP address detected from a remote request, which sometimes is different
than the one reported here. If this happens and you're not concerned about security breaches
within your organization, chop off the last section of the IP address and use that. When the two IP
addresses differ, they usually differ only in the last part of the IP address. Dropping this last part of
the IP address is a quick and easy way to enable the connection without having to hunt down the
exact IP address seen in the reverse direction.
Setting up Node 2
Setting up the second node in our example is similar to setting up the first. Access this copy of
HTML/OS, pull up its Control Panel, and follow these steps:
1. Click Identity and enter B for the node name. Click the Save Settings button
when done. Now the node identity is set for this copy of HTML/OS.
2. Then click Configuration. You need to set up connections for Nodes 1 and 3.
3. To set up Node 1, type A in the Node box, http://www.d1.com/cgi/htmlos in the
URL box, and 128.50.123.2 in the IP box; then click Save.
4. To set up Node 3 type C in the Node box, http://www.d3.com/scripts/htmlos.exe in
the URL box, 128.50 in the IP box, and click Save.
Setting up Node 3
To set up the third node in our example access this copy of HTML/OS, pull up its Control Panel, and
follow these steps:
-179-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
1. Click Identity and enter C for the node name and click Save Settings.
2. Then click Configuration on the left menu to set up connections for Nodes 2 and
3.
3. Type A in the Node box, http://www.d1.com/cgi/htmlos in the URL box,
128.50.123.2 in the IP box, and click Save.
4. Enter B in the Node box, http://www.d2.com/cgi-bin/htmlos.cgi in the URL box,
128.50.50.135 in the IP box, and click Save again.
Enabling Network Access
Once you set up your connections, you need to enable them, which involves using a network
security setting. In the Control Panel of any of your three copies follow these steps:
1. Click Security to view the security options.
2. Then click Network Allow. You will see a list of remote nodes as shown in Figure
15.2. Next to each is a pull-down menu containing the two options: Limited Access
and Full-Access.
Figure 15.2: To Give a Node Full Network Access to Your Node Set Their Node to FullAccess in the Network Allow List Page of the HTML/OS Control Panel.
3. Set these to Full-Access.
4. Click Save Settings when done.
Repeat for the other two nodes. Now the nodes are fully networked together. Later, in the
"Controlling Network Access" section of this chapter, you learn how to provide only limited network
access between nodes, but for now, and in situations where staff with access to one node already
has access to the others, this setting is fine.
Testing the Network Setup
After setting up the network nodes and enabling access you will want to test the connectivity. Here's
a page you can use to test the networking. This will test the ability of HTML/OS, on a specific node,
to access your other nodes. The code for the page is as follows:
<html>
<title>Six-Line Network Test Page</title>
Network Test:<ul>
<form method=post ACTION=testnetwork>
-180-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Enter Remote Node Name: <input type=text name=remote_node><br>
<input type=submit value="Test Access to Node">
</form>
<< IF mytest != "ERROR" THEN
DISPLAY mytest /DISPLAY
mytest="ERROR"
/IF
>>
</html>
<<overlay testnetwork
remote_file= remote_node+":/junk.txt"
COPY "12345" TO FILE=remote_file /COPY
mytest="Writing 12345 to "+remote_file:
"+TAGRESULTS[1,2]+"<br>"
COPY FILE=remote_file TO temp /COPY
mytest=mytest+
"Got the value: "+temp+"<br>"
GOTO PAGE
AM
FL
Y
"Reading "+remote_file: "+TAGRESULTS[1,2]+"<br>"+
TE
>>
Copy this file to each node in the network. Run the page to test network access. The page asks the
user for a node identity. Enter a remote node name and click Test Access To Node. The page tries
to write the characters 12345 to a file on the remote server. It then tries to read the file it wrote. As it
does this it saves the status messages returned by the Overlay tags that write and read from the
remote server. These status messages are displayed for you at the bottom of the page after each
test.
If network access is okay then you'll see a message like the following:
Writing 12345 to B: OK
Reading B: OK
If you do not get this, you have not successfully accessed the network. Use the error messages
displayed to fix network configuration problems, if any.
Using the Network Error Log—If after using the testing page in the previous section of this
chapter you are still unable to establish a network connection, use the network error log in the
Control Panel. The log is written to when incoming or outgoing network accesses fail. To access
the log open the Control Panel. Click Network to access your Network options and then click Error
Log. Error lines in the report can include valuable information. For example, if a remote access
was denied by HTML/OS because the IP address of the originating request didn't match the IP
address setting in the Network-Configuration section of the Control Panel, then it will be indicated
here along with the IP address of the originating request.
Controlling Network Access
In the prior sections you set up a three-node Web network with full network access. This allows any
node in the network to access files and databases in any other node. Often however you do not want
-181Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
unrestricted access between nodes. Full-access to a remote node does just that: It gives users of the
remote node full access to your system. With full access, even if they do not know the password of
the node, they can get it by networking to the server. To limit access run the Control Panel and follow
these steps:
1. Click Security; then click Network allow.
2. Change the node to Limited Access and click Save Settings.
3. Then click Database allow. You'll see a screen like that shown in Figure 15.3.
Figure 15.3: The Network Database Allow List Page of the HTML/OS Control Panel is
Where You Place Limitations on Other Nodes Accessing Your Databases.
The top of the page displays the current access security of each node defined in the Network
Configuration page. To change the settings, use the options provided at the bottom of the page. You
can set two kinds of access: database-specific and database nonspecific, known here as the default.
Default Access
Use the Set Defaults section of the page to set up access security to databases, in general. You may,
for example, not mind it if specific nodes or all nodes have Read-Only or even Read/Write access to
your databases. Note that you can apply these to all nodes or only specific nodes, and if a specific
node needs to be denied access to your databases, you can set this as well.
As an example, suppose you wish to give Node C read access to all of your databases. To do that
select C, select Read Only, and click Save.
Database-Specific Access
Use the Change Access section of the page to set up access security for specific databases. The
default settings discussed in the previous section have secondary priority to these database-specific
security settings. If for example by default all nodes have access to all databases but you add a node
specific restriction, then the node specific restriction will have priority over the default settings.
Extending the example in the previous section, suppose you still want Node C to have read access
to your databases, however you want to give Node C write access to the /work/postings
database, and deny Node C access to the /mypasswords database as follows:
1. Use the Change Access section of the page.
2. Select C, and then select Read/Write.
-182-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
3. Type /work/postings in the database text box, and click Save.
4. Then select C again.
5. Select No access, type /mypasswords into the text box, and click Save. The
access rights for Node C will be described at the top of the page. It will read as
follows:
6.
Default read-only access provided to all databases.
7.
No access to: /mypasswords
8.
Read/Write access to: /work/postings
Serving Network Databases
Often, when setting up a network, you do not have access to any copies of HTML/OS other
than your own. At the same time, you wish to give access to a database residing on your
copy of HTML/OS without jeopardizing the security of your copy.
If this is what you want to do, then you want to be a Database Service Provider (DSP), a
fancy way of saying you want to serve one or more databases to customers in a controlled
and secure fashion. This section shows you how.
First, you set up your copy of HTML/OS as a DSP. To do that, give your node an identity
such as DSP. Then collect the node names and IP addresses of your first customers
wishing to access your database. If they do not have node names, you can select some for
them. To set up customers you do not need their URLs. Note also that you do not need to
set up customer nodes; you simply need to forward customers your network parameters
and their new node name, if necessary. Once you are set up, issue a DSP confirmation email to your customers that looks something like this:
Dear Marketing Department Web Designer,
Access to DSP databases has been configured. Please place the following network
parameters in the Network section of the Control Panel in your copy of HTML/OS:
1) Set your Node Identity to: M1
2) Set up our DSP node in your Configuration:
Node: DSP
URL: http://dsp.ourdomain.com/cgi-bin/htmlos.cgi
IP: 128.356.65
3) You have been authorized for the following:
Database: Access
dsp:/sales/orders: Read Only
dsp:/sales/incoming: Read/Write
Thank you and best regards.
Jan Jackson
DSP Administrator
To set up the access level shown in this sample e-mail, do the following:
1. Access the Control Panel in your copy of HTML/OS.
2. Click Network and then click Configuration.
3. Type M1 for the Node, leave the URL blank, enter their IP address, and click
Save.
4. Then on the left menu bar, click Security and Network Allow.
5. On the Network Allow page, change the access setting for the node M1 to
Limited-Access, and click Save Settings.
-183-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
6. Then on the left menu bar, under Security, click Database Allow. You'll see a
screen that displays the network access parameters for the nodes in your
system.
7. Make sure the Default Settings for All Nodes is set to None.
8. In the Change Access section, you can select a node, access level and enter a
database name. First select M1, Read Only, type /sales/orders and click
Save.
9. Finally select M1, select Read/Write, type /sales/incoming, and click Save.
Now M1 is set up. Repeat for other customers, keeping everything the same
except the customer node name.
Accessing Network Databases
You access databases located at remote nodes much the same way as you access regular
databases. For example, consider the seven-line database report provided in Chapter 13, Building
Database Reports. If the database used in the report resides on Node B, you could write the report
as follows:
<html>
<title>Seven-line Network Database Report </title>
<< myresults=DBFIND("B:/work/contacts",myquery,1,100,
"contact_company,contact_name,contact_phone")
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088 colspan=4 align=center>
<b>My Report</b></td></tr>
<tr>
<td>Column One</td>
<td>Column Two</td>
<td>Column Three</td>
</tr>
<<
FOR NAME=myresults ROWNAME=myrow DO
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
</table>
</html>
The only change in the report that's necessary is the need to precede the database name with the
name of the remote node (indicated in bold). The problem with this however, is that it does not
-184-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
address a fact of life in systems that work with networks—that the system administrator of Node B
may one day change the network name from B to something else. If all you have is this one page,
renaming will not be a problem. But what if you have many pages and many references to Node B?
One way around this is to build a network node assignment file. If a node name changes, you simply
make a single change in a single file. In your programs, you use your own node variables, which are
set in this file. As an example, suppose you have three different nodes you read or write to. In your
file you can write:
node_dsp="A:"
node_finance="B:"
node_hr="C:"
Rewriting the seven-line database to take advantage of this file, you now have the following code:
<html>
<title>Eight-line Network Database Report</title>
<< EXPAND FILE="/nodes.txt" /EXPAND
myresults=DBFIND(node_finance+"/work/contacts",myquer,1,100,
"contact_company,contact_name,contact_phone")
>>
<table border=1 cellspacing=0 cellpadding=2>
<tr><td bgcolor=#000088 colspan=4 align=center>
<b>My Report</b></td></tr>
<tr>
<td>Column One</td>
<td>Column Two</td>
<td>Column Three</td>
</tr>
<<
FOR NAME=myresults ROWNAME=myrow DO
DISPLAY "<tr>" /DISPLAY
DISPLAY "<td>"+myrow[1]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[2]+"</td>" /DISPLAY
DISPLAY "<td>"+myrow[3]+"</td>" /DISPLAY
DISPLAY "</tr>" /DISPLAY
/FOR
>>
</table>
</html>
Changes are shown in bold. At the top of the file, the nodes.txt file is expanded. The EXPAND tag
runs the instructions in the page as though the code had been written at the location where the file
was expanded. The file defines the variable node_finance, which is used in the subsequent
-185-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
DBFIND tag to set the node name of the database. Further information on DBFIND and EXPAND is
available in Appendix D, HTML/OS Tag Reference Guide.
Writing to Network Databases
As you see here, network databases are easy to access. In general the DB tags used to access
network databases are the same as those used locally. In a few DB tags however, you need to
specify additional parameters. This need to respecify occurs with the DBADD and DBEDIT tags, the
two most common tags for writing to databases. With these tags, you must also specify the names of
the fields you are adding or editing.
As an example, consider the eight-line database editor discussed toward the end of Chapter 11, The
Web Database. A network-friendly version of the editor follows:
<html>
<title>10-Line Network Database Editor</title>
<b>10-Line Network Database Editor</b><br><br>
Records in Network Database:
<< EXPAND FILE="/nodes.txt" /EXPAND
myrecords=DBFIND(node_finance+"/tinydb","",1,100,"record")
DISPLAY COLTOLIST(myrecords,1," ") /DISPLAY
>>
<form method=post action=dbopts>
Record ID: <input type=text name=record size=10>
<input type=submit name=mybutton value="Load">&nbsp;&nbsp;
<input type=submit name=mybutton value="Save">
<input type=submit name=mybutton value="Delete">
<input type=submit name=mybutton value="Add">
Full Name: <input type=text name=f1><br>
E-mail Address: <input type=text name=f2><br>
Phone Number: <input type=text name=f3><br>
</form>
</html>
<<OVERLAY dbopts
fieldlist="f1,f2,f3"
IF mybutton="Load" THEN
temp=DBGETREC(node_finance+"/tinydb",record)
ELIF mybutton="Save" THEN
temp=DBEDIT(node_finance+"/tinydb",record,fieldlist)
ELIF mybutton="Delete" THEN
temp=DBDELETE(node_finance+"/tinydb",record)
-186-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ELIF mybutton="Add" THEN
temp=DBADD(node_finance+"/tinydb",fieldlist)
/IF
GOTO PAGE
>>
Changes are shown in bold. Like the eight-line network database report of the previous section, you
expand node names by expanding the /nodes.txt file. You tack the node_finance variable onto
the beginning of all database names throughout the code, and in the DBADD and DBEDIT tags, you
define a fieldlist parameter, and finally at the top of the Overlay you add dbopts. Now the
database editor edits a remote database located on node_finance node, defined at the top of this
"Accessing Network Databases" section, as the B Node, rather than one that is local.
Network Failure Recovery
When writing to databases on the same physical equipment as the copy of HTML/OS, you rarely
need to concern yourself with network or equipment failure. After all, if the equipment fails so too
does your program. Backup recovery systems are usually the remedy in cases when the equipment
hosting your copy of HTML/OS fails.
When working in network environments recovering from equipment failure is more complicated,
since the copy of HTML/OS may still be active while a remote networked node is down.
Generally the topic of network recovery requires a lot of thought and planning. Especially when
product orders, membership sign-ups, customer requests, and other kinds of critical data are at
stake. It is best to develop a preventative scheme that performs two specific duties. First, when a
network goes down, the site should write out critical data to a pending area so you can recover it
after the network is up again. Second, you need to implement a scheme for bringing the system back
to its proper state.
The first step involves isolating the points in your program that write critical data to a networked node.
When the remote network is accessed, you need to test whether the network connections succeeded
and take action in your code if the network is inaccessible. This is easier than it sounds. To do this
you use the return status of the DB tags. The return status, depending on the tag, is either returned
or placed in TAGRESULTS variable. In the case of DBADD and DBEDIT, a status result is returned. As
an example, suppose you have a DBADD tag that posts critical information to a remote database. The
code might be something like this.
fieldlist="f1,f2,f3"
temp=DBADD(homeoffice:/mydb",fieldlist)
If the homeoffice node goes down, what to do? The best way to handle this is to set up a database
identical to the remote database on the local copy of HTML/OS, for example, /pending/mydb.
When the network goes down, you posted the record to the local database rather than the networked
database. To do this you write the following code:
fieldlist="f1,f2,f3"
temp=DBADD(homeoffice:/mydb",fieldlist)
IF temp[1,1] != "TRUE" THEN
-187-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
temp=DBADD("/pending/mydb")
/IF
If the network goes down, temp[1,1] is not TRUE, so posted information writes to the
/pending/mydb database instead of homeoffice:/mydb.
The complicated part is recovering from the network failure. When the node comes up, how do you
restore the data? In general there are two ways to do this: You can place a block of code in your
program that recovers automatically, or you can set up a recovery page for a system administrator.
Both have their advantages and disadvantages. An automated system requires no intervention.
However, automated recovery, if not designed and tested properly, can cause more problems than it
solves. Also, it is a good idea to check the data before you recover it. You cannot do this with
automated systems. For this reason, here we discuss only recovery pages, designed to be used by
system administrators. The code for such a page is as follows:
<html>
<title>Twelve-line Network Failure Recovery System</title>
<font size=5>Twelve-line Network Failure Recovery
System</font><br><li> Data #1 Status:
<< records=DBFIND("/backup/mydb","",1,5,"record")
IF records[1,1] != "" THEN
N=ROWS(records)
DISPLAY
"<font color=red>"+N+" records pending.</font><br>"+
"<a HREF=recover1>Recover</a>"
/DISPLAY
ELSE
DISPLAY "OK" /DISPLAY
/IF
>>
</html>
<<overlay recover1
records=DBFIND("/backup/mydb","",1,5,"record")
FOR NAME=records ROWNAME=r DO
temp=DBGETREC("/pending/mydb", r[1])
IF temp THEN
t2=DBADD("homeoffice:/mydb",fieldlist)
IF t2[1,1]="TRUE" THEN
t3=DBDELETE("/pending/mydb",r)
/IF
/IF
/FOR
-188-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/IF
GOTO PAGE
>>
The twelve-line recovery system displays a line that says Data #1 Status: OK, or if it finds
pending records, it displays the message Data #1 Status: NOK plus a recover link. When the
user clicks the link, the on-click Overlay recover1 runs. The DBFIND tag then searches the
/pending/mydb database and extracts a list of records. The FOR loop transfers each record in
/pending/mydb to the database at the homeoffice node. As the records transfer successfully,
they are deleted from the local database, /pending/mydb.
The twelve-line recovery system explained here recovers failed DBADD attempts. To recover failed
DBEDIT attempts, you use a similar technique. Suppose you have the following code:
fieldlist="f1,f2,f3"
temp=DBEDIT(homeoffice:/mydb",fieldlist)
To protect these edits, you need to save the edit to a local database. Here too you create a local
database. But unlike the previous example, you also need to store the record number of the record
being edited. When setting up the local database, add an extra field to store this record number. Call
it newrecord, for example. Once you do that rewrite the code as follows:
fieldlist="f1,f2,f3"
temp=DBEDIT("homeoffice:/mydb",record, fieldlist)
IF temp[1,1] != "TRUE" THEN
fieldlist=f1,f2,f3,newrecord"
myrec=DBFIND("/pending/mydb2", "r2="record,1,1,"newrecord")
IF myrec=record THEN
temp=DBEDIT("/pending/mydb2",record,fieldlist)
ELSE
temp=DBADD("/pending/mydb2",fieldlist)
/IF
If DBEDIT fails, the record changes are saved to /pending/mydb2. This code creates a new record
if it finds a record_pending field already containing record. Otherwise it overwrites the record.
The recovery page for this is the same as the recovery page for records for DBADD failures except
that the database name here is /pending/mydb2 instead of /pending/mydb.
Summary
In this chapter you learned about database networking. You learned that once a network is set up all
you need to do to supply access to a database is to precede the database with the name of the node
on which it resides.
You also learned that operating across a network provides both opportunities and potential pitfalls.
Pitfalls include the need to manage potential network failures. Advantages include the ability of an
organization to provide network database sharing and cooperation between departments—giving
departments within an organization greater control and autonomy over their Web operations. Web
-189-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
networking also gives organizations the ability to build systems that sit on both sides of their firewall,
allowing Web-sites and intranets to operate even if the other is experiencing network or hardware
failures.
Next, in Part IV, E-Commerce Programming, you look at how to build world-class shopping carts and
shopping system. You'll find that, once again, HTML/OS simplifies and extends your abilities as a
Web developer, giving you the power to build the most sophisticated e-commerce sites.
Exercises
In the exercises that follow, you practice setting up Web networks and network-based database
systems. These exercises are best performed in a computer lab with multiple, preinstalled copies of
HTML/OS. Answers to these exercises are provided on this book's companion Web site as
described in the book's Preface.
Exercise 1
Set up a two-node system with full-access between the nodes using what you learned in the
"Network Setup" section of this chapter. What happens when you enter a bad IP address, a bad URL,
or a bad node name?
Exercise 2
Using the two-node system you built in exercise 1, build a system that takes advantage of server
jumping—the ability to create a series of Web pages that span more than one node. Build a page
that bounces back and forth between two servers. Use the knowledge base on the Aestiva site at
http://www.aestiva.com/ support/ for a description of server jumping.
Exercise 3
Using the networking techniques you learned, add a Publish button to the six-line editor discussed at
the top of Chapter 8, Building Text Editors. When the user clicks the Publish button, copy the page to
a file of the same name on the remote server. Use the two-node system set up in exercise 1 to
accomplish this.
-190-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Part IV:
E-Commerce Programming
Chapter List
Chapter 16: Designing E-Commerce Systems
Chapter 17: Building Product Navigation Pages
Chapter 18: Building Product Detail Pages
Chapter 19: Building Shopping Cart Pages
Chapter 20: Building Checkout Pages
TE
AM
FL
Y
Chapter 21: Building Back-End Management Systems
-191Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 16: Designing E-Commerce Systems
E-commerce Web sites vary extremely in size and the features and kinds of product they offer. Some
sell memberships, others sell software downloads. Most sell products and services. Some sell only a
single product; others sell millions of products. Sites can be B2B, B2C, B2G, G2G, and C2C or a bit
of each. (In Web-speak B means Business, C means Consumer, and G means Government.) In Part
IV, E-Commerce Programming, you learn how to build these kinds of Web sites.
In this chapter you start by learning about the kinds of Web pages used when building e-commerce
sites. You begin with a one-page, one product e-commerce site. Then you build a simple databasedriven e-commerce site. Then, we provide a list of advanced e-commerce features in preparation for
the other chapters of Part IV of this book, where the Web pages used in highly advanced ecommerce sites are explained and brought to life.
A Ten-Line E-Commerce Web Site
Consider for starters a one-page Web site where product information and a form to purchase the
product are provided on a single Web page. The Web page requires only ten HTML/OS
instructions—two in an Overlay that displays a message to the user and eight in an on-click Overlay
that does the order processing. The Web page displays product information explicitly typed into the
page and an HTML form that captures the purchaser's name and e-mail address, shipping address,
and credit card information. The ten-line e-commerce Web site is shown in Figure 16.1 and the code
for that page follows.
<html>
<title>Ten-line E-commerce Web Site</title>
<font size=5>Buy the official <i>Best Web Sites Made Easy</i>
book.</font><br>
Published by Top Floor Publishing
<ul type=square><li>First edition.
<li>Make your Web development friends jealous.
<li>Price slashed from <strike>$35.95</strike>
to only <b>$29.95</b>. <font size=2>plus $6.00 S&amp;H.</font>
<li>Money back guarantee. Ships in 24 hours.
<li>We pay the sales tax!
<li>MasterCard, Visa, and AmEx accepted.
</ul>
<< IF msg != "ERROR" THEN DISPLAY msg+"<br>" /DISPLAY /IF
msg="ERROR"
>>
<form method=post ACTION=postorder>
<table bgcolor=#CECECE border=0 cellspacing=0>
<tr><td colspan=2><b>Order Form</b></td></tr>
<tr><td align=right valign=top>Full Name</td>
-192-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<td><input type=text name=shop_name size=30></td></tr>
<tr><td align=right valign=top>E-Mail</td>
<td><input type=text name=shop_email size=30></td></tr>
<tr><td align=right valign=top>Shipping<br>
<font size=1>Enter complete<br>shipping address</font></td>
<td><textarea name=shop_ship cols=30
rows=3></textarea></td></tr>
<tr><td align=right valign=top >Credit Card</td>
<td valign=top><input type=text name=shop_card></td></tr>
<tr><td align=right valign=top >Expires</td>
<td><input type=text name=shop_expires><font
size=1>(MM/DD/YYYY)</font>
</td></tr>
<tr bgcolor=#FFFFFF><td colspan=2 align=center>
<i>A total of $35.95 ($29.95 + 6.00 S&H) will
appear on your credit card.</i></td></tr>
<tr bgcolor=#FFFFFF><td colspan=2 align=center>
<input type=submit value="COMPLETE ORDER NOW">
</td></tr></table></form>
</html>
<<overlay postorder
IF shop_name="" OR shop_email="" OR
shop_ship="" OR shop_card="" OR
shop_expires="" THEN
msg="<font color=red>Missing Info. Try Again.</font>"
ELSE
myorder=
"============================"+LF+
"Order Date: "+TODAY+LF+
"Full Name: "+shop_name+LF+
"E-mail: "+shop_email+LF+
"Card: "+shop_card+" ("+shop_expires+")"+LF+
"Shipping: "+LF+shop_name+LF+
"============================"+LF+LF
MAIL myorder TO ADDRESS="[email protected]"
SUBJECT="Order ("+now+")"
/MAIL
APPEND myorder To FILE="order_log.txt" /APPEND
msg="Order Placed. Thank you. E-mail questions to
[email protected]"
-193-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
shop_card=""
shop_expires=""
shop_ship=""
/IF
GOTO PAGE
>>
Figure 16.1: The Ten-Line E-Commerce Web Site has all the Basic Capabilities of an E-Commerce Site.
This page provides the basic functionality required in an e-commerce site. When a user accesses
the page, it displays information about a product. The user is presented with an order form to fill out.
When the user clicks a Purchase button, the on-click Overlay postorder runs, and the order is
processed. If one of the entries in the HTML form is missing, an error message is set and the page is
redisplayed, informing the user of the error and allowing the user to make the necessary corrections
and resubmit the page. If the HTML form is filled out in its entirety, an e-mail message is composed
and e-mailed to the sales office using the MAIL tag. The APPEND tag backs up the order to a file. A
Thank you message is composed in the code too, so when the page is redisplayed, the user sees
the message. You write the entire page with a few IF-THEN statements and the four Overlay tags:
DISPLAY MAIL, APPEND, and GOTO.
Of course, most e-commerce sites are larger. They sell more products and provide the user a more
sophisticated ordering procedure. In the next section you look at splitting this single page ecommerce site into a more typical multipage site.
A Database-Driven E-Commerce Site
When selling multiple products a larger e-commerce site is necessary. To build such a site you split
the functionality of the one-page site discussed in the previous section across multiple Web pages.
In addition, you add product navigation along with a shopping cart where a temporary order list for
the user is saved.
Database-driven e-commerce sites have the defining feature that product information is stored in a
database rather than being hard-coded into different HTML documents. This way of constructing an
e-commerce site is helpful since it allows you to limit the number of Web pages the site requires. For
example, a database-driven Web site with one thousand products may have less than ten Web
-194-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
pages. The same site without the database would have over one thousand pages. As a result,
database-driven Web sites are easier to manage than those that are not.
In this section you take a look at a basic database-driven e-commerce site built with the following
Web pages:
ƒ
Page
menu.html
Primary Purpose
Product navigation
detail.html
Product detail
cart.html
Shopping cart
checkout.html
Checkout
thanks.html
Confirmation
Most advanced e-commerce Web sites use these five kinds of Web pages so this section gives you
an introduction to the general layout of these kinds of Web pages. The features and makeup of these
pages will be different depending on the site being designed but their primary purpose will be the
same. In this section you build minimal versions of each of these five Web pages. Later, in the
"Advanced Customization" section of this chapter you go into more detail regarding the possible
features that you can add to each of these pages.
The minimal versions of these pages, described next, assumes a product database called
products with the fields prod_title, prod_price, prod_desc, and prod_image.
A Three-Line Product Navigation Page
A simple product navigation can be one that lists products on the screen. When users click a product,
a page provides them detail on the product. You can build a page to do this by starting with the twoline database report discussed in Chapter 12, Building Query Pages. Here's the page:
<html>
<title>Three-Line Product Navigation Page</title>
<b>My Database-Driven Store</b><br>
Product Menu:<ul>
<table border=0>
<<
myresults=DBFIND("products","",1,50,"record,prod_title,prod_pric
e")
DISPLAY
LAYOUT(myresults,"<tr>",
"<td><a href=detail.html name=rec
value=",[1],">",
[2],"</a></td>",
"<td>",[3],"</td>",
"</tr>")
/DISPLAY
>>
-195-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</table>
</ul>
</html>
Changes are shown in bold. You add a DBFIND search to the page to extract the products. You
convert the entries in the first column of the table to hypertext links. When a user clicks a link, the
rec variable is set to the record ID of the product and the user is directed to detail.html. This is
the same technique used in the "Linking Reports to Detail Page" section at the end of Chapter 13,
Building Database Reports.
A Six-Line Product Detail Page
A simple product detail page is one that displays information on a selected product along with an
Add To Cart button to add the item to a shopping cart. Here is a page that displays the details
available on the selected product with a hypertext link to add the item to the cart:
<<temp=DBGETREC("products",rec)>>
<html>
<title>Six-Line Product Detail Page </title>
<b>My Database-Driven Store</b><br>
Product Detail:<br>
<a href=menu.html>Return to Menu</a><ul>
<b><<prod_title>></b><br>
<table border=0>
<tr>
<td valign=top>
Price: <<prod_price>></td>
<td valign=top><img src=/images/<<prod_image>>></td>
</tr>
<tr><td colspan=2>
<a href=addtocart name=rec value=<<rec>>>ADD TO
CART</a></td></tr>
</table>
</ul>
</html>
<<overlay addtocart
temp=DBGETREC("products",rec)
APPEND ROW(rec,prod_title,1,prod_price) TO mycart /APPEND
GOTO "cart.html"
>>
The first Overlay on the page uses DBGETREC to load variables from a record in the products
database using the record rec, which is set when the user clicks a product link on the menu.html
-196-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
page. The variables loaded from the products record are displayed throughout the page. Then, at
the bottom of the page, a hypertext link enables users to add the item to a shopping cart. The
shopping cart is really just a variable, called mycart, with four columns and multiple rows—one row
per item in the cart. The first column contains the record ID of a product. The second has the product
name; the third contains the quantity ordered, and the last column contains the unit price of the item
ordered. In more sophisticated carts, this variable has more columns, but the idea is the same.
When a user clicks the Add To Cart link, the database record loads again using the DBGETREC tag to
ensure the variables associated with the record match those containing field values. Then the
APPEND tag adds a row to mycart. Note how the ROW tag is used as the first parameter of the
APPEND tag. This is a good technique for appending multicolumn rows to multicolumn, multirow
variables like mycart. Also note that mycart, like all HTML/OS variables, starts with the value
ERROR. After a row is appended you still have a row in mycart containing the value ERROR. Here
that bad row is left in mycart. It is stripped out of mycart on the shopping cart page. After the
APPEND instruction, the user is directed to cart.html to view the contents of the shopping cart.
Note that this detail page is similar to the detail page described in the "Linking Reports to Detail
Pages" section at the end of Chapter 13, Building Database Reports. The main difference is that
here you use a DBGETREC tag instead of a DBGET tag, because here a record ID determines the
record instead of a company_name field.
A Twenty-Line Shopping Cart
The previous product detail page used an Add To Cart link. It assumed a shopping cart variable
called mycart with four columns. Here is a simple shopping cart page that lists the contents of that
mycart variable. This page also includes Delete links to remove items from the cart, a State Tax
option to include or not include sales tax, and a Shipping Formula to calculate shipping costs. The
code for the page is as follows:
<< # Structure of mycart:
# column 1 - record ID
# column 2 - product name
# column 3 - quantity
# column 4 - unit price
/#
mycart=GETCOLNOTEQ(mycart,1,"ERROR")
mycart=GETCOLNOTEQ(mycart,1,"")
IF mycart[1,1]="" THEN GOTO "nocart.html" /IF
IF ISNUMBER(tax)="FALSE" THEN tax=0 /IF
shipping=MAX(15,(3+3*rows(mycart)))
>>
<html>
<title>Twenty-Line Shopping Cart Page </title>
<b>My Database-Driven Store</b><br>
Shopping Cart:<br>
<a href=menu.html>Return to Menu</a><ul>
<form method=post ACTION=changecart>
<table border=1><tr>
<td>Product</td><td>Unit
-197-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Cost</td><td>Quantity</td><td>Subtotal</td>
</tr>
<< i=1 subtotal=0
FOR NAME=mycart ROWNAME=x DO
DISPLAY
"<tr><td><a href=detail.html name=rec VALUE="+x[1]+">"+
x[2]+"</a> (<a href=dcart NAME=drec
VALUE="+x[1]+">Delete</a>)"+
"</td><td>"+x[4]+"</td>"+
"<td><input type=text name=mycart[3,"+i+"]
size=4></td>"+
"<td
align=right>"+FORMAT((x[3]*x[4]),"comma")+"</td></tr>"+LF
/DISPLAY
subtotal=subtotal+(x[3]*x[4])
i=i+1
/FOR
>>
<tr><td colspan=3>Subtotal</td>
<td align=right><<FORMAT(subtotal,"comma")>></td></tr>
<tr><td colspan=3>State Sales Tax:
<a href=<<page>> NAME=tax value="0.0625">Yes</a>
<a href=<<page>> NAME=tax value="0.0000">No</a>
</td><td align=right><<FORMAT(tax*subtotal,"comma")>></td></tr>
<tr><td colspan=3>Delivery:
</td><td align=right><<FORMAT(shipping,"comma")>></td></tr>
<tr><td colspan=3>TOTAL</td>
<td
align=right><<FORMAT(subtotal+tax*subtotal+shipping,"comma")>></
td></tr>
<tr><td colspan=4 align=right>
<input type=submit name=mybutton value="Recalculate">
<input type=submit name=mybutton value="Checkout">
</td></tr>
</table>
</form>
</html>
<<overlay changecart
i=1 subtotal=0
WHILE mycart[1,i] != "" DO
-198-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF ISINTEGER(mycart[3,i])!= "TRUE" OR
mycart[3,i] =0 OR mycart[3,i] < 0 THEN
mycart[1,i]=""
ELSE
subtotal=subtotal+mycart[3,i]*mycart[4,i]
/IF
i=i+1
/WHILE
mycart=GETCOLNOTEQ(mycart,1,"")
shipping=MAX(15,(3+3*rows(mycart)))
IF mybutton="Checkout" THEN
GOTO "checkout.html"
ELSE
GOTO PAGE
/IF
>>
<<overlay dcart
mycart=GETCOLNOTEQ(mycart,1,drec)
GOTO PAGE
>>
The top of this page contains an Underlay that serves three functions. First it clears empty rows and
the ERROR in the first cell of mycart (as mentioned in the previous section). Then it tests to see
whether mycart is empty. If mycart is empty, it directs the user to nocart.html. After that the
Underlay initializes the tax variable. Last it runs the following line:
shipping=MIN(15,(3+3*rows(mycart)))
You can change this shipping formula to fit your needs. This particular formula sets the shipping cost
to $3.00 plus $3.00 for each item in the cart, with a maximum shipping cost of $15.00. This sets the
shipping cost between $6.00 and $15.00. In more advanced e-commerce systems, multiple shipping
options are often provided along with more complex calculations. In the most advanced systems,
costs are calculated based on the Zip codes of the organization and destination and data such as
total shipping weight is taken into account.
After the Underlay at the top of the page, an HTML table is displayed. First an HTML table header is
displayed. After that an HTML table row is displayed for each product in the cart. One row for state
tax and a row for shipping follow. The last row displays a total. The part of the Overlay that displays
the items in the cart uses the following code:
DISPLAY
"<tr><td><a href=detail.html name=rec VALUE="+x[1]+">"+
x[2]+"</a> (<a href=dcart NAME=drec
VALUE="+x[1]+">Delete</a>)"+
"</td><td>"+x[4]+"</td>"+
-199-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
"<td><input type=text name=mycart[3,"+i+"] size=4></td>"+
"<td
align=right>"+FORMAT((x[3]*x[4]),"comma")+"</td></tr>"+LF
/DISPLAY
This code is inside a FOR loop. The first part of the DISPLAY statement displays a hypertext link to
detail.html. Note the name-value pair that sets rec to the record ID. The text in the link is the
product name.
This is followed with a similar link that goes to dcart, an on-click Overlay for deleting a product from
mycart, which is followed by a unit cost, x[4].
After this you fill an HTML table cell with an HTML form input text box by using the following code:
"<td><input type=text name=mycart[3,"+i+"] size=4></td>"
Note how the name of the input box is dynamically generated. When you are writing input text boxes
directly in a page, you can write for example <input type=text name=mycart[3,4]>, which
places an input text box on the screen for the name mycart[3,4]. When the user submits the
HTML form, the value in this box is saved into mycart[3,4]. Here however, this HTML statement
is generated within a DISPLAY statement, so you paste together the line as shown.
To the right of this quantity input, a subtotal is displayed. To ensure numbers are displayed properly
you right-align them in the cells of the HTML table, and you use the FORMAT tag to display numbers
with two decimal places.
Below the product rows of the HTML table are rows for setting and displaying the state tax, the
delivery charge, a subtotal, and a total.
Below this, you provide Recalculate and Checkout buttons. When a user clicks Recalculate, the
changecart on-click Overlay runs. This Overlay checks the quantities placed in the cart and
recalculates totals. The code uses a WHILE loop that checks each cell possibly modified by the user.
If a quantity value is not a positive integer then the first cell in that product row changes to an empty
string. Then below the WHILE loop, using a GETCOLNOTEQ tag, deletes all rows with an empty string
in the first column. New totals and shipping costs are calculated and depending on whether the onclick Overlay was run by the user clicking Recalculate or Checkout, the page is redisplayed or the
user is directed to checkout.html.
Below the changecart on-click Overlay is the dcart on-click Overlay, which runs when the user
clicks the Delete link next to a product. This on-click Overlay uses the GETCOLNOTEQ tag to delete
the row in mycart with drec in the first column and redisplays the page.
A Nine-Line Checkout Page
When the Checkout button on the previous Shopping Cart page is clicked the user is sent to a page
where the user can complete the order. The checkout process typically does the following:
ƒ
ƒ
Collects delivery information
Collects payment information
-200-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
ƒ
Sends order to company for fulfillment
Backs up order
You can build a simple page that performs these tasks by starting with the page discussed in the
"Ten-Line E-Commerce Web Site" section at the beginning of this chapter. Deleting the unnecessary
parts of the page (with new or modified elements in bold), you get the following code:
<html>
<title>Checkout Page</title>
<b>My Database-Driven Store</b><br>
Checkout Form:<br>
<a href=menu.html>Return to Menu</a><ul>
<< IF msg != "ERROR" THEN DISPLAY msg+"<br>" /DISPLAY /IF
msg="ERROR"
total=subtotal + tax x subtotal + shipping
>>
<form method=post ACTION=postorder>
<table bgcolor=#CECECE border=0 cellspacing=0>
AM
FL
Y
<tr><td colspan=2><b>Checkout Form</b></td></tr>
<tr><td align=right valign=top>Full Name</td>
<td><input type=text name=shop_name size=30></td></tr>
<tr><td align=right valign=top>E-Mail</td>
<td><input type=text name=shop_email size=30></td></tr>
<tr><td align=right valin=top>Shipping<br>
TE
<font size=1>Enter complete<br>shipping address</font></td>
<td><textarea name=shop_ship cols=30
rows=3></textarea></td></tr>
<tr><td align=right valign=top >Credit Card</td>
<td valign=top><input type=text name=shop_card></td></tr>
<tr><td align=right valign=top >Expires</td>
<td><input type=text name=shop_expires><font
size=1>(MM/DD/YYYY)</font>
</td></tr>
<tr bgcolor=#FFFFFF><td colspan=2 align=center>
<i>A total of <<FORMAT(total,"comma")>>(
<<FORMAT(subtotal,"comma")>> + <<FORMAT(shipping,"comma")>>
S&H)
will appear on your credit card.</i></td></tr>
<tr bgcolor=#FFFFFF><td colspan=2 align=center>
<input type=submit value="COMPLETE ORDER NOW">
</td></tr></table></form>
-201Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</html>
<<overlay postorder
IF shop_name="" OR shop_email="" OR
shop_ship="" OR shop_card="" OR
shop_expires="" THEN
msg="<font color=red>Missing Info. Try Again.</font>"
ELSE
myorder=
"============================"+LF+
"Order Date: "+TODAY+LF+
"Full Name: "+ship_name+LF+
"E-mail: "+ship_name+LF+
"Card: "+ship_card+" ("+ship_expires+")"+LF+
"Shipping: "+LF+ship_name+LF+
"ITEMS ORDERED:"+LF
FOR NAME=mycart ROWNAME=x DO
myorder=myorder+
" Qty "+x[3]+" of "+x[1]+": "+FORMAT(x[3]*x[4],"comma")+LF
/FOR
myorder=myorder+
"SUBTOTAL: "+ FORMAT(subtotal,"comma")+LF+
"Shipping: "+ FORMAT(shipping,"comma")+LF+
"TOTAL: "+ FORMAT(total,"comma")+LF+
"============================"+LF+LF
MAIL myorder TO ADDRESS="[email protected]"
SUBJECT="Order ("+now+")"
/MAIL
APPEND myorder To FILE="order_log.txt" /APPEND
msg="Order Placed. Thank you. E-mail questions to
[email protected]"
shop_card=""
shop_expires=""
shop_ship=""
/IF
GOTO "thanks.html"
>>
This code collects contact and shipping information, and credit card information. When the user
clicks the Complete Order button, the postorder on-click Overlay runs. It uses an IF-THEN
statement to check the submitted information. If the data is validated, the code composes the text for
the order, placing it in the myorder variable. It then e-mails myorder to the owner of the site, backs
-202-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
up the myorder to a file using the APPEND tag, and then directs the user to the thank you page,
thanks.html. Note that more sophisticated pages would also send a confirmation e-mail to the
user of the site. This is left as an exercise at the end of this chapter.
A One-Line Confirmation Page
The thank you or confirmation page is needed so the user knows that the order has been processed.
It is best to separate the page from the checkout page to ensure the user gets the feedback that the
order was successfully placed. A simple confirmation page is as follows:
<< mycart="" >>
<html>
<title>Thank You Page</title>
<b>My Database-Driven Store</b><br>
<a href=menu.html>Return to Menu</a>
<ul><br><br>
<font size=2>Your order has been processed.</font><br>
<font size=5>Thank you.</font><br><br>
If you have questions please call 1-212-555-6789<br><br>
- Staff</ul>
</html>
The only HTML/OS instruction used on this page is the one at the top that sets mycart to the empty
string. The rest of the page is a simple document that displays a Thank you message for the user.
Additional helpful features can be added to pages like this. These are just some of the advanced
features covered in the next section.
Advanced E-Commerce Sites
In the previous section you looked at the construction of a database-driven e-commerce site. The
site was composed of five pages, each with a different purpose. The navigation page was used to
select a product. The product detail page provided detail on selected products. The shopping cart
page informed the users of their selections, and a checkout page processed the order. The
confirmation page provided information to the users after order placement.
Highly sophisticated e-commerce sites provide additional pages of use by those running the site.
They also improve on the pages discussed in this chapter, often placing multiple features on the
same page. Web sites, such as Puritan's Pride (http://www.puritanspride.com/) in Figure 16.2,
place multiple navigation tools and product detail on the same page.
-203-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 16.2: Sophisticated E-Commerce Sites Merge Navigation and Product Information Together on
the Same Page.
In this section you look at the main components of more sophisticated e-commerce sites and the
kinds of features you may want to add to your site. This section can serve as a source of ideas. In
the remaining chapters of Part IV, E-Commerce Programming, you learn how to build many of the
features discussed in this section.
The Main System Components
When building an advanced e-commerce site you work with three main system components: Web
pages, databases, and settings data. Web pages provide presentation and functionality but depend
on underlying data. You store the underlying data in databases and settings files. Until now, the only
database discussed in this chapter was a product database and no settings files were used. A
sophisticated site uses multiple databases and settings files to store configuration information.
Site Databases and Settings Files
Sophisticated e-commerce sites often include the following databases:
ƒ
ƒ
ƒ
ƒ
Product database
Order database
Order Items database
User database
The Product database stores information on each product. You expand the fields in each record to
include any and all information directly associated with each product. In addition to the basic fields,
such as product name and price, fields for product options, related products, special pricing options,
and special shipping surcharges are stored here.
The Order database is where you store orders the user has placed. It gives you an archive of orders.
You save customer information, shipping information, the date the order was placed, and other
information pertaining to the order here. The Order database goes along with an Order Items
database, which stores the specifics of each line item ordered. If a user places an order containing
four products, for example, a record is added to the order database and four other records, one for
each item ordered, are added to the Order Items database. The Order Items database record would
include a field that stores a unique order number, relating the records back to its associated record in
the order database.
-204-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The User database is where you store information on specific users. You usually store fields for a
login ID, password, and an e-mail address in this database. Other fields may store a pricing level
(retail or wholesale, for example), the date of the user's last order, the total number of orders placed,
and so on.
Settings files are standard text files used by the HTML documents on the Web site. These text files
make it easier to manage a Web site. For example, you can use settings files to store a table of
shipping options in a pull-down menu for providing users with shipping options. They can contain a
cross-reference between two-letter U.S. state codes, and the full names of states, again in a pulldown menu format. Setting files can be e-mail templates—files that store the content of messages
that are automatically e-mailed to users when they order a product or when staff members process
their order.
The Web Pages
Web pages are the presentation and dynamic elements of a site. In the previous section you
familiarized yourself with five kinds of Web pages. In addition to them, sophisticated sites may have
Web pages for customer-order tracking, providing receipts, printing purchase orders, and providing
staff functions such as product record editing, determining which products are on sale or should be
highlighted, issuing sales reports, and more. In the following section you look at the many of the
features possible when building a sophisticated Web site.
Advanced Customization
Sophisticated e-commerce sites serve as interfaces between organizations and their customers, staff,
and visitors. They reflect the individuality of an organization. No two e-commerce sites are they same.
The features and Web pages need to be changed and updated as operations and organizational
policies change. An advanced e-commerce site requires constant attention, improvement, and
customization.
When building a new e-commerce site, it is important to realize you cannot build it over night. A lot of
thought is needed—the same kind of thought needed when making everyday organizational
decisions. You must carefully select the ideas and features that go into a site from hundreds of
possibilities.
With so many possible designs and features, a good starting point when designing an e-commerce
site, is to select those features or design elements you want on each page of the e-commerce part of
the site. Narrowing down the features and designs you want can prevent you from skipping those
that are truly important to the site. The following is a general wish list you can use for this purpose.
Features are organized by the kind of Web pages typically found in the e-commerce section of a
Web site. We do not list those Web site features that are purely informational or those disconnected
from transactions associated with the buying of products or services.
Product Navigation Pages
ƒ
Dynamic categories. Category lists that display only those categories containing
products.
ƒ
Image grid. Products that display a multicolumn table of images.
ƒ
Search box. A text box component that yields a search result page.
ƒ
Multiple categories. A category search that attaches to a subcategory search.
ƒ
Multiselection feature. Product lists that allow the user to select multiple products
at a time.
ƒ
Prioritization. Product lists sorted by a priority key so that certain products always
appear at the top.
-205-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
Add to cart. Add To Cart buttons that appear on navigation pages.
Product Detail Pages
ƒ
Variable pricing. Automatic pricing based on user type.
ƒ
Options. Product options, such as color of a sweatshirt, applied by users when
they add the item to the shopping cart. Options are factors that don't change the
price of a product.
ƒ
Add-ons. Add-ons to a product, such as 50X zoom lens, applied by users when
they add the item to the shopping cart. Add-ons change the price of the purchase.
ƒ
Product collections. Detail pages that display series of products, all with a
common description. Each product has its own description and an Add To Cart
button.
ƒ
Real-time inventory. Detail pages that test the availability of an item, notifying the
user if the item is unavailable.
ƒ
Related items. Lists of products related to the one displayed, typically displayed at
the bottom of a detail page.
ƒ
Extra detail. Extra detail related to the current detail displayed on the page,
typically displayed at the bottom of the page or as a link to a pop-up page.
ƒ
Zoom and related photos. Links that allow the user to zoom into graphics
elements already on the page.
ƒ
Sales announcements. Special messaging on a page containing promotional
content relating to the product being viewed. A message may include information
on possible savings if minimum quantities are ordered, or if orders are placed by a
specific date, for example.
Shopping Cart Pages
ƒ
Related products. Related products that display along side the cart when certain
items appear in the shopping cart. For example, if a toy is found in the cart, a link
for purchasing batteries is provided.
ƒ
Specials and promotions. Specials and promotions that display on the page. For
example, if free shipping is provided on orders over $50 then a cart with $35 worth
of merchandise in the cart might display: "Free shipping on orders over $50. Add a
few more items to your order and save."
ƒ
Printable purchase order. A printable purchase order for use by business
customers.
ƒ
Cookied carts. Shopping cart entries that are saved so that, even if they are
abandoned, upon returning to the site days or weeks later, the items remain in the
user's cart.
ƒ
Reminder service. An option to permit users to compose a reminder e-mail with
links to re-order specific items in their cart or remind them of specific holidays.
Checkout Pages
ƒ
Login. A page with a Forgot your password? feature so users can re-use
information previously entered, such as shipping and billing information, in the
checkout process.
ƒ
Real-time authorization. A feature that allows credit card or checks to be issued
over the Web. When a user places an order, a connection is made to a bank,
completing a bank transaction on the fly, without the need to process the order at a
later date.
-206-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
Automated download feature. A product download button for sites that sell
software or information.
ƒ
Shipping calculators. Useful when building sites that give users multiple shipping
options. For example, sites that give the user the ability to choose among Ground,
Next Day, and Two-Day delivery options.
ƒ
Address book. An address book entry that avoids requiring the user to type in
shipping addresses on repeat orders.
Confirmation Pages
ƒ
Receipt page. A page that provides the user a printable receipt of the order.
ƒ
Reminder notification. A page that gives users the ability to send themselves a
reminder to buy another product. The user specifies a date, and the site sends
them a reminder on that date with the message they specified.
Other Pages
ƒ
Order tracking. Pages that allow customers to see whether the products they
ordered shipped out. May include a FedEx, UPS, or U.S Postal Service tracking
numbers and links.
ƒ
Customer order history. A page where customers can view past orders. Orders
can be summarized and include an Add To Cart button for easy re-ordering.
ƒ
Order pickup. A page for staff to pickup and process orders.
ƒ
Database editors. Record editors for modifying product and user information.
ƒ
Settings pages. Pages for setting categories, products on sales, editing e-mail
templates, and setting configuration parameters.
ƒ
Reporting. Reports on page access, failed user searches, the use of sales
promotions, and so on.
ƒ
Coupons. A feature that provides users the ability to manage a coupon database.
Goes along with a Coupon entry box placed on a checkout page.
This list includes many of the features you have grown to expect from the world's largest and most
sophisticated e-commerce sites. You will want to add many of them to your e-commerce site. In
addition you will want to add features specific to your industry and drop those features not applicable
to your situation. Finally, since this list is long, you will want to prioritize the list to meet objective time
and delivery goals.
Summary
This chapter introduced the world of e-commerce sites. You learned that building advanced
database-driven Web sites is not rocket science—that such sites can be built using five types of Web
pages, each requiring on average, about a dozen lines of HTML/OS instructions. You also learned
that those pages can vary greatly because of the variety of features you can add to them.
In practice, the individuality you give your site will reflect the individuality of your organization.
Construction will take a lot longer than the time it takes to cut and paste the pages here into a site
and get them to work together. You will want to pay careful attention to how each is constructed and
vary them to fit your individual needs. Furthermore, you will want to experiment with different
features and designs.
-207-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
In the next five chapters of Part IV, E-Commerce Programming, you'll learn how to build many of the
features listed in the "Advanced Customization" section of this chapter. Next, in Chapter 17, Building
Product Navigation Pages, you learn how different navigation pages are built.
Exercises
In the following exercises you extend the capabilities of the ten-line e-commerce Web site provided
near the beginning of this chapter. Answers to these exercises are provided on this book's
companion Web site as described in the book's Preface.
Exercise 1
The "Ten-Line E-commerce Web Site" section of this chapter provides the code for a single page ecommerce site. The page avoids tax calculation by saying tax is included. Expand this page so it
supports tax calculation.
Exercise 2
The ten-line e-commerce site does not do sophisticated data validation. Use the data validation
techniques described in the "Commonly Used Validation Schemes" section of Chapter 10, HTML
Forms Processing to accomplish this goal.
Exercise 3
The ten-line e-commerce site does not e-mail a confirmation to the user when a purchase is made.
Add this feature to the page.
-208-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 17: Building Product Navigation Pages
Overview
Chapter 16, Designing E-Commerce Systems introduced a three-line navigation page in the "Your
First Database-Driven E-Commerce Site" section. In this chapter you explore the different ways to
construct navigation pages. An emphasis is placed on e-commerce sites but you can also apply what
you learn here to the needs of information databases, portal Web sites, and article archives. The
concepts are universal.
You start by exploring the kinds of features typically included on a navigation page and how to build
them. Then you build an image-based navigation page. After that you learn about navigation pages
that allow users to select multiple products at a time. This is followed with discussions on category
lists, hierarchical trees, and search boxes.
Use Navigation Pages Only When Necessary—Product navigation pages are not required on
sites with few products. Do not introduce navigation components unless they are useful to the user
visiting the site. If you have only a few products it is doubtful the user needs navigation aids. If you
plan on having many products don't introduce the navigation elements until they are needed.
Selecting Navigation Features
One of the most important considerations when building an e-commerce site is how to make
products on the Web site easy to locate. The process must be intuitive and natural. The amount of
energy and thought required on the part of a user to find a product should be minimized. The best
navigation pages are single uncluttered Web pages that allow the users to find what they are looking
for. Well-designed navigation pages include any combination of the following four navigation
components:
ƒ Products capsules
ƒ Product lists
ƒ Category lists
ƒ Search boxes
The way you arrange these different components depends on the number of products you have and
the ability you have to support navigation components. Sites may need only one of these
components. Other sites will give the user many options, such as the King Arthur Flour Web site
(http://www.kingarthurflour.com/) shown in Figure 17.1. The Web page shown includes two category
lists and a sophisticated search box. Let's look at each component and see how and when it should
be used.
-209-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 17.1: The Recipe Book on this Web Site Uses Multiple Navigation Tools to Ensure Recipes are
Easy to Find.
Product capsules are mini product detail pages that point to pages with greater detail. A product
capsule can include a small graphic, a one-line product title, and perhaps a one-line description. If
the product is a commodity that is well recognized by the user, the capsule can include an Add To
Cart link. If not, you should include a link to a detail page in the capsule.
Product capsules are a great way to give selected products in your inventory preferred visibility,
however they interfere with the navigation aspect of the page. Only a limited number of product
capsules can be placed on a navigation page before the page becomes too busy. Also, product
capsules need to be managed. They should change at least daily. You can do this automatically by
randomly selecting products, when the page is launched, from a list of products reserved for placing
on the product navigation page, or you can do this manually by providing a back-end tool for
selecting the product capsules you want on the page.
Product lists are typically lists of single-line product names that link to product detail pages. On a
Web site with less than 20 products the navigation page is often a product list. Other components
are not necessary. When you have so many products that you can split them into distinct categories,
product lists are no longer effective navigation aids; replace them with a list of categories. When the
number of products grows further, perhaps to thousands, product lists become necessary once
again—but not as a primary navigation tool. You use such lists as alternatives to product capsules.
Use them like product capsules—as a way to focus special attention on specific products.
Category lists are typically lists of category names presented on a navigation page that link to a list
of products. Product categorization is useful when you have more than 20 products. If you have
thousands of products, a category list may have different levels. First the user selects a main
category. Then the user selects a subcategory, and so on. The depth of the categories depends on
the number of products you have.
Search boxes are most often input boxes with a Find button for searching the inventory of a Web site.
The best search boxes are simple, requiring the user simply to place one or more words in the box to
perform a search and yield a product list of results. Search boxes are useful on sites with hundreds
or thousands of products. In most cases they are not useful when you have fewer products because,
in those cases, product categorization works better—always yielding a search result for the user.
The exception to this rule is the site that features complex products that cannot be described by their
product name. For example, a site that sells about one hundred works of art may provide a search
that allows the user to specify a topic, or an artist name.
-210-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Perhaps the most important aspect of search boxes is not what the user sees, but how you actually
process the search request. The best searches are intelligent—meaning you create an algorithm that
searches your products and yields results in most circumstances, placing the most appropriate
search result at the top of the page. This is easier said than done and highly dependent on the
nature of the data being searched. However, you do get some tips on how to build intelligent
searches in the "Building Search Boxes" section later in this chapter.
Building Product Lists
The "A Database-Driven E-Commerce Site" section of Chapter 16, Designing E-Commerce Systems,
introduced a three-line navigation page. The page displays a simple list of products. Here you
display a product list consisting of images, lists that are prioritized, lists that display specials, lists
with Add To Cart links next to each item in the list, and lists where the users can select more than a
single item. You can also extend the examples provided here with Page Up and Page Down features.
To do that, use the technique described in the "Seven-Line Database Report" section of Chapter 13,
Building Database Reports.
A Twelve-Line Image Selection Page
AM
FL
Y
To display a list of images, you could take the three-line navigation page and substitute product
names with images. In doing so however, the images would appear in a single column. More often
you want to display the images in multiple columns across the page. To do this, you convert the
LAYOUT tag used in the navigation page with a WHILE loop. The code follows:
<title>Twelve-line Image Navigation Page </title>
<b>My Database-Driven Store</b><br>
Image Menu:<ul>
<table border=0>
<<
e")
i=1 j=1 no_cols=3
TE
x=DBFIND("products",1,50,"record,prod_title,prod_price,prod_imag
WHILE myresults[1,I] != "" DO
IF j=1 THEN DISPLAY "<tr>" /DISPLAY /IF
DISPLAY
'<td><a href=detail.html name=rec
value="'+myresults[1, i]+'">'+
'<img border=0 src="/images/'+myresults[4]+
'</a><br>'+
x[2]+ '</td>'
/DISPLAY
IF j=no_cols THEN DISPLAY "</tr>" /DISPLAY /IF
i=i+1 j=j+1
IF j > no_cols THEN j=1 /IF
/WHILE
IF j != 1 THEN DISPLAY "</tr>" /DISPLAY /IF
-211Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
>>
</table>
</ul>
</html>
This page displays images, each with a product title underneath it, three columns wide. The number
of columns is adjustable. The trick here is to use three different DISPLAY statements in a WHILE
loop: one to display the HTML <tr> tag when needed, one to display the content of each cell, and
one to display the HTML </tr> tag, when needed.
Sites such the Schweitzer Linen (http://www.schweitzer-lenen.com) shown in Figure 17.2, use
image selection. Here images are more helpful navigation tools than words.
Figure 17.2: This Web Site Uses Images as a Navigation Tool.
A Six-Line Product List with Product Specials
When displaying products, it is often useful to give increased visibility to certain products. Two ways
to do this are by placing select products in bold and placing certain products at the top of the list. To
place certain products at the top of the list, you sort the products returned from DBFIND using a sort
key. Alternatively, you can use the DBFINDSORT tag. To emphasize items on sale, you display them
in bold. Both cases are accomplished using special fields in the database. The prod_key field
creates sort prioritization. The prod_onsale field determines the products to display in bold. Here's
the code to do that:
<html>
<title>Six-Line Product List With Specials</title>
<b>My Database-Driven Store</b><br>
Product Menu:<ul>
<table border=0>
<< myresults=DBFIND("products",1,50,
"record,prod_title,sort_key,prod_onsale")
myresults=SORTCOL(myresults,3)
FOR NAME=myresults ROWNAME=x DO
-212-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF x[4]="TRUE" THEN b[1]="<b>" b[2]="</b>" ELSE b="" /IF
DISPLAY "<tr><td>"+b[1]+
"<a href=detail.html name=rec value="+x[1]+">"+
x[2]+"</a>"+b[2]+"</td><td>"+x[3]+"</td></tr>"
/DISPLAY
/FOR
>>
</table>
</ul>
</html>
The first line in bold sorts the results of the DBFIND tag. The second bolded line sets up b, a twocolumn by one-row variable. If the line belongs to an item on sale, the two cells in the b variable are
filled with the <b> and </b> HTML bold tags. If not, b is set to the empty string. As a result, products
on sale appear bold while others remain unchanged. The technique used here is an example of how
you add fields to a database to give you greater control over the display of the products. In general,
product emphasis requires the addition of special fields in the product database. How you use those
fields when you display the products is up to you. The same fields, for example, can be used to
display onsale images next to items on sale.
A Seven-Line Multiselect Product List
Product navigation is about finding and locating a product or service. In some cases however, the
concept is also about giving the user the ability to select multiple products and view them side by
side on a detail page. This requires you give the user the ability to select multiple products at once
from a menu. At other times you may want to allow the user to select multiple products at once and
add them to a shopping cart.
In both of these cases you need to give users the ability to select multiple products from a list. To do
this, you use check boxes. Instead of providing a list of hypertext links, you provide the user
Compare and Add To Cart buttons, and check boxes next to each item. Because a multiselect
feature is more often an add-on to a standard product list, the links used to select an individual
product are best retained. The code for this is as follows:
<html>
<title>Seven-Line Multi-Select Product List </title>
<b>My Database-Driven Store</b><br>
Product Menu:<ul>
<form method=post ACTION=redirect>
<input type=submit name=mybutton value="Compare">
<input type=submit name=mybutton value="Add To Cart"><br>
<table border=0>
<< myresults=DBFIND("products",1,50, "record,prod_title")
FOR NAME=myresults ROWNAME=x DO
DISPLAY '<tr>'+
'<td><input type=checkbox name=myrecords
-213-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
value="'+x[1]+'"></td>'+
"<td><a href=detail.html name=rec value="+x[1]+">"+
x[2]+"</a></td></tr>"
/DISPLAY
/FOR
>>
</table>
</form>
</ul>
</html>
<<overlay redirect
IF mybutton="Compare" THEN
GOTO "mdisplay.html"
ELSE
GOTO "cart.html"
/IF
>>
This Web page starts with an HTML form linked to the redirect on-click Overlay, which,
depending on the button the user clicks, redirects the user to mdisplay.html or cart.html. This
Web page is similar to the page discussed in the previous "A Six-line Product List with Product
Specials" section except it includes an HTML form with check boxes and two submit buttons.
This page uses multiple check boxes with the myrecords name. It is important to remember that,
when the HTML form is submitted, this myrecords name will be a single column, multirow variable
filled with record numbers—one for each selection. See the section "Using Checkbox" in Chapter 10,
HTML Forms Processing for more information on how this HTML form component works.
After the HTML form is submitted, the user is directed to either mdisplay.html or cart.html.
The purpose of this page is to create and fill the myrecords variable. How the myrecords variable
is handled is up to the code placed in mdisplay.html and cart.html. The code to do that is left
as an exercise at the end of this chapter.
Building Category Lists
Category lists can be written manually, read from a text file, or composed on the fly from the
products database. In this section you learn all three methods. At the end of this section you also
look at category lists that have depth—meaning, categories with subcategories with subcategories,
etc.
A Zero-Line Category List
First, consider building a category list manually. Here you add a field to the products database called
prod_cat that stores the name of a category. Then you manually type into the HTML document a
list of hypertext links, one for each category. You use no Overlays, producing a zero-line category list
as shown in the following example:
<html>
<title>Zero-Line Category List </title>
-214-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<b>My Database-Driven Store</b><br>
Categories:<ul>
<a href=list.html name=mycat value="tshirts">T-Shirts</a><br>
<a href=list.html name=mycat value="pants">Pants</a><br>
<a href=list.html name=mycat value="shorts">Shorts</a><br>
</ul>
</html>
Here each link, when clicked, sets the mycat variable and directs the user to list.html, a product
listing page. That page lists the products in the selected category. To do that, you replace the
DBFIND tag on any product listing page with one that extracts only those products in the selected
category. For example, starting with the three-line navigation page introduced in the "Your First
Database-Driven E-Commerce Site" section of Chapter 16, Designing E-Commerce Systems, you
get the following:
<html>
<title>Three-Line Product List From Category</title>
<b>My Database-Driven Store</b><br>
Category: <b><<mycat>></b><ul>
<table border=0>
<< s='prod_cat="'+mycat+'"'
myresults=DBFIND("products",s,1,50,"record,prod_title,prod_price
")
DISPLAY
LAYOUT(myresults,"<tr>",
"<td><a href=detail.html name=rec value=",[1],">",
[2],"</a></td>",
"<td>",[3],"</td>",
"</tr>")
/DISPLAY
>>
</table>
</ul>
</html>
Changes are shown in bold. A category name is displayed at the top of the page and a search string
appears in the second parameter of the DBFIND tag to limit searches to items in the selected
category.
Editable Categories
The zero-line category list is fine if categories do not change. But on large sites they may change
often. In that case you may want to store the possible categories in a text file and allow staff
-215-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
members to edit the text file, which they can do using an HTML editor, like those discussed in
Chapter 8, Building Text Editors, or using a comma-delimited text file editor like that discussed in the
"Fifteen-Line Spreadsheet Editor" section of Chapter 10, HTML Forms Processing.
The category page reads the category text file and displays the category lines using a layout tag.
Replace the three hypertext links in the zero-line category page with the following Overlay:
<<
COPY FILE="mycats.txt" TS="|" TO mycats /COPY
DISPLAY
LAYOUT(mycats,'<a href=list.html name=mycat value="',[1],
'">',[1],'</a><br>')
/DISPLAY
>>
The Overlay reads the text file into the mycats variable. That variable is used in a LAYOUT tag to
display the category listing lines dynamically. Now staff members have a text file to edit that is
divorced from the Web page. Changing the text file changes the list of categories listed on the Web
page without you having to edit the Web page itself.
Dynamic Categories
Another way to build a category list is by extracting the possible categories from the products
database. This method has the advantage that empty categories are not listed. It is most definitely a
Web faux pas to display empty categories on Web pages. To extract the categories used in a
product database, you use the DBUNIQUE tag. Replace the three hypertext links in the zero-line
category page with the following Overlay:
<<
mycats=DBUNIQUE(products,"","prod_cat")
DISPLAY
LAYOUT(mycats,'<a href=list.html name=mycat value="',[1],
'">',[1],'</a><br>')
/DISPLAY
>>
This piece of code is identical to the previous code used when extracting categories from a text file,
except here, the categories are extracted from the database.
Hierarchical Categories
On large Web sites category lists may lead to other category lists. Products may appear at every
level. Portals and search engines often organize information this way. The Famous UncleWebster
Web site at http://www.unclewebster.com/ does this, for example. See Figure 17.3.
-216-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 17.3: The UncleWebster Web Site Uses a Hierarchical Tree as a Navigational Aid.
Each category leads to more categories that lead to more categories. This is also known as a
hierarchical tree. When coding this kind of drill-down navigation, it is useful to define the location of
the product in the hierarchical tree using a prod_cat field. You use a single field. Define the entry
as a category path, preceded with the depth of the path. For example:
ƒ
4|Macintosh|Monitors|Color|15 Inch
This indicates a category with four levels. The main category is Macintosh. Within this is the category
Monitors, and within that is the category Color, and within that is the category 15 Inch. The category
depth (4) is placed at the beginning of this field as a convenience. It becomes useful when you are
extracting categories of a particular depth—a task you need to perform later on in the programming.
Navigation involves three distinct user aids. First, you provide users a clickable category tree that
tells them their current category. Clicking a category name in the tree allows users to navigate up the
tree. Second, you provide users a list of categories to access. Last, you provide the users a list of
products.
Building the Clickable Tree
The clickable tree that displays on the screen depends on the current category and category depth of
the user. The user starts in the home category at depth 0. You begin the link with a link to this home
directory and follow that with links to each of the subcategories defined in the prod_cat field. The
code is as follows:
<< mycatcol=LISTTOCOL(mycat,"|")
mylevel=mycatcol[1,1]
IF ISINTEGER(mylevel)!="TRUE" THEN mylevel=0 mycatcol=0 /IF
DISPLAY
'<A HREF="'+page+'" NAME=mycat value="0">Home</a>&gt;'
/DISPLAY
i=1
WHILE i <= mylevel DO
-217-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
x=GETTABLE(mycatcol,1,1,2,i+1)
x=i+"|"+COLTOLIST(x,1,"|")
y=mycatcol[1,i+1]
DISPLAY
'<A HREF="'+page+'" NAME=mycat
value="'+x+'">'+y+'</a>&gt;'
/DISPLAY
i=i+1
/WHILE
>>
The LISTOCOL tag converts the vertical bar-separated list into a one-column table. This makes the
data easier to manage. The first cell of this variable is the depth level. If the mycat variable has not
yet been set (the first time on the page) then the level is set to 0. This is followed by the display of a
link to the home category, which is followed by a WHILE loop that displays the links stored in the
mycatcol variable. Each time through the loop a link is displayed. The tricky part of the loop is
correctly setting the value for the mycat variable (x) and the value displayed (y). If, for example, the
current value of mycat is 3|Macintosh|Monitors|Color, you want to set x to 1|Macintosh in
the first link. The second time through the loop, you want to set x to 2|Macintosh|Monitors, and
so on. To do this, you use the GETTABLE tag to extract the first up to the i'th column of mycatcol.
Then you create x by pasting together the category depth, a vertical bar, and a vertical barseparated list. You create the list with the COLTOLIST tag. See Appendix D, HTML/OS Tag
Reference Guide, for details on this Overlay tag. Finally, i is incremented and the loop is repeated.
The final result is a link that looks as follows:
ƒ
Home>Macintosh>Monitors>Color>
A user clicking a link in this navigation component sets mycat to a value like those stored in
prod_cat fields.
Building the Category List
The list of categories in the current category is extracted from the database using the DBUNIQUE tag,
the same tag used earlier in the "Dynamic Categories" section of this chapter. Here however you are
looking for subcategories, not a specific category specified on a previous page. The technique is the
same except that here, if the current category is 2|Macintosh|Monitors, for example, then you
seek all categories beginning with 3|Macintosh|Monitors. The code for this is as follows:
s=REPLACE(mycat,mylevel+"|",(mylevel+1)+"|")
s='prod_cat ~ "'+s+'"'
mycats=DBUNIQUE(products,s,"prod_cat")
IF mycats[1,1] != "" THEN
DISPLAY "Categories:<br>" /DISPLAY
FOR NAME=mycats ROWNAME=x DO
y=CUT(x[1],(mylevel+1)+"|")
y=REPLACEALL(y,"|","&gt;")
-218-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
DISPLAY
'<A HREF="'+page+'" NAME=mycat
value="'+x[1]+'">'+y+'</a><br>;'
/DISPLAY
/FOR
/IF
>>
The first REPLACE tag increments the leading category depth in the mycat variable so a search
string can be composed and used in the following DBUNIQUE tag. The DBUNIQUE tag extracts
unique subcategories, which a FOR loop then displays. Note that for this technique to work at least
one product must be in the database at this level. If you are not placing products at every level, be
sure to place a blank record at each level. Note also how the CUT and REPLACEALL tags convert the
category lists extracted from the database into categories you can display on the screen.
Building the Product List
You use the DBFIND tag to extract the products at the current category level. The code is similar to
that used in the three-line product list from category page, described earlier in this section. The code
is as follows:
y=CUT(x[1],(mylevel+1)+"|")
y=REPLACEALL(y,"|","&gt;")
DISPLAY "Category: <b>"+y+"</b><br>" /DISPLAY
s='prod_cat="'+mycat+'"'
myresults=DBFIND("products",s,1,50,"record,prod_title,prod_price
")
DISPLAY
LAYOUT(myresults,"<tr>",
"<td><a href=detail.html name=rec value=",[1],">",
[2],"</a></td>",
"<td>",[3],"</td>",
"</tr>")
/DISPLAY
Changes are shown in bold. The main change converts the mycat variable into text that can be
displayed. This code, along with the code provided in this section for displaying a navigation tree and
a list of categories are needed when you are building sophisticated hierarchical navigation pages for
use in large e-commerce sites and directory portals.
Building Search Boxes
Search boxes are useful on sites with many products. Often they are placed on pages that also
include product lists and category lists. The Overlays used when building search boxes are the same
-219-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
as those used when building query pages—like those discussed in Chapter 12, Building Query
Pages.
When building search boxes, it is best to keep the search simple—at least for the user. A single input
text box is preferred. The fields actually searched, and the algorithm behind the search, do not need
to be simple however.
At the least, you want the words placed in the search box to be matched against words in one or
more fields of the products database. To do that, set up your search as a group search. Setting up
group searches is discussed in the "Using Group Fields" section of Chapter 12. This is known as an
intelligent search since the multiple words are matched against words in multiple fields. Users are
not required to think about the nature of the words they type into the search box.
A more intelligent search is obtained by automatically changing the search result if no search results
are found, or if too many results are found. We leave this as an exercise at the end of this chapter.
Summary
In this chapter you learned how to build navigation pages. You learned that product navigation is
generally customized to fit the needs of each site. While this chapter provides you with many
sophisticated ways to build navigation pages, this chapter is not, and cannot be, exhaustive. In
general, navigation pages are limited only by your imagination.
You've also seen that navigation elements are often not stand-alone pages. In this book we have
separated the different e-commerce components into different pages for convenience, but in highly
sophisticated sites these components are intermingled.
The main thing to remember when building your site is to provide the user options that fit the needs
of your site. The elements should be intuitive, and lead your users to the products or services they
seek with a minimum of thought, a minimum of hunting around, and as few clicks as possible.
Next, in Chapter 18, Building Product Detail Pages, you learn the different ways to display content on
the page once a user has been able to locate a product. That next chapter, along with this one, gives
you the ability to build a wide variety of navigation and product detail pages—an important part of ecommerce applications.
Exercises
In the following exercises you extend and build additional navigation pages using what you learned
in this chapter. Answers to these exercises are provided on this book's companion Web site as
described in the book's Preface.
Exercise 1
The multiselect product list, which was discussed in the "A Seven-Line Multiselect Product List"
section of this chapter, links to mdisplay.html and cart.html. Set up a products database
and the pages mdisplay.html and cart.html, so that together, the three pages constitute a
functional Web site. To test your creation, fill the database with at least three or four records.
Exercise 2
The discussion in the "Building Search Boxes" section near the end of this chapter recommends
building a search box using the techniques learned in the "Using Group Fields" section of Chapter 12,
Building Query Pages. Add a search box that takes advantage of group fields to the Web page you
built in Exercise 1. In your query, use the =& comparison operator.
-220-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 3
TE
AM
FL
Y
Make the search in Exercise 2 more intelligent by doing the following: If the number of search results
found is less than five, repeat the search using a ~& operator. If after searching within this operator,
you still get too few results, use a ~| operator. Take the results you find with these searches and
combine them together. Display the results found using the first search operator above those found
using the second type of search operator. Now you have a highly intelligent search with results
sorted by relevance.
-221Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 18: Building Product Detail Pages
Overview
Product detail pages can vary a lot in sophistication. Most e-commerce sites simply display the
information about a product along with an Add To Cart link. The page described in the "A Six-Line
Product Detail Page" section of Chapter 16, Designing E-Commerce Systems, is such an example.
As you get more sophisticated however, you learn that these pages can offer a lot more to users.
Extra information can include manufacturer articles, specification sheets, reviews, photos, and links
to related Web pages or Web sites.
You also learn that products are often related to other products. Users interested in one product may
be interested in other products by the same manufacturer, other products (perhaps hot sellers) in the
same category, other products priced similarly to the one they selected, and so on.
In this chapter you look at some of the ways you can expand and improve detail pages. You also
look at how to display products with options or add-ons and how those products work with shopping
carts.
Adding Extra Detail
Extra detail is most often added to Web pages in three ways. You can add links to other Web pages
containing the extra information, add the information in sections at the bottom of the page, or add the
extra detail to the main part of the page. In this section you explore all three of these situations.
The Image Zoom
The saying "a picture is worth a thousand words" holds true on the Web as well as off the Web.
However, large pictures slow down the rendering of Web pages. As a result, it is often a good idea to
limit image photos on detail pages to 15 kilobytes or smaller, and link them to larger photos. This is
known as image zoom. Many e-commerce sites use a standard hypertext link to provide the image
zoom. The HTML instruction looks something like the following line:
<a href=p_01_big.jpg><img border=0 src=p_01.gif></a>
When the user clicks the image on the product detail page, a larger image is displayed. The image
appears in the upper-left corner of the browser. It is displayed without a title. To return to the page,
the user must click the browser's Back button. This is a somewhat crude way to do image zooms.
A better way to do this is to display the larger photo on a Web page, centered, along with a product
title and a link back to the previous page. To do this you link the image in the detail page to a zoom
page. You write the following:
<a href=zoom.html name=rec value=<<rec>>><img border=0
src=p_01.gif></a>
When the user clicks the image, the zoom.html page is displayed. You write that page as follows:
<html>
<title>One-Line Zoom Image Page</title>
<<temp=DBGETREC("products",rec)>>
<html>
-222-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<title>Six-Line Product Detail Page </title>
<table border=0 width=100% height=100%>
<tr valign=center><td align=center>
<table border=1 cellpadding=10 cellspacing=0>
<tr><td><img src=/images/<<prod_image_big>>><br>
<<prod_title>><br>
<a href=detail.html>Continue Shopping</a><br>
</td></tr></table>
</td></tr></table>
</html>
The first and only HTML/OS instruction on this page loads the specified record. You include the
image and title by displaying field values and provide a link so the user can return to the
detail.html page. Using HTML tables lets you center the image on the page.
Related Items
Related items are products of interest to those viewing the currently selected product. Related items
are an example of cross-referencing as explained in the accompanying note, "Cross-Referencing."
They are most often displayed below the product detail. You can store them in the products
database as a list of SKUs (product IDs). Assume a database with a prod_related field for this.
Assume also that each record has a prod_sku field for the product SKU.
Cross-referencing—Displaying related items is an example of cross-referencing. You can use the
same technique used here to link a detail page to other kinds of cross references. For example, a
video database can contain fields with names of actors in the video. Instead of displaying related
products, one displays the names of the actors in the video. Clicking an actor's name can bring up
information on the actor along with other videos featuring the same actor.
When the product loads on the product detail page, you wish to convert the prod_related field
into a list of related products. To do that you must find all the records in the products database with
matching SKUs. One way to do this is to create a group field called prod_skugroup with the
member prod_sku. By doing this, you make it easy to extract a list of SKUs. For example, the
search string for a group field search might be
prod_skugroup = "111,222,333"
That would extract records with SKUs equal to 111, 222, and 333 respectively. See the "Using
Group Fields" section in Chapter 12, Building Query Pages, for an explanation. Using this technique,
the code for displaying a list of related products on a detail page is as follows:
IF prod_related="" THEN
myresults=""
ELSE
s='prod_skugroup =| "' + prod_related + '"'
myresults=DBFIND(products,s,1,10,"record,prod_title")
DISPLAY "Related Products:<br>" /DISPLAY
DISPLAY LAYOUT(myresults,
-223-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
"<A HREF=detail.html NAME=rec VALUE=",[1],">",
[1],"</a><br>"
/DISPLAY
/IF
Here, if the prod_related variable is empty, related products are not listed. If one or more SKUs
are in the prod_related variable, a DBFIND search is done and the products are listed, each with
a link to the detail.html product page.
Real-Time Inventory
Real-time inventory is an example of extra product detail that's merged into the primary display on
the page. Typically you accomplish this by looking at an inventory field in the products database.
You can display the inventory, but more importantly, in the event the inventory for an item is
unavailable, you notify the user and deactivate the Add To Cart button. This is all you need to do on
the detail page to handle real-time inventory. The two Overlays to do this can be as simple as those
shown here:
<< # Overlay to inform user whether item is in stock /#
DISPLAY "Inventory: " /DISPLAY
IF prod_inventory=0 THEN
DISPLAY "<font color=red>Not In Stock</font>" /DISPLAY
ELSE
DISPLAY "<font color=red>In Stock</font>" /DISPLAY
/IF
>>
<< # Overlay to display Add To Cart button /#
IF prod_inventory != 0 THEN
DISPLAY
'<input type=submit name=mybutton value="Add To Cart">'
/DISPLAY
ELSE
DISPLAY
'<input type=submit name=mybutton value="Not In Stock">'
/DISPLAY
/IF
>>
This first Overlay uses an IF-THEN statement to inform the user whether the item is in stock. The
second Overlay also uses an IF-THEN statement, this time changing the Add To Cart button to a
Not In Stock button if prod_inventory is 0. Most of the programming associated with real-time
inventory is not here but in the final checkout process. There you need to ensure inventory is
available at order time. There is no need to do checking of this kind in this page or any page
preceding final checkout. The checkout process is described in Chapter 20, Building Checkout
Pages.
-224-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Product Collections
In many Web sites it makes more sense to display many products on a single detail page than place
the products on different Web pages. This is particularly important when the products are similar or
need to be compared against each other. These pages are called product collections. The Baby
Heirlooms site (http://www.babyheirlooms.com/) shown in Figure 18.1 demonstrates the use of
product collections.
Figure 18.1: This Baby Store Uses Product Collections to Display Related Products on the Same Web
Page.
When working with product collections, it is important to organize the fields in the products database
correctly. The recommended way to do this is as follows:
ƒ Place each product in a collection in a different record.
ƒ Define a collection ID that's the same for each record in a collection.
ƒ Define a collection title field that's the same for each.
ƒ Define a product suffix field that's unique to each.
ƒ Define a collection description field that's the same for each.
ƒ Define a product description field that's unique to each.
Displaying multiple products on a page is a matter of displaying the duplicated fields of one of the
records at the top of the page and following that with smaller detail pages, one for each record in the
collection, each with its own Add To Cart button. The code for this is left as an exercise at the end of
this chapter.
Displaying product collections may require added programming in the product navigation page. If you
wish to display each product in a collection, modification is unnecessary although you will want to
display product titles as the concatenation of the collection title and individual product title. To display
one listing per collection you will need to delete duplicate entries. To do this you use the following
algorithm:
# myresult format:
# column 1 - collection ID
# column 2 - product ID
# column 10 - unused
/#
-225-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
myresult=SORTCOL(myresult,1)
i=1 last_cid="*"
WHILE myresult[1,i] != "" DO
IF myresult[1,i] != last_cid THEN
last_cid=myresult[1,i]
ELSE
IF last_cid != "" THEN myresult[10,i]="DELETEME" /IF
/IF
i=i+1
/WHILE
myresult=GETCOLNOTEQ(myresult,10,"DELETEME")
This code takes myresult, a generic search result, and sorts it by the collection ID. Then, using a
WHILE loop, it scans the search result, marking column 10 with the value DELETEME when a
collection ID has changed, provided the collection ID is not empty. After the loop, the GETCOLNOTEQ
tag deletes the marked entries. This code is general in nature. You can use it any time you need to
delete duplicated entries from a list.
Displaying product collections may also require added programming when building record editors for
the database. In your record editor, when you edit a duplicated field in a record, you will want to
make sure you apply the same modification to the duplicated fields of all records in the same
collection. Backend systems, the pages used by office staff members to manage their site, are
described further in Chapter 21, Building Back-End Management Systems. Record editors are
described in Chapter 14, Building Database Editors.
Options and Add-Ons
Many e-commerce sites sell products and services with options and add-ons that need to be
determined at the time the item is added to a shopping cart. When the user selects options or addons, the same item is added to the cart except that the item description is changed perhaps along
with its price. The distinction between an option and an add-on for purposes of this book's discussion
is covered in the accompanying "Options versus Add-On" note.
Options versus Add-On—In this book the word option specifically refers to a product attribute that
needs to be selected at purchase time but does not change the price of the selection, for example,
the color of an iMac computer. An add-on is an item that the user may add to the same product, at
purchase time. Each add-on adds a surcharge to the base price of the product, for example, an
extended warranty for the iMac.
You store the specification of available options and add-ons in fields in the record for the product.
There are a number of ways to do this. In each case, you need to consider the following three
development issues:
ƒ How to format the data in the options or add-on field
ƒ How to convert this field into HTML form components on the Web page
ƒ How to add the user selection to the shopping cart
-226-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Formatting Your Data
The first issue relates to how the options or add-ons are stored in the products database. If your
products database only needs to support a single kind of option, a field called prod_options that
contains a comma-delimited list of option values is sufficient. In this case you format the content of
the field as follows:
option1,option2,option3
If there's a chance an option might contain a comma (,) character, use a vertical bar as your delimiter.
If your products database needs to support two kinds of options, add another options field. For
example, a clothing database might have options for sizes and colors. Defining the fields
prod_sizeoptions and prod_coloroptions could handle that situation.
In larger databases you might have an arbitrary number of options. For example, some products
might be clothing. Other products might be electrical appliances that require that the user to select
between American and European voltages (110 vs. 220 volts). Still other products may have other
kinds of options. In this case you want more flexibility. You can do this by defining one field that lists
your option types and another that lists the option values for each. The field name and format of the
two fields would be as follows:
prod_opt_titles title1,title2
prod_opt_values opt1,opt2,opt3|opt1,opt2,opt3
For example, the fields of one record may have the values:
prod_opt_titles Colors, Sizes
prod_opt_values Red,White,Black|Small,Medium,Large,Xlarge
Another record in the same database may have fields with the following values:
prod_opt_titles Voltage
prod_opt_values 110 (USA),220 (European)
The prod_opt_values field contains two option lists. It uses two separators. Here a comma
separates one option from the next, and a vertical bar separates one set of options from the next.
When working with add-ons, you can define the two fields, prod_addon_values and
prod_addon_surcharge. The format would be as follows:
prod_addon_values
1 Year Warranty, 4 Year Extended Warranty
prod_addon_surcharge 0,85.00
Both are comma-delimited lists. The first contains the titles of each add-on. The second contains the
surcharges associated with each.
Converting to HTML Form Components
Once you know how your data is stored in your options or add-on fields, you are ready to take that
data and convert it to HTML components on your product detail page. You can convert these fields
to pull-down menus or radio button. Here you convert the data to pull-down menus.
-227-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Consider the case of a single options list formatted as a comma-delimited list. Such a page is shown
in Figure 18.2.
Figure 18.2: Product Detail Pages Can be Outfitted with Product Option Menus, as Shown Here.
You provide the options using a pull-down menu placed beside the Add To Cart button. The options
are specified when a user adds the item to the shopping cart. To build this page, you can start with
the page described in the "A Six-Line Product Detail Page" section of Chapter 16, Designing ECommerce Systems." The code for that page, modified to display product options, is as follows:
<<temp=DBGETREC("products",rec)>>
<html>
<title>Sixteen-Line Product Detail Page With Options</title>
<b>My Database-Driven Store</b><br>
Product Detail:<form method=post ACTION=addtocart>
<a href=menu.html>Return to Menu</a><ul>
<b><<prod_title>></b><br>
<table border=0>
<tr>
<td valign=top><<prod_desc>></td>
<td valign=top>
<<IF ISFILE("/images/"+prod_image) THEN
DISPLAY "<img border=0 src=/images/"+prod_image+">"
/DISPLAY
/IF
>>
</td>
</tr>
<tr><td colspan=2>
<<IF prod_coloroptions != '' THEN
DISPLAY "Colors: <select name=addopt>" /DISPLAY
myopts=LISTTOCOL(prod_coloroptions,",")
DISPLAY
LAYOUT(myopts, '<option value="',[1], '">',[1])
-228-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
/DISPLAY
DISPLAY "</select>" /DISPLAY
/IF>>
<input type=submit value="Add To Cart"></td></tr>
</form></ul>
</html>
<<overlay addtocart
temp=DBGETREC("products",rec)
IF addopt != "ERROR" AND addopt != "" THEN
prod_title=prod_title+' ('+addopt+')'
/IF
APPEND ROW(rec,prod_title,1,prod_price) TO mycart /APPEND
addopt=''
GOTO "cart.html"
>>
Changes are shown in bold. First the code replaces the Add To Cart link with an Add To Cart button
in an HTML form. An image is displayed only if it is found using an IF-THEN statement. Then the
Overlay displays an Options pull-down menu (in the event options are found). The LISTTOCOL tag in
the Overlay converts the list of options into myopts, a one-column variable. Then the top of an
HTML select component, which sets the addopt variable, is displayed in the page. Using the
LAYOUT tag, the option lines in the select statement (formatted as <option
value=option>option) are displayed. This is followed by the close of the select statement. For
information on this HTML form component, see the "Form Components" section of Chapter 10,
HTML Forms Processing.
After the user submits the HTML form, addopt changes the product title of the item added to the
cart. If the user specifies an option or add-on, the title of the item is modified to reflect the option
selected.
If the data in the options fields is more complex, you need more coding to convert the data to pulldown elements. Consider the case of add-ons. The code to do that conversion is as follows:
IF prod_addon_values != '' THEN
addon_titles=LISTTOCOL(prod_addon_values,",")
addon_surcharges=LISTTOCOL(prod_addon_surcharges,",")
DISPLAY "<select name=add_addon>" /DISPLAY
DISPLAY '<option value="">select addon' /DISPLAY
i=1
WHILE addon_titles[1,i] != "" DO
DISPLAY
'<option
value="'+addon_titles[1,i]+'">'+addon_titles[1,i]+
' (Add $'+FORMAT(addon_surcharges[1,I],"money",2) +
-229-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
')'
/DISPLAY
i=i+1
/WHILE
DISPLAY "</select>" /DISPLAY
/IF
Here both fields are converted to one-column variables using LISTTOCOL tags. Here too a pull-down
menu is displayed. This code uses a WHILE loop. Note that each pull-down element is displayed with
its corresponding surcharge. Using this code, a product detail page with support for product add-ons
follows this paragraph. The addtocart on-click Overlay needed to process this is similar to the one
used in the sixteen-line, product detail page with options page described earlier in this section. Here
however, you must extract surcharges and apply them to the cart. The code for the on-click Overlays
is as follows:
<<overlay addtocart
temp=DBGETREC("products",rec)
IF add_addon != "ERROR" AND add_addon != "" THEN
col1=LISTTOCOL(prod_addon_values,","),
col2=LISTTOCOL(prod_addon_surcharges,",")
temp=MERGE(col1,col2)
temp=GETCOLEQ(temp,1,add_addon)
IF ISNUMBER(temp[2]) THEN
prod_title=prod_title+' ('+temp[1]+')'
prod_price=prod_price+temp[2]
/IF
/IF
APPEND ROW(rec,prod_title,1,prod_price) TO mycart /APPEND
add_addon=''
GOTO "cart.html"
>>
In this on-click Overlay the add_addon variable contains the title of the add-on selected. Here
LISTTOCOL tags convert the values in the add-on fields into two one-column variables that are then
pasted together using the MERGE tag. The MERGE tag is similar to the APPEND tag, except it appends
horizontally rather than vertically. After the two columns are merged, the add_addon variable in the
GETCOLEQ tag extracts the row from the merged table with the selected add-on. The title and
surcharge are extracted and used to adjust the values of prod_title and prod_price, which are
added to the mycart variable in the subsequent APPEND tag.
Summary
In this chapter you learned how to build advanced product detail pages. Like the navigation pages
described in the previous chapter, here too you can customize the pages to fit the individual needs of
the site. Perhaps the most important technique you learned is how to work with options and add-ons.
-230-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Although a tricky area, the ability to handle options and add-ons is a must-learn for the advanced
Web developer
It should be noted that this chapter introduced a few ways to process options and add-ons. You may
need to change the way you add the options or add-ons to your shopping cart. For example, instead
of altering the content of the product title and price, you may want to add the data to new columns in
mycart, used specifically for storing options or add-ons. This gives you more flexibility in the way
mycart is displayed on the shopping cart page.
The shopping cart is the topic of Chapter 19, Building Shopping Cart Pages, coming up next. There
you learn how to build different kinds of shopping carts.
Exercises
In the following exercises, you build other kinds of detail pages using what you learned in this
chapter. Answers to these exercises are provided on this book's companion Web site as described in
the book's Preface.
Exercise 1
Exercise 2
AM
FL
Y
Build a product detail page for a hypothetical Web site for selling books. The detail page should list
the title of the book and the author along with other titles by the same author. Make each title a link.
Exercise 3
TE
In Exercise 1 you built a small on-line bookstore. Using the outline provided in the "Product
Collections" section of this chapter, build a database to support book collections. When a book's
detail page is displayed, present information on the book and an Add To Cart link for each book in
the collection. Using what you learned in Chapter 17, Building Product Navigation Pages, build a
product list page that links to this detail page. Modify the page so only one book is displayed per
collection.
The "Options and Add-ons" section of this chapter describes how to set up the fields for product
databases with unlimited types of options. Build a product page that takes the values in the two fields
used to store these options and convert them to one or more pull-down menus, preceding each with
the title for that option. Make sure the addtocart on-click Overlay works as well. Hint: Break up the
options into multiple options lists and use a FOR loop to loop across all of the option lists, each time
generating a pull-down menu.
-231Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 19: Building Shopping Cart Pages
Overview
Fundamentally, a shopping cart is a temporary holding area for products being purchased. It's also
the place where users make adjustments to an order and you calculate the costs and display them
on the page for the user. The page described in the "Twenty-Line Shopping Cart" section of Chapter
16, Designing E-Commerce Systems, does all those things.
However, as a site gets more advanced, the purpose of the shopping cart expands. It becomes a
place to post promotional messages that inform the user of sales, and special opportunities. The cart
is not simply a temporary area that disappears when the user leaves the site, but a personal area the
user can return to in the future. Figure 19.1 shows a shopping cart page for Courtesy CareFree
Garden (http://www.carefreegardens.com/) with highlighted product capsules. In this chapter you
explore how to extend the capabilities of shopping carts to include these capabilities.
Figure 19.1: This Garden Supply Store Displays Promotional Products on Its Shopping Cart Page.
Promotional Messages
Promotional messages are a type of on-page marketing. By reminding the user of certain specials or
related products, the Web site may increase sales. As an example, when buying a toy, it is easy to
forget the need to buy batteries. Reminding the user to buy batteries is helpful to the buyer and the
Web site. Note that you can display these reminders on the product page, as explained in the
"Related Items" section of Chapter 18, Building Product Detail Pages. But you can also display them
on the shopping cart page.
Another type of sales reminder might prompt the user to take advantage of a special offer. For
example, when users buy one item, you might want to inform them that if they buy another they get it
at fifty percent off. Or, for example, when users place items totaling more than $100 in their cart, they
get free shipping. Most on-page promotions fall into the following three categories:
ƒ Product specific promotions— A Click Here To Buy Batteries button is an example of a
product-specific promotion.
-232-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ Category-triggered promotions— A banner that advertises printer paper whenever the
cart contains a printer, is an example of a category-triggered promotion (the category
being printers).
ƒ Cart-triggered promotions— Get Free Shipping on Orders Over $100 is an example of a
cart-triggered promotion.
Twenty-Four Line Cart with Shipping Promotion
To add a free shipping promotion, you need to do two things: First you need to modify your shipping
calculation to reflect the promotion. Second you need to alert the user of the promotion. Starting with
the code provided in the "Twenty-Line Shopping Cart Page" section of Chapter 16, Designing ECommerce Systems," you get the following:
<< # Structure of mycart:
# column 1 - record ID
# column 2 - product name
# column 3 - quantity
# column 4 - unit price
/#
mycart=GETCOLNOTEQ(mycart,1,"ERROR")
mycart=GETCOLNOTEQ(mycart,1,"")
IF mycart[1,1]="" THEN GOTO "nocart.html" /IF
IF ISNUMBER(tax)="FALSE" THEN tax=0 /IF
shipping=MAX(15,(3+3*rows(mycart)))
>>
<html>
<title>Twenty Four-Line Cart With Shipping Promotion</title>
<b>My Database-Driven Store</b><br>
Shopping Cart:<br>
<a href=menu.html>Return to Menu</a><ul>
<form method=post ACTION=changecart>
<table border=1><tr>
<td>Product</td><td>Unit
Cost</td><td>Quantity</td><td>Subtotal</td>
</tr>
<tr><td colspan=4 align=center>
<b>Free Shipping On All Orders Over $100</b></td>
</tr>
<< i=1 subtotal=0
FOR NAME=mycart ROWNAME=x DO
DISPLAY
"<tr><td><a href=detail.html name=rec VALUE="+x[1]+">"+
x[2]+"</a> (<a href=dcart NAME=drec
-233-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
VALUE="+x[1]+">Delete</a>)"+
"</td><td>"+x[4]+"</td>"+
"<td><input type=text name=mycart[3,"+i+"]
size=4></td>"+
"<td
align=right>"+FORMAT((x[3]*x[4]),"comma")+"</td></tr>"+LF
/DISPLAY
subtotal=subtotal+(x[3]*x[4])
i=i+1
/FOR
IF subtotal > 100 THEN
shipping=0
/IF
>>
<tr><td colspan=3>Subtotal</td>
<td align=right><<FORMAT(subtotal,"comma")>></td></tr>
<tr><td colspan=3>State Sales Tax:
<a href=<<page>> NAME=tax value="0.0625">Yes</a>
<a href=<<page>> NAME=tax value="0.0000">No</a>
</td><td align=right><<FORMAT(tax*subtotal,"comma")>></td></tr>
<tr><td colspan=3 name=shipping>Delivery:
</td><td align=right><<FORMAT(shipping,"comma")>></td></tr>
<tr><td colspan=3>TOTAL</td>
<td
align=right><<FORMAT(subtotal+tax*subtotal+shipping,"comma")>></
td></tr>
<tr><td colspan=4 align=right>
<input type=submit name=mybutton value="Recalculate">
<input type=submit name=mybutton value="Checkout">
</td></tr>
</table>
</form>
</html>
<<overlay changecart
i=1 subtotal=0
WHILE mycart[1,i] != "" DO
IF ISINTEGER(mycart[3,i])!= "TRUE" OR
mycart[3,i] =0 OR mycart[3,i] < 0 THEN
mycart[1,i]=""
ELSE
-234-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
subtotal=subtotal+mycart[3,i]*mycart[4,i]
/IF
i=i+1
/WHILE
mycart=GETCOLNOTEQ(mycart,1,"")
shipping=MAX(15,(3+3*rows(mycart)))
IF subtotal > 100 THEN shipping=0 /IF
IF mybutton="Checkout" THEN
GOTO "checkout.html"
ELSE
GOTO PAGE
/IF
>>
<<overlay dcart
mycart=GETCOLNOTEQ(mycart,1,drec)
GOTO PAGE
>>
Changes are shown in bold. The first change modifies the shipping variable after the total in the cart
is known. A simple IF-THEN statement accomplishes this. Then an HTML table row alerts the user
of the shipping promotion. Then, in the changecart on-click Overlay, an IF-THEN statement, like
earlier on the Web page, modifies the shipping calculation. This is an example of a cart-specific
promotion.
Thirty-Five Line Cart with Product Promotion
Product-specific promotions require that you first add a field to the product database to indicate the
promotion being applied to the item. Typically it's a good idea to give promotions their own codes so
you can develop multiple product promotions. For example, suppose the product promotion is called
"A." You can add code to the shopping cart page that provides information on the promotion when
such a product is detected and you change your calculations accordingly. For example, if your offer
is a buy one and get the next 50 percent off promotion, you write the shopping cart instructions as
follows:
<< # Structure of mycart:
# column 1 - record ID
# column 2 - product name
# column 3 - quantity
# column 4 - unit price
# column 5 - promotion code
/#
mycart=GETCOLNOTEQ(mycart,1,"ERROR")
mycart=GETCOLNOTEQ(mycart,1,"")
IF mycart[1,1]="" THEN GOTO "nocart.html" /IF
-235-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF ISNUMBER(tax)="FALSE" THEN tax=0 /IF
shipping=MAX(15,(3+3*rows(mycart)))
promos=GETCOLNOTEQ(mycart,5,"")
promo_message=""
IF promos[1,1]!= "" THEN
FOR NAME=promos ROWNAME=promo DO
IF promo[5]="A" THEN
promo_message=promo_message+
"Buy One <u>"+promo[2]+"</u> and get the next half
off!<br>"
/IF
/FOR
/IF
>>
<html>
<title>Thirty Five-Line Cart With Shipping Promotion</title>
<b>My Database-Driven Store</b><br>
Shopping Cart:<br>
<a href=menu.html>Return to Menu</a><ul>
<form method=post ACTION=changecart>
<table border=1><tr>
<td>Product</td><td>Unit
Cost</td><td>Quantity</td><td>Subtotal</td>
</tr>
<< IF promo_message != "ERROR" THEN
DISPLAY "<tr><td colspan=4 align=center><b>"+
promo_message+"</b></td></tr>" /DISPLAY
/IF
>>
<< i=1 subtotal=0
FOR NAME=mycart ROWNAME=x DO
DISPLAY
"<tr><td><a href=detail.html name=rec VALUE="+x[1]+">"+
x[2]+"</a> (<a href=dcart NAME=drec
VALUE="+x[1]+">Delete</a>)"+
"</td><td>"+x[4]+"</td>"+
"<td><input type=text name=mycart[3,"+i+"]
size=4></td>"+
"<td align=right>"
/DISPLAY
-236-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF x[3]<2 THEN
DISPLAY FORMAT((x[3]*x[4]),"comma")+"</td></tr>"+LF /DISPLAY
+=0
ELSE
t=ROUNDDOWN(x[3]/2)*x[4]*0.50
DISPLAY
FORMAT(((x[3]-t)*x[4]),"comma")+"</td></tr>"+LF
/DISPLAY
/IF
subtotal=subtotal+((x[3]-t)*x[4])
i=i+1
/FOR
>>
<tr><td colspan=3>Subtotal</td>
<td align=right><<FORMAT(subtotal,"comma")>></td></tr>
<tr><td colspan=3>State Sales Tax:
<a href=<<page>> NAME=tax value="0.0625">Yes</a>
<a href=<<page>> NAME=tax value="0.0000">No</a>
</td><td align=right><<FORMAT(tax*subtotal,"comma")>></td></tr>
<tr><td colspan=3 name=shipping>Delivery:
</td><td align=right><<FORMAT(shipping,"comma")>></td></tr>
<tr><td colspan=3>TOTAL</td>
<td
align=right><<FORMAT(subtotal+tax*subtotal+shipping,"comma")>></
td></tr>
<tr><td colspan=4 align=right>
<input type=submit name=mybutton value="Recalculate">
<input type=submit name=mybutton value="Checkout">
</td></tr>
</table>
</form>
</html>
<<overlay changecart
i=1 subtotal=0
WHILE mycart[1,i] != "" DO
IF ISINTEGER(mycart[3,i])!= "TRUE" OR
mycart[3,i] =0 OR mycart[3,i] < 0 THEN
mycart[1,i]=""
ELSE
IF mycart[4,i] = "" THEN
-237-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
subtotal=subtotal+mycart[3,i]*mycart[4,i]
ELSE
t=ROUNDDOWN(x[3]/2)*x[4]*0.50
subtotal=subtotal+((x[3]-t)*x[4])
/IF
/IF
i=i+1
/WHILE
mycart=GETCOLNOTEQ(mycart,1,"")
shipping=MAX(15,(3+3*rows(mycart)))
IF mybutton="Checkout" THEN
GOTO "checkout.html"
ELSE
GOTO PAGE
/IF
>>
<<overlay dcart
mycart=GETCOLNOTEQ(mycart,1,drec)
GOTO PAGE
>>
Changes are shown in bold. The Underlay uses GETCOLNOTEQ to extract those lines in the cart with
a promotion. If a promotional item is found, a promotion message is composed. Note the IF-THEN
statement placed inside the FOR loop. It sets a message for promotion type A. If you have other
promotion codes, add more IF-THEN statements in the FOR loop to handle them.
Later in the document the promo_message variable is displayed. Then, in the changecart on-click
Overlay, the subtotal calculation is modified to take into account the type A promotion. You can add
more promotion types along side this IF-THEN statement to perform the necessary calculations.
Saving Shopping Carts using Cookies
Saving a shopping cart so it is not lost when the user leaves your sites (abandons the cart), can be
accomplished by using the Cookie Tracking feature built into HTML/OS. Cookie Tracking is
explained in the "Cookie Tracking" note.
Cookie Tracking—Cookie Tracking is a setting you can turn on in the HTML/OS Control Panel.
When you turn it on, if HTML/OS is able to read a cookie successfully, it sets ISCOOKIE to TRUE.
HTML/OS also gives the user a unique ID, which it saves in the HTMLOS . COOKIEID variable.
Once HTMLOS . COOKIEID is saved in a cookie, HTML/OS does not change it, so it can reliably
identify a specific computer returning to the site. To use this feature, access the Control Panel from
the HTML/OS desktop. Click System. Click Cookies. Then select On and click the Save Settings
button. Now Cookie Tracking is on.
-238-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Cookie tracking requires that a user's cart, previously saved in a file, be loaded when the user first
enters the site. Then, as the cart is changed, this file is updated, ensuring that users who abandon
their carts can return later. The code for the entrance page of the site is as follows:
IF mycart="ERROR" THEN
IF ISCOOKIE="TRUE" THEN
IF ISFILE("/cookies/"+HTMLOS.COOKIEID+".txt") THEN
COPY FILE="/cookies/"+HTMLOS.COOKIEID+".txt" TS="," TO
mycart /COPY
ELSE
mycart=""
/IF
/IF
This code initializes the cart using the cookie file containing the cart. Note that we arbitrarily named
the cookie file with the name of the HTMLOS.COOKIEID variable followed with a .txt extension and
we placed it in the /cookies directory. In the cart itself, or whenever the mycart variable is
modified, you need the following code to update this file:
IF ISCOOKIE="TRUE" THEN
COPY mycart TO FILE="/cookies/"+HTMLOS.COOKIEID+".txt" TS=","
/COPY
/IF
This ensures that users can leave the site at any time and return to find their carts intact. The
algorithm outlined here works fine except for two problems. First, what happens when you have so
many people visiting your site that the /cookies directory fills up with tens of thousands of files?
This is dangerous since directories with so many files can slow down a site. Second, what if the
returning users have products in their carts that are no longer available, or products with out-of-date
prices?
To solve the first potential problem it is important to understand that most hardware operating
systems slow down when a directory has too many files in it, because responses to file requests
slow down. Most operating systems organize files internally as series of lists. The more files, the
longer the lists are that have to be scanned when a file needs to be located on the hardware. One
solution is to limit the number of files you have in a directory by spreading them across many
directories. If you have too many directories, limit the number of directories by nesting them within
one another. Now, when the hardware operating system looks for a file, it needs to scan a few short
lists rather than one very long list spread all across the hard disk.
To code a solution, you will need to set up different directories for your cookies so that no single
directory has too many files. This is left as an exercise at the end of this chapter.
To ensure that items in the cart are still valid, the previous code placed at the entrance of the site
needs to contain additional code that checks each item in the cart. If an item is no longer in the
products database, it is omitted from the cart. If it is found, settings such as price are updated. The
code for this is as follows:
-239-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF mycart="ERROR" THEN
IF ISCOOKIE="TRUE" THEN
IF ISFILE("/cookies/"+HTMLOS.COOKIEID+".txt") THEN
COPY FILE="/cookies/"+HTMLOS.COOKIEID+".txt" TS="," TO
mycart /COPY
i=1
FOR NAME=mycart ROWNAME=x DO
sstr='prod_name="' + x[2] + '"'
r=DBFIND("products",sstr,1,1,"record,prod_price")
IF r[1,1]!="" THEN
mycart[1,i]=r[1]
mycart[4,i]=r[2]
ELSE
mycart[1,i]=""
/IF
i=i+1
/FOR
mycart=GETCOLNOTEQ(mycart,1, "")
ELSE
mycart=""
/IF
/IF
Changes are shown in bold. The DBFIND tag searches the products database. If an item is found,
it updates the first and fourth column entries. If not, it places an empty string in the first column entry
of the cart. Those rows marked with an empty string are later deleted using a GETCOLNOTEQ tag
immediately following the end of the FOR loop.
Building Advanced Shipping Options
Most shopping systems give the user the ability to select from a list of shipping options. The options
are presented to the user on the shopping cart page or on a checkout page. By placing the shipping
information on the shopping cart page, you give the user the ability to determine a true cost for the
items in the cart. By placing the option on a checkout page, you deny the user a true cost but you
simplify the cart for the user and merge the function with related tasks, such as capturing a shipping
address. In some cases the shipping cost will depend on the destination requiring you place it on a
checkout page. Here you look at placing shipping options on the shopping cart page. Specifically,
you look at giving a user the ability to select ground, overnight, or express shipping.
Adding shipping options is similar to adding a tax calculation to a shopping cart page except that you
use a pull-down menu to offer the selections (rather than some hypertext links) and that you must
change the shipping charge accordingly. Starting with the twenty-line shopping cart of Chapter 16,
Designing E-Commerce Systems, you get the following page:
<< # Structure of mycart:
-240-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
# column 1 - record ID
# column 2 - product name
# column 3 - quantity
# column 4 - unit price
/#
mycart=GETCOLNOTEQ(mycart,1,"ERROR")
mycart=GETCOLNOTEQ(mycart,1,"")
IF mycart[1,1]="" THEN GOTO "nocart.html" /IF
IF ISNUMBER(tax)="FALSE" THEN tax=0 /IF
IF shipcode="ERROR" THEN shipcode="A" /IF
IF shipcode="A" THEN
shipping=MAX(15,(3+3*rows(mycart)))
ELIF shipcode="B" THEN
shipping=MAX(25,(8+4*rows(mycart)))
ELSE
shipping=MAX(49,(15+5*rows(mycart)))
AM
FL
Y
/IF
>>
<html>
<title>Thirty-Line Cart With Shipping Options</title>
<b>My Database-Driven Store</b><br>
Shopping Cart:<br>
TE
<a href=menu.html>Return to Menu</a><ul>
<form method=post ACTION=changecart>
<table border=1><tr>
<td>Product</td><td>Unit
Cost</td><td>Quantity</td><td>Subtotal</td>
</tr>
<< i=1 subtotal=0
FOR NAME=mycart ROWNAME=x DO
DISPLAY
"<tr><td><a href=detail.html name=rec VALUE="+x[1]+">"+
x[2]+"</a> (<a href=dcart NAME=drec
VALUE="+x[1]+">Delete</a>)"+
"</td><td>"+x[4]+"</td>"+
"<td><input type=text name=mycart[3,"+i+"]
size=4></td>"+
"<td
align=right>"+FORMAT((x[3]*x[4]),"comma")+"</td></tr>"+LF
/DISPLAY
-241Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
subtotal=subtotal+(x[3]*x[4])
i=i+1
/FOR
>>
<tr><td colspan=3>Subtotal</td>
<td align=right><<FORMAT(subtotal,"comma")>></td></tr>
<tr><td colspan=3>State Sales Tax:
<a href=<<page>> NAME=tax value="0.0625">Yes</a>
<a href=<<page>> NAME=tax value="0.0000">No</a>
</td><td align=right><<FORMAT(tax*subtotal,"comma")>></td></tr>
<tr><td colspan=3>
Shipping: <select name=shipcode>
<option value="A">Ground
<option value="B">Two-Day
<option value="C">Overnight
</select>
</td><td align=right><<FORMAT(shipping,"comma")>></td></tr>
<tr><td colspan=3>TOTAL</td>
<td
align=right><<FORMAT(subtotal+tax*subtotal+shipping,"comma")>></
td></tr>
<tr><td colspan=4 align=right>
<input type=submit name=mybutton value="Recalculate">
<input type=submit name=mybutton value="Checkout">
</td></tr>
</table>
</form>
</html>
<<overlay changecart
i=1 subtotal=0
WHILE mycart[1,i] != "" DO
IF ISINTEGER(mycart[3,i])!= "TRUE" OR
mycart[3,i] =0 OR mycart[3,i] < 0 THEN
mycart[1,i]=""
ELSE
subtotal=subtotal+mycart[3,i]*mycart[4,i]
/IF
i=i+1
/WHILE
mycart=GETCOLNOTEQ(mycart,1,"")
-242-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF shipcode="A" THEN
shipping=MAX(15,(3+3*rows(mycart)))
ELIF shipcode="B" THEN
shipping=MAX(25,(8+4*rows(mycart)))
ELSE
shipping=MAX(49,(15+5*rows(mycart)))
/IF
IF mybutton="Checkout" THEN
GOTO "checkout.html"
ELSE
GOTO PAGE
/IF
>>
<<overlay dcart
mycart=GETCOLNOTEQ(mycart,1,drec)
GOTO PAGE
>>
Changes are shown in bold. To handle the different shipping options, you use a shipcode variable,
which can contain the values A, B, or C, corresponding to ground, express, or overnight delivery
respectively. The shipcode variable is first initialized to the value A. Then, depending on the value,
different shipping formulas are used.
Inside the page, a pull-down menu gives the user the ability to select a shipping option. When the
user clicks Recalculate the HTML form is resubmitted, thereby setting a new value for shipcode.
The changecart on-click Overlay also includes code, shown in bold, to recalculate the shipping
charge when the user clicks the Checkout button.
Summary
In this chapter you learned how to extend the capabilities of shopping carts. Next, in Chapter 20,
Building Checkout Pages, you learn how to combine the information stored in the cart with that
information used to complete the ordering process. The next chapter, along with this one, gives you
the ability to build a wide variety of ordering systems—an important component of e-commerce
applications.
Exercises
In the following exercises you build pages that compliment those described in this chapter. Answers
to these exercises are provided on this book's companion Web site as described in the book's
Preface.
Exercise 1
In the "Saving Shipping Carts Using Cookies" section of this chapter, you learned that placing
thousands of files in a single directory is not advisable. Build a function that takes any filename and
places it in one of two hundred directories. Hint: To do this, add up the ASCII values of the
-243-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
characters in the filename (use the GETASCII and MIDDLE tags) and find modulus 200 of the sum,
in other words, the integer portion of the sum divided by 200. (Use the MOD tag.) Use this to
determine the directory name in which to place the file.
Exercise 2
The "Building Advanced Shipping Options" section of this chapter used calculations hard-coded into
the Web page. Modify the page so shipping options are read from a delimited text file. Place in the
columns a shipping code, a shipping title, and the parameters X, Y, and Z used in the formula
shipping=MAX(Z,X + Y*rows(mycart)). Modify the Web page so modifications to the shipping
file change the shipping options and shipping calculations presented to users.
Exercise 3
Starting with the code provided in the "Fifteen-Line Spreadsheet Editor" section of Chapter 10, HTML
Forms Processing, build an editor for the shipping parameters of Exercise 2. Now you have a backend page that gives staff members the ability to edit shipping options and shipping calculations.
-244-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 20: Building Checkout Pages
Overview
Order checkout is the part of an e-commerce site associated with the completion of a customer
transaction. At the least, checkout pages include the features discussed in the "A Nine-Line
Checkout Page" section of Chapter 16, Designing E-Commerce Systems. But as the sophistication
of the e-commerce site grows, so too does the feature list of this type of page.
Sophisticated e-commerce sites spread the checkout pages across multiple pages, which increases
the number of pages a user must navigate and goes against the Web design rule: Don't have the
user navigate unnecessary pages. However, in this particular case, the design rule is superceded
with a more important design rule: Don't make the user think about more than one transaction at a
time. Since advanced e-commerce sites provide users more options, requiring users to make more
decisions and supply more information, it is best to split the order checkout process into distinct
pages, each encompassing a smaller transaction.
In this chapter the checkout process is spread across a shipping page, a payment page, a review
page, and an order confirmation page. You look at the code needed for each page and the
sophisticated features you may add to them.
A Six-Line Shipping Page
The shipping page is where you collect the shipping address information. In this chapter you look at
a checkout process split across multiple pages. As a result, users need page navigation options,
which we accomplish here with Previous and Continue buttons. You place these buttons in the
HTML form so changes made to that form are captured, even if the user clicks a button. The
Continue button directs the user to the review.html page, if they have been to that page.
Otherwise the Continue button directs the user to the next page. A six-line shipping page follows:
<html>
<title>Six-Line Shipping Page</title>
<b>My Database-Driven Store</b><br>
<b>Checkout</b><br>
Shipping Information (Page 1 of 3):<br>
<form method=post ACTION=postpage>
Enter shipping address below.<br>
<table bgcolor=#CECECE border=0 cellspacing=0>
<tr><td align=right>Street</td>
<td><input type=text name=ck_street1></td></tr>
<tr><td align=right>&nbsp;</td>
<td><input type=text name=ck_street2></td></tr>
<tr><td align=right>City, State, Zip</td>
<td><input type=text name=ck_city size=20>
<input type=text name=ck_state size=4>
-245-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<input type=text name=ck_zip size=10></td></tr>
<tr><td align=right>Country</td>
<td><input type=text name=ck_country></td></tr>
<tr bgcolor=#FFFFFF><td colspan=2 align=center>
<input type=submit name=mybutton value="Previous">
<input type=submit name=mybutton value="Continue">
</td></tr></table></form>
</html>
<<overlay postpage
IF mybutton="Previous" THEN GOTO "cart.html"
ELSE
IF review_flag="TRUE" THEN GOTO "review.html"
ELSE GOTO "purchase.html"
/IF
/IF
>>
This page contains an HTML form to capture shipping information and Previous and Continue
buttons. When a user clicks a Previous button the postpage on-click Overlay runs and directs the
user to cart.html. A user who clicks the Continue button is directed to review.html if the
review_flag variable is TRUE (which is set to TRUE on the review.html page). Otherwise the
user is directed to purchase.html, the next checkout page.
On an advanced e-commerce site you also want to add data validation. To do that use the
techniques described in the "Commonly Used Validation Schemes" section of Chapter 10, HTML
Forms Processing. If you want the user to select a shipping method, use the technique discussed in
the "Building Advanced Shipping Options" section of Chapter 19, Building Shopping Cart Pages.
This shipping page is where you place other kinds of information or features pertaining to shipping.
As an example, you may want to add to this page an option that allows the user to determine
whether to split orders if one of the items is backordered. Or you may want to give the user the ability
to save the shipping address for future use or load a shipping address from previously saved
addresses.
A Twelve-Line Payment Page
The payment page is where you capture contact information and payment information. The page is
similar to the shipping page in the preceding section except you seek to capture payment information
rather than shipping information. You also want to place a billing address on the page in case the
address is different from the shipping address. The page is as follows:
<html>
<title>Twelve-Line Payment Page</title>
<b>My Database-Driven Store</b><br>
<b> Checkout</b><br>
-246-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
PaymentInformation (Page 2 of 3):<br>
<form method=post ACTION=postpage>
Full Name:<br>
<input type=text name=ck_bname><br>
E-Mail Address:<br>
<input type=text name=ck_bemail><br>
Enter Billing Address or click <a href=sameas>same as
shipping</a>.<br>
<table bgcolor=#CECECE border=0 cellspacing=0>
<tr><td align=right>Street</td>
<td><input type=text name=ck_bstreet1></td></tr>
<tr><td align=right>&nbsp;</td>
<td><input type=text name=ck_bstreet2></td></tr>
<tr><td align=right>City, State, Zip</td>
<td><input type=text name=ck_bcity size=20>
<input type=text name=ck_bstate size=4>
<input type=text name=ck_bzip size=10></td></tr>
<tr><td align=right>Country</td>
<td><input type=text name=ck_bcountry></td></tr>
<tr><td colspan=2>Payment Information:</td></tr>
<tr><td align=right>Credit Card</td>
<td><input type=text name=ckb_card size=15>(MM/YYYY)</td></tr>
<tr><td align=right>Exp Date</td>
<td><input type=text name=ckb_expdate
size=15>(MM/YYYY)</td></tr>
</table>
<input type=submit name=mybutton value="Previous">
<input type=submit name=mybutton value="Continue">
</form>
</html>
<<overlay postpage
IF mybutton="Previous" THEN GOTO "cart.html"
ELSE
IF review_flag="TRUE" THEN GOTO "review.html"
ELSE GOTO "purchase.html"
/IF
/IF
>>
<<overlay sameas
ck_bstreet1=ck_street1
-247-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ck_bstreet2=ck_street2
ck_bcity=ck_city
ck_bstate=ck_state
ck_bzip=ck_zip
ck_bcountry=ck_country
GOTO PAGE
>>
The page requests contact information and a billing address and provides a Same As Shipping link
so the user can load the shipping address. When the user clicks that link, the sameas on-click
Overlay, from which the variables are loaded from the prior page, runs. Below this you capture the
billing address credit card information and provide Previous and Continue buttons. Users who click
the Previous button are directed to shipping.html, the previous page. When users click the
Continue button they are directed to review.html, the next checkout page.
As on the shipping page, you will want to add data validation to this page. Here too, like the shipping
page, you may want to give the user the ability to save multiple addresses for future use. This is left
as an exercise at the end of this chapter.
A Twelve-Line Final Review Page
Once the order information has been captured you want to give the user the ability to review and
make any necessary changes needed to the ordering information before completing the order. To do
this, provide a page that summarizes the content of the cart, the shipping information, and the
payment information. Next to each type of information provide links to the corresponding pages
where the information can be edited. A sample page is shown in Figure 20.1.
Figure 20.1: The Checkout Pages of this Software Company Include this Review Page so a User Can
Edit the Order before Finalizing It.
Assuming the same variables as the cart described in the "Twenty-Line Shopping Cart Page" section
of Chapter 16, Designing E-Commerce Systems, the code for the page shown in Figure 20.1 is as
follows:
<<review_flag="TRUE">>
-248-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<html>
<title>Twelve-Line Final Review Page</title>
<b>My Database-Driven Store</b><br>
<b> Checkout</b><br>
Final Review (Page 3 of 3):<br>
Review information below. Click Complete Order when
done.<br><br>
<a href=cart.html>Products</a>:<ul>
<table border=0><tr>
<td>Product</td><td>Unit
Cost</td><td>Quantity</td><td>Subtotal</td>
</tr>
<< FOR NAME=mycart ROWNAME=x DO
DISPLAY
"<tr><td>"+x[2]+"</td><td>"+x[4]+"</td><td "+x[3]+"</td>"+
"<td
align=right>"+FORMAT((x[3]*x[4]),"comma")+"</td></tr>" LF
/DISPLAY
/FOR
>>
<tr><td colspan=3>Subtotal</td>
<td align=right><<FORMAT(subtotal,"comma")>></td></tr>
<tr><td colspan=3>State Sales Tax:
</td><td align=right><<FORMAT(tax*subtotal,"comma")>></td></tr>
<tr><td colspan=3 name=shipping>Delivery:
</td><td align=right><<FORMAT(shipping,"comma")>></td></tr>
<tr><td colspan=3>TOTAL</td>
<td align=right>
<<FORMAT(subtotal+tax*subtotal+shipping,"comma")>></td></tr>
</table>
</ul>
<a href=shipping.html>Shipping Information</a>:<ul>
<table border=0 cellspacing=0>
<tr><td align=right>Street</td><td><<ck_street1>></td></tr>
<tr><td align=right>&nbsp;</td><td><<ck_street2>></td></tr>
<tr><td align=right>City, State, Zip</td><td><<ck_city>>,
<<ck_state>> <<ck_zip>></td></tr>
<tr><td align=right>Country</td><td><<ck_country>></td></tr>
</table>
-249-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</ul>
<a href=payment.html>Payment Information</a>:<ul>
Full Name: <<ck_bname>><br>
E-Mail Address: <<ck_bemail>><br>
<table border=0 cellspacing=0>
<tr><td>ADDRESS</td></tr>
<tr><td align=right>Street</td><td><<ck_bstreet1>></td></tr>
<tr><td align=right>&nbsp;</td><td><<ck_bstreet2>></td></tr>
<tr><td align=right>City, State, Zip</td><td><<ck_bcity>>,
<<ck_bstate>> <<ck_bzip>></td></tr>
<tr><td align=right>Country</td><td><<ck_bcountry>></ d></tr>
<tr><td>Credit Card</td>
<td><<ckb_card size=15>> Exp: <<ckb_expdate>></td></tr>
</table>
</ul>
<form method=post ACTION=order.html>
<input type=submit value="Complete Order">
</form>
</html>
The first instruction sets review_flag to TRUE so users editing the ordering information with
shipping.html or payment.html return to review.html when they click the Continue button
on those pages. Then cart information is displayed along with a link to cart.html. After that
shipping and payment information is displayed, with links to shipping.html and payment.html
respectively. After that a Complete Order button links to order.html, the page that processes the
order.
Order Processing
Final order processing takes place after the user clicks the Complete Order button in the code
presented in the preceding section. This is the point where you can assume the order is final. During
order processing, you need to perform a number of tasks, after which you provide the user a Thankyou notice and perhaps a receipt. These tasks need to be done quickly, since they are done
between the instant the user clicks Complete Order and receives the Thank-you page. The tasks can
involve any of the following, in the order shown here:
1.
2.
3.
4.
5.
6.
7.
Check product inventory to ensure items in the cart are in stock.
Post order to a real-time transaction processor (such as Authorize.net).
Post order to backup log.
Add records to order history and transaction databases.
Add or edit customer database record.
Compose and send e-mail confirmation to order desk.
Compose and send e-mail confirmation to user.
To check product inventory, the first task listed, use a FOR loop to scan the items in the cart. If
inventory on an item is insufficient, abort order processing, compose a message for the user, adjust
-250-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
the order to reflect what is available for ordering, and then direct the user to a page for posting the
adjusted order.
Orders can be posted to a real-time credit card processor bureau, such as Authorize.net
(http://www.authorize.net/), if you have contracted for such a service. This is a Web-based version of
a credit card swiping machine you see in restaurants. Using instructions provided to you by the realtime credit card processing bureau, you send credit card and price information over the Web. The
bureau processes these requests in real time, returning a result to you in a fraction of a second. The
result you receive tells you whether the transaction succeeded. You use this result to continue your
order processing or abort the transaction.
In general, credit card processing services bureaus provide two ways to interact with them. Either
they tell you how to organize your information and post it to them over the Web (using the NETWEB
tag, for example) or they provide you software to run at your site. The NETWEB tag is described in the
Official HTML/OS Print Manual. Code examples for hooking up to real-time processing bureaus,
such as Authorize.net, can be found in the knowledge base on the Aestiva Web site at
http://www.aestiva.com/support/.
AM
FL
Y
It should be noted that many e-commerce sites do not need real-time credit card processing,
because it is often against the policies of the bank to deduct money from customer credit card
accounts until the ordered item is shipped. Many e-commerce sites deal in physical goods where
real-time credit card processing does not make sense.
TE
Tasks 3, 6, and 7 in the numbered list earlier in this section have been described in previous
chapters. You accomplish Task 3, posting the order to a backup log, by using the APPEND tag as
shown in the "Ten-Line E-Commerce Web Site" section in Chapter 16, Designing E-Commerce
Systems. Tasks 6 and 7, composing and sending e-mail confirmations to the order desk and user
are examples of sending e-mails. Composing a sales e-mail is described in the same "Ten-Line ECommerce Web Site" section. Extending this so the e-mails sent out can be defined using a
template, is described in the "Using an E-Mail Template" section of Chapter 9, Building Login Pages.
Tasks 4, adding records to order history and transaction databases, and 5, adding or editing
customer database records, depend on your back-end system. Task 4 is required if you keep an
order history for your staff or users. Typically you save your order history in two database tables, one
that saves basic order information and the other for the items ordered. You use an order number
unique for each order to cross-reference the tables. If you use real-time credit card processing, you
may also want to post the payment in a database reserved for financial transactions. Later, in a
back-end system, you will need this database to obtain payment history for the customer or post
credits and other payment adjustments, if it becomes necessary.
Once some or all of Tasks 1 through 7 are completed, assuming the transaction is successful, the
user is directed to a thank you or confirmation page.
Order Confirmation
During order confirmation, you provide the users with a Thank-you message and explain how to
contact you in the event they experience a problem. The Thank-you message reinforces to the user
that the order has been successfully processed.This feedback is important since it gives the user a
sense of closure. The page can be extremely simple, like that presented in the "One-Line
Confirmation Page" section of Chapter 16, Designing E-Commerce Systems. Or you can be more
helpful by providing the user a printable receipt.
-251Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
To provide a printable receipt, add a link to the thank you page directing users to a new Web page.
On the page, organize the ordering information the way a typical receipt would look and provide a
Print This Page message at the top.
Summary
In this chapter you learned about checkout pages organized as a sequence of pages that capture
ordering information followed by a review page where the user can either go back and edit the
information or post the order. Of course, this particular design is not the rule. You have the freedom
to organize checkout pages any way you want.
You may want the user to log in and provide shipping and payment information before being able to
submit an order. This is a good way to organize your pages if you expect a lot of repeat business. It
makes it easy for you to add quick-order buttons and single-step checkouts that take advantage of
the information already on file for a user.
If orders include soft items, such as information packets or software that can be downloaded from
the Web, you want your checkout pages to include a way to download the items purchased.
In general, the way you organize your checkout page will depend on the nature of your products and
services, the kinds of customers you have, and your own creativity.
Next, in Chapter 21, Building Back-End Management Systems, you explore the kinds of systems
used by staff to manage e-commerce sites. The chapter completes this discussion on e-commerce
leaving you with an appreciation for the intricacies of e-commerce development and more
importantly, the tools to build advanced e-commerce sites.
Exercises
In the exercises that follow you build many of the components discussed in this chapter. Answers to
these exercises are provided on this book's companion Web site as described in the book's Preface.
Exercise 1
Add data validation to the six-line shipping page and the twelve-line payment pages described at the
beginning of this chapter. Hint: Use the code described in the "Commonly Used Validation Schemes"
section of Chapter 10, HTML Forms Processing.
Exercise 2
Sophisticated Web sites such as Amazon.com (http://www.amazon.com/) give users the ability to
keep multiple shipping and delivery addresses on file. Using cookies gives users such an option. Add
a pull-down menu followed with Save, Load, and Delete buttons. Clicking Save should add the
current entry to the list. Clicking Load should load the selected entry into the HTML form. Clicking
Delete should delete the selected entry. Save the list in a comma-delimited file using the
HTMLOS.COOKIE ID describe in the "Saving Shopping Carts Using Cookies" section of Chapter 19,
Building Shopping Cart Pages.
Exercise 3
Some sophisticated e-commerce sites require users to log in before making a purchase. Using what
you learned in Chapter 9, Building Login Pages, precede the shipping and payment page described
in this chapter with a login page. Hook this up to a cart.html page so, upon checking out, users
must log in, if they have not already done so.
-252-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 21: Building Back-End Management
Systems
Overview
No book on advanced Web sites would be complete without a discussion of back-end systems.
Back-end means the part of the Web site used for performing office tasks, such as those associated
with maintaining the pages for users of the Web site. We use the terms back-end and back-office
interchangeably.
As a site gets more advanced, so does the need for Web pages office workers can use to maintain
the site. As a site takes on more functions and provides more kinds of information, the production
department imparts the maintenance of that information to other individuals in the organization.
The tasks performed by these other individuals are done with back-office Web pages that
correspond to pages or features of the Web site. For example, product capsules appearing on the
home page would correspond to back-end pages for selecting the products to be placed on that
page. Or the automated e-mail messages sent to customers when they order a product would
correspond to a back-end Web page for editing the e-mail template used to compose the automated
e-mail.
This chapter concentrates on back-end Web pages for maintaining Web sites. But you could easily
apply the concepts here to the construction of any kind of back-office system.
Web-based computing is ideally suited to the construction of Web-based offices since office tasks
are most often related to communication and the handling or processing of information—two tasks
ideally suited to Web-based computing.
Collaboration, document updates, document exchange, e-mail broadcasts, calendars, database
lookup, workflow systems are tasks that can be handled by Web-based systems. With Web-based
computing you can customize the tasks and combine them in different ways.
This chapter serves not only as an introduction to back-end systems as they pertain to e-commerce
sites, but also as an introduction to how you build Web-based offices in general. First, we look at the
benefits of running a Web-based office. Then, in the "The SBS Back-End System" section, we
describe the features of a fictitious back-end system for managing an e-commerce site. The SBS
system described in this chapter provides pages for updating product information, editing the Web,
picking up and processing orders, and editing e-mail templates.
-253-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Benefits of Web-Based Offices
One of the main benefits of Web-based offices is the ability to customize the automation of office
tasks. Web-based offices can be built, rebuilt, and adapted to changing operational needs in the way
you update Web sites constantly to keep up with changing business needs.
The Web-based development environment gives businesses of all sizes the ability to build
sophisticated solutions customized to meet their current and future organizational needs. Unlike
approaches based on canned solutions, where the software purchased dictates offices tasks, custom
solutions enable office workers to perform tasks more efficiently. For example, suppose an
equipment retailer needs to provide office workers with the ability to search for customers whose
orders are late so the customers can be alerted by e-mail of the delays. The retailer with a canned
solution would need this feature built into the canned system. But the software may not have that
feature, or the e-mails that go out may need to be combined with information in another product. The
result is an inability to serve the customer to the detriment of customer and business.
Although canned solutions help businesses in the short term, custom solutions (provided they can be
modified quickly) offer greater long-term benefits. The Web-approach frees the organization from the
canned software mentality that systems cannot be given new features when necessary, and that
manual intervention is needed when transferring information back and forth between different
software products.
Another benefit of Web-based offices is that the business tasks are performed using a Web browser.
This frees office workers to use any computer in the office or work from home. It means a PC can
break or fail without disrupting business operations.
Customizing internal business operations using Web-based systems is an empowering experience
that allows you to build a more flexible organization and take advantage of market opportunities that
you may otherwise not explore. For example, any opportunity that includes the need to exchange
data with another organization can benefit from the construction of custom systems. Most legacy
systems suffer from development cycles of 90 to 120 days or more. New features can be added to a
Web-based system in as little as a day.
Customizing your back-end systems also forces you to analyze and understand the operations in
your organization. When building a back-end system you often find needless operations that you can
streamline. It's a cleansing process with benefits that often yield unexpected rewards. For example,
in the course of constructing a back-end system to process customer support requests you might
find you already have the answers to many customer questions and can make them available over
the Web. Reexamining your operations often leads to improvements in customer relations and
business operations. The construction of Web-based systems involves re-examining your operations.
Perhaps the most interesting aspect of back-office Web development is that the quick-development
process often makes implementation of advanced back-office pages less costly than purchasing outof-the-box enterprise solutions. You can have your cake and eat it too. On one hand you get the
customization capabilities. On the other hand you get something affordable. What many don't realize
is that the development of back-end Web pages is often easier than building Web sites. Here's why:
ƒ Back-end systems rarely need to support as many users as a high-access Web site. You
may have thousands of users on your Web site but you don't have that many people on
a back-end system.
ƒ Back-end systems do not need large amounts of graphical design. Most often a list of
tasks is all that's needed (although placing a nice header and footer on your pages
won't hurt.)
ƒ Only one task needs to be handled per Web page. This separation of duties makes the
pages fast to develop or modify. HTML editors are not needed either. Contrast this with
-254-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
the pages of an advanced Web site that often perform many tasks (providing users
product navigation, search, product detail, and promotional information, for example)
and may contain many graphical design elements.
Back-End Design Considerations
The design of back-end systems varies from organization to organization. And, of course, the design
of such systems involves many topics unique to your situation, so they cannot be discussed in this
book. But some common issues directly related to Web development are important and worthy of
mention. In this section we discuss the need to distribute control between users, the need to
modularize and document development work, and the need to protect your code from future changes
outside of your control.
Distributing Control
If you operate in a large organization you need to be keenly aware of security and that different
departments may need control over their own information, their own systems, and perhaps their own
development. A technical support department may not want to give other departments the ability to
edit or change its information. A sales department does not want to give other departments access to
its coupon databases. In these business environments you want to set up copies of HTML/OS in
each department. This topic is discussed in the "Serving Network Databases" section of Chapter 15,
Database Networking. In general, each department can maintain its own databases and give access
to its respective databases by opening read or read/write access to other users in the organization
on a per-request basis.
Keeping Tasks Limited
When building back-end system components keep the tasks simple. You may not be the last one
modifying the page.
As an example, if you are building a page to collect incoming orders you can provide a page to view
orders and another to pick up and print orders. Using different pages for each of the tasks makes it
easier for someone else to make changes later.
Document, Document, Document
It is important to document your Web pages. Although HTML/OS code is easy to read (since it's
English-like), you should document each task in your code and add a description within your code of
the fields used in the database. This is helpful to you or anyone modifying the page later. Liberal use
of Read Me files is helpful as well. If code in your back-end application connects to a database in
another department, make sure the parameters used to access the network database are clearly
accessible so they can be changed if necessary.
Using HTML Editors Carefully
HTML editors can build Web pages that don't work properly on some browsers. Since back-end
systems rarely go through the compatibility testing that Web sites go through, it may make more
sense to avoid using HTML editors. There is little excuse for building systems that suddenly don't
work when you try to log into them with a Macintosh, for example. See the accompanying note,
"Don't Tolerate Browser Bigotry."
Don't Tolerate Browser Bigotry—People who tout one browser over another, or one computer
over another hurt organizations since they place greater importance on their biases than on
whether a system works. Such people should be reminded that companies are better served by
making the Web their standard than the software from a single company.
-255-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Assume Databases Will Change
Design your databases and Web pages with the expectation that fields will be added in the future. To
do this, avoid using Overlay tags that are dependent on field structure. In this book we avoided tags
that depend on field structure so this is not a tough thing to watch out for, but such dependent
Overlay tags do exist. For example, you want to use DBFIND, not DBSEARCH. The DBFIND tag, as
you have seen in the code examples in this book, has a field list parameter. If you add new fields to
the database, the tag returns the same result. This is not true for DBSEARCH, which relies on the
order of the fields in the database.
The SBS Back-End System
In this section you look at the back-end pages for managing a fictitious e-commerce site that sells
retail products. The system is called Simple Back-End System or SBS. You can download the SBS
system along with its front-end Web site from the companion Web site of this book. (See the Preface
for details.) The home page of the SBS system is shown in Figure 21.1.
Figure 21.1: The SBS System Automates the Tasks Needed to Maintain a Web Site.
The SBS system provides four main functions. A Product Database section allows marketing staff to
edit, delete, and add products. A Web Site Pages section allows staff in a production department to
edit and view the HTML documents that make up parts of the Web site. An Order Management
section allows an order processing department to pick up and manage orders. A Marketing section
allows staff to edit the automated e-mail template used to define confirmation e-mails.
The SBS pages reside in the /apps/sbs/bin directory. The Web site itself resides in the
/apps/sbs directory. A directory list of the SBS pages is shown in Figure 21.2.
-256-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 21.2: The SBS System Uses Only the HTML Documents Shown in this Listing.
The Order State—The order state is a code that represents the status of an order in a workflow
system. When an order first arrives it is in one state. After it is taken out of inventory it is in another
state. After it has been shipped it is in yet another state. The order states are stored in an
orderstates.txt file. You associate an order coder with each order to use when you are
building screens that list products at specific points of the workflow process.
In this section you will be looking at several SBS documents to get a taste of how to build back-end
systems. The files used by the SBS system and their functions are shown in the following list. The
files shown in bold are described in this chapter.
ƒ
menu.html
editlist.html
Back-end options menu (see Figure 21.1)
List products to edit
editproduct.html
Edit product in products database
noyes.html
Delete product confirmation page
editfile.html
Generic text file editor
download.html
Download orders
orderstates.txt
Display valid order states
orders.html
Select order by order state
orderqueue.html
View orders in an order state
orderqueue_form.txt
HTML code to display a search form
orderqueue_display.txt
HTML code to display an order
orderedit.html
Edit order record
orderprint.html
View print-ready order record
orderview.html
View archived order
-257-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The SBS system uses two databases: a products database and an orders database. The
products database stores products that can be ordered on the Web site. The fields for the
products database are as follows:
ƒ
PRODNAME
Name of product
PRODNUMBER
Product number
PRODPRICE
Product price
PRODTITLE
Product title
PRODDESC
Product description
PRODIMAGEFILE
Product image file
PRODSHIPPING
Product shipping surcharge
The orders database stores the information about each order (such as credit card number and
shipping address) as well as the individual items ordered (product name and price, for example).
Each record has room for only a single item ordered, so an order with four items, for example, would
occupy four records in the database. Note that many ordering systems would place the items
ordered in a separate orderitems database. To keep this system simple a single database is used.
The fields for the orders database are as follows:
ƒ
ORDER_NUMBER
Order number
ORDER_STATUS
Order state
ORDER_EMAIL
Order e-mail address
ORDER_DATE
Date order placed
ORDER_SHIPTYPE
Shipping method requested
ORDER_CC
Credit card number
ORDER_CCEXP
Credit card exp date
ORDER_NAME
Credit card holder name
ORDER_ADDRESS
Shipping street address
ORDER_CITY
Shipping city
ORDER_STATE
Shipping state
ORDER_ZIP
Shipping Zip code
ORDER_PHONE
Shipping phone number
ORDER_TAXABLE
Order item tax flag
ITEM_ID
Order item id
ITEM_QTY
Order item quantity
ITEM_NAME
Order item name
ITEM_PRICE
Order item price
ITEM_OPTIONS
Order item options
ITEM_SHIPPING
Order item shipping charge
-258-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ITEM_DELETE
Order item delete flag
Editing the Product Database
E-commerce systems with many products require a way for office staff to edit the list of products.
Product information can be stored in a legacy spreadsheet or database and periodically uploaded,
which requires maintaining two product databases. An alternative approach is to let office staff edit
the Web-based product database directly. That is done here. Two menu options for editing the
product database are provided as shown in Figure 21.1. The code for this menu page is as follows:
<html>
<title>SBS Home Page</title>
<br><font size=6 face=arial>
<b>S B S</b></font><br>
<b>S</b>imple <b>B</b>ack-End <b>S</b>ystem.</font><br><br>
<table border=0><tr><td valign=top>
<ul><font size=2 face=arial,helvetica>
Product Database<ul>
<li><a href="editlist.html">Edit/Delete Product</a>
<li><a href="addproduct">Add Product</a><br><br>
</ul>
Web Site Pages<ul>
<li><a href="../index.html">View Web Site</a>&nbsp;&nbsp;
<li><a href="editfile.html" name=e value="data/settings.txt">
Settings</a>
<li><a href="editfile.html" name=e value="index.html">Edit Home
Page</a>
<li><a href="editfile.html" name=e value="category.html">Edit
Product Category Page</a>
<li><a href="editfile.html" name=e value="product.html">Edit
Product Detail Page</a>
<li><a href="editfile.html" name=e value="search.html">Edit
Product Search Page</a>
<li><a href="editfile.html" name=e value="cart.html">Edit
Shopping Cart Page</a>
<li><a href="editfile.html" name=e value="cart_empty.html">Edit
Empty Cart Page</a>
<li><a href="editfile.html" name=e value="checkout.html">Edit
Checkout Page</a>
<li><a href="editfile.html" name=e value="confirm.html">Edit
Thank You Page</a><br></ul>
</td><td valign=top>
-259-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<ul><font size=2 face=arial,helvetica>
Order Management<ul>
<li><a href="orders.html">Check Orders</a></ul><br><br>
Marketing<ul>
<li><a href="editfile.html" name=e value="data/email.txt">Edit
E-mail Template</a>
</ul></ul></td></tr></table>
</html>
<<overlay addproduct
PRODNAME="Untitled."
PRODNUMBER=""
PRODPRICE=""
PRODTITLE=""
PRODDESC=""
PRODIMAGEFILE=""
PRODSHIPPING=""
sc_conf=""
sc_rec="ERROR"
GOTO "editproduct.html"
>>
The menu page contains the Edit/Delete Product and Add Product options. When users select the
Add Product option, they are directed to the addproduct on-click Overlay, which sets up the
variables for a new product record and launches editproduct.html. The editproduct.html
page is used for editing new as well as existing records.
When users select the Edit/Delete Product option, they are directed to editlist.html to select a
product record. The editlist.html page is as follows:
<<
IF button="Menu" THEN button="" GOTO "menu.html" /IF
sc_products=DBFIND("/apps/sbs/data/products",
"",1,500,'RECORD,PRODNAME,PRODNUMBER,PRODPRICE')
>>
<html>
<form method=post action=<<page>>>
<input type=text name=sc_search size=20>
<input type=submit name=button value="Menu">
</form>
<table border=1 cellpadding=2 cellspacing=0>
<tr>
<td><b>Product Name</b></td>
-260-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<td><b>Product ID</b></td>
<td align=right><b>Price</b></td>
<td>&nbsp;</td>
</tr>
<<
FOR NAME=sc_products ROWNAME=sc_product DO
DISPLAY
'<tr>'+
'<td><a href="editproduct.html" name=sc_rec '+
'value='+sc_product[1]+'>'+sc_product[2]+'</a></td>'+
'<td>'+sc_product[3]+'</td>'+
'<td align=right>'+sc_product[4]+'</td>'+
'<td align=right><a href="noyes.html" name=sc_delno '+
'value='+sc_product[1]+'>delete</a></td></tr>'
/DISPLAY
/FOR
</table>
</html>
AM
FL
Y
>>
<html>
TE
The first Overlay on the page uses DBFIND to fill the sc_products variable with the records in the
products database. Then a FOR loop displays the products on the page. Each product name is a
hypertext link to editproduct.html. In addition, a Delete hypertext link, which links to the
noyes.html page, is placed next to each product name. The page is a four-line confirmation page
that gives the user a chance to confirm the delete. The code is as follows:
<title>Four-Line Confirmation Page</title>
<< sc_stat=DBGETREC("/apps/sbs/data/products",sc_delno)>>
Delete <u><<prodname>></u>?<BR><BR>
<a href="doit" name=sc_delno value=<<sc_delno>>>Yes</a>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<a
href="editlist.html">No</a>
</html>
<<overlay doit
stat=DBDELETE("/apps/sbs/data/products",sc_delno)
sc_delno=""
GOTO "editlist.html"
>>
-261Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
When a user clicks Delete, the sc_delno variable is set. The noyes.html page asks whether to
delete the record. If the user clicks Yes, the doit on-click Overlay is called and the DBDELETE tag
deletes the record. If the user clicks No, the user returns to the editlist.html page.
When the user selects a record on the editlist.html page, the editproduct.html page
launches. This is the Web page staff use when editing new or existing records. The code for the
page follows:
<html>
<title>Eight-line SBS Product Editor</title>
<<sc_stat=DBGETREC("/apps/sbs/data/products",sc_rec)>>
<br><form method=post action=editrec>
<table border=1 cellpadding=2 cellspacing=0>
<tr><td colspan=2 align=right>
<input type=submit name=button value="Save">
<input type=submit name=button value="Quit">
<input type=submit name=button value="Quit To Products"></td>
</tr><tr><td><b>Name</b></td><td>
<input type=text name=prodname size=20></td></tr>
<tr><td><b>ID</b></td><td>
<input type=text name=prodnumber size=10></td></tr>
<tr><td><b>Price</b></td>
<td> <input type=text name=prodprice size=20></td></tr>
<tr><td><b>Title</b></td><td>
<input type=text name=prodtitle size=20></td></tr>
<tr><td valign=top><b>Description</b></td><td>
<textarea name=proddesc rows=4 cols=50></textarea></td></tr>
<tr><td><b>Image file</b></td><td><input type=text
name=prodimagefile size=20></td></tr>
<tr><td><b>Shipping</b></td><td>
<input type=text name=prodshipping size=20></td></tr>
</table></form>
</html>
<<overlay editrec
IF button="Quit" THEN
GOTO "menu.html"
ELIF button="Quit To Products" THEN
GOTO "editlist.html"
ELSE
IF ISINTEGER(sc_rec) THEN
stat=DBEDIT("/apps/sbs/data/products",sc_rec)
-262-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ELSE
stat=DBADD("/apps/sbs/data/products")
/IF
/IF
GOTO PAGE
>>
At the top of the page a DBGETREC loads the record into the variable environment. Once loaded, it
fills the HTML form displayed on the page. A similar technique is used in the "Eight-Line Database
Record Editor" section of Chapter 11, The Web Database.
The HTML form on the page is preceded with Save, Quit, and Quit to Products buttons. When the
user clicks a button, the editrec on-click Overlay runs. If the user clicks Quit or Quit to Products,
the user is directed to menu.html or editlist.html respectively. If the user clicks Save, the
sc_rec variable is inspected. If it's a number, you assume you are editing an existing record and
hence, you save the record with a DBEDIT tag. If not, you can assume you are editing a new record
and hence you create the new record with a DBADD tag.
Editing Documents
The SBS system also includes options to edit the HTML documents that make up the Web site. To
provide this feature you build an editfile.html page similar to that described in the "Six-Line
Text Editor" section of Chapter 8, Building Text Editors. The page is as follows:
<< sc_editfile="/apps/sbs/"+e
COPY FILE=sc_editfile TO sc_editdata /COPY
>>
<html>
<form method=post action=editstuff>
<table border=0 cellpadding=0 cellspacing=0>
<tr><td><b>File: <<e>></b></td>
<td align=right>
<input type=submit name=button value="Save">
<input type=submit name=button value="Reload">
<input type=submit name=button value="Menu"></td>
</tr><tr><td colspan=2>
<textarea name=sc_editdata rows=17 cols=69></textarea>
</td></tr></table>
</form>
</html>
<<overlay editstuff
IF button = "Reload" THEN
GOTO PAGE
ELIF button = "Save" THEN
-263-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
COPY sc_editdata TO FILE=sc_editfile /COPY
GOTO PAGE
ELIF button = "Menu" THEN
GOTO "menu.html"
/IF
GOTO PAGE
>>
The variable e, which contains the name of the file being edited, is set in menu.html when a file is
selected. The editor assumes filenames in menu.html are specified relative to the /apps/sbs
directory. It sets sc_filename at the top of the file accordingly. Like the text editor described in
Chapter 8, when a user clicks an option, the appropriate action is taken in the on-click Overlay at the
bottom of the document.
This editor gives product staff quick access to the HTML documents that make up the Web site. It is
also used for editing the e-mail template—the file that defines the automated e-mail message sent
when a product is ordered. It is always a good idea to give office staff the ability to edit this document.
The Ordering System
Ordering systems gives office staff the ability to manage incoming orders. In this section we describe
a sophisticated ordering system that includes workflow. The concept of workflow was first discussed
in the "Document Collaboration" section of Chapter 8, Building Text Editors, which described a
system based on text files. Here a database-based system is described.
The workflow system works as follows: All orders include a field called order_status. The field is
set to QUEUE when an order is placed. The field can also have the values CANCEL, SHIP, HOLD,
or DONE. When office staff members pick up incoming orders, they can selectively view orders
depending on the order's order state. They can view orders in the incoming queue (orders with
order_status equal to QUEUE), or they can view orders on hold, orders completed, and so on. In
addition, staff members can move orders between one state and another. After they ship an order,
for example, they would move the order to the DONE state. If they cancel orders, they move them to
the CANCEL state. If they must put an order on hold, they move it to the HOLD state.
From a programming perspective, the process of moving orders between order states is simply a
matter of changing the contents of the order_status field. Viewing orders of a specific order state
is a matter of setting up the query on the orders database so the page displays only those records
with an order_status equal to the specified order state.
In the SBS system, the order states are stored in the orderstates.txt file. The file is a twocolumn, delimited text file with the following contents:
QUEUE,"Input Queue"
SHIP,"Orders To Ship"
DONE,"Orders Shipped"
CANCEL,"Cancelled Orders"
HOLD,"Order On Hold"
-264-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
The file defines five order states. Different workflow systems can use different sets of possible states.
The first column contains the codes you store in the order_status field of the orders database.
The second column contains a text description you used on Web pages and in links to describe the
order state. The ordering system is launched from the menu.html page, which provides a Check
Orders hypertext link to orders.html.
The orders.html page gives office workers the ability to access orders in different order states.
They may wish to view only incoming orders. At other times, they may need to search completed
orders. The orders.html page used in this SBS system is dynamically created using the
orderstates.txt file. The code for the page is as follows:
<html>
<title>Three-Line SBS Order States Page</title>
<< COPY FILE='orderstates.txt' TS=',' TO sc_states /COPY >>
<a href="menu.html">Menu</a>
<ul>
<< FOR NAME=sc_states ROWNAME=sc_row DO
DISPLAY '<li><a href="orderqueue.html" '+
'name=sc_state value="'+sc_row[1]+'">'+
sc_row[2]+'</a>'+LF /DISPLAY
/FOR
>>
</ul>
</html>
The page reads the orderstates.txt delimited text file into the sc_states variable. Then, using
a FOR loop, it displays hypertext links to each state. When the user clicks a state, the sc_state
variable is set to the order state and the page orderqueue.html runs.
The orderqueue.html page is the main page used to view orders. The page lists products for a
specified order state and contains options to move orders to different order states. A sample screen
shot is shown in Figure 21.3. The code for the page follows.
Figure 21.3: The Workflow System Used by SBS Includes a Page Where Staff Can Move Orders among
Different Order States.
<<
-265-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
IF sc_day="ERROR" THEN
sc_searchdate=TODAY
ELSE
sc_searchdate=sc_month+'/'+sc_day+'/'+sc_year
/IF
sstr='ORDER_DATE>="'+sc_searchdate+'" AND '+
'ORDER_STATUS="'+sc_state+'"'
flist='ORDER_NUMBER,ORDER_NAME,ORDER_DATE,ITEM_QTY,'+
'ITEM_PRICE,ORDER_TAXABLE,ITEM_DELETE'
sc_orders=DBFIND('/apps/sbs/data/orders',sstr,1,100,flist)
sc_linkstates=GETCOLNOTEQ(sc_states,1,sc_state)
>>
<html>
<title>Thirty Five-Line Order List Page With Workflow</title>
<b><<GETTABLE(GETCOLEQ(sc_states,1,sc_state),2,2,1,1)>></b><br>
<a href="menu.html">Menu</a>&nbsp;&nbsp;<a
href="orders.html">Orders</a>
<< DISPLAY FILE="orderqueue_form.txt" /DISPLAY >>
<table border=0>
<<
i=0 sc_total=0 lastorder="ERROR"
IF sc_orders!="" THEN
DISPLAY "<tr bgcolor=#EEEEEE>"+
"<td><u>Order Number</u></td>"+
"<td><u>Name</u></td><td><u>Date</u></td></tr>"
/DISPLAY
FOR NAME=sc_orders ROWNAME=sc_order DO
IF lastorder != sc_order[1] THEN
IF lastorder!="ERROR" THEN
EXPAND FILE="orderqueue_display.txt" /EXPAND
/IF
sc1=sc_order[1] sc2=sc_order[2] sc3=sc_order[3]
lastorder=sc_order[1]
/IF
/FOR
IF lastorder!="ERROR" THEN
EXPAND FILE="orderqueue_display.txt" /EXPAND
/IF
/IF
>>
-266-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
</table>
</html>
<<OVERLAY changestate
sc_newstate=CUT(CHOPRIGHT(sc_orderdata,'|'),'|')
sc_ordernum=CUT(CHOPLEFT(sc_orderdata,'|'),'|')
sstr='ORDER_NUMBER="'+sc_ordernum+'"'
stat=DBFILL('/apps/sbs/data/orders',sstr,'ORDER_STATUS',sc_newst
ate)
GOTO PAGE
>>
The Underlay at the top of orderqueue.html page composes a search string to determine which
orders should be listed on the page. The DBFIND tag extracts orders. The instruction
sc_linkstates=GETCOLNOTEQ(sc_states,1,sc_state) fills sc_linkstates with a column
of order states with all states other than the current state. The variable is used later in the code to
generate a list of order states to which an order may be moved.
Below this Underlay is the display part of the Web page. The page uses the following Overlay to
generate its page title dynamically. Depending on the current order state, a different title is displayed.
Note that sc_states was filled in a prior page with the contents of the orderstates.txt file, so
the Overlay effectively displays the title associated with the current order state.
<<GETTABLE(GETCOLEQ(sc_states,1,sc_state),2,2,1,1)>>
Below this Underlay an HTML form is placed on the page, so the user can list only orders since a
specified date. The HTML form is stored in the orderqueue_form.txt file and placed in the page
using the instruction:
DISPLAY FILE="orderqueue_form.txt" /DISPLAY
The contents of this file could have been written directly into the page, but here, as a way to cut
down the size of the page, the code for the search box is stored in a separate file. The contents of
the file follow:
<form method=post action=orderqueue.html>
Show orders after <select name=sc_month>
<option value=01>Jan<option value=02>Feb<option value=03>Mar
<option value=04>Apr<option value=05>May<option value=06>Jun
<option value=07>Jul<option value=08>Aug<option value=09>Sep
<option value=10>Oct<option value=11>Nov<option value=12>Dec
</select>
<select name=sc_day>
<option value=01>01<option value=02>02<option value=03>03
<option value=04>04<option value=05>05<option value=06>06
<option value=07>07<option value=08>08<option value=09>09
-267-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<option value=10>10<option value=11>11<option value=12>12
<option value=13>13<option value=14>14<option value=15>15
<option value=16>16<option value=17>17<option value=18>18
<option value=19>19<option value=20>20<option value=21>21
<option value=22>22<option value=23>23<option value=24>24
<option value=25>25<option value=26>26<option value=27>27
<option value=28>28<option value=29>29<option value=30>30
<option value=31>31</select>,
<select name=sc_year>
<option value=2001>2001<option value=2002>2002<option
value=2003>2003
<option value=2004>2004<option value=2005>2005<option
value=2006>2006
</select>
<input type=submit name=button value="Go">
</form>
When the user specifies a date with this HTML form and clicks the Go button, the page redisplays, a
new search string is calculated, and the orders after or on the selected date are displayed.
Below this portion of the page the orders are displayed. A FOR loop loops across sc_orders, the
records obtained using the DBFIND tag in the Underlay. This FOR loop is a bit tricky since the orders
database contains records for each order item, not simply one record per order. You want to display
only one link per order. To do that you use a variable called lastorder, which is initialized to
ERROR. As you loop across sc_orders, when the order number of the order in the loop changes
from the previous, you display the order line. The code to display the order line could be placed in
the Web page, but to save space, the code is expanded into the Overlay using the EXPAND tag. The
code is stored in the orderqueue_display.txt file. The code is as follows:
DISPLAY
'<tr><td>'+sc1+'</td><td>'+sc2+'</td><td>'+sc3+'</td></tr>'+
'<tr><td colspan=3><br>Options: '+
'<a href="orderedit.html" name=sc_ordernum value="'+sc1+'">'+
'Edit order</a>&nbsp;|&nbsp;'+
'<a href="orderprint.html" name=sc_ordernum value="'+sc1+'">'+
'Print order</a>&nbsp;|&nbsp;'+
'<a href=orderview.html name=sc_ordernum value="'+sc1+'">'+
'View original</a>'+
'<br>Move To:'
/DISPLAY
DISPLAY
LAYOUT(sc_linkstates,'&nbsp;|&nbsp;',
-268-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
'<a href="changestate" name=sc_orderdata value="',
[1],'|',sc1,'">',[2],'</a>')
/DISPLAY
DISPLAY
'</td></tr><tr><td colspan=3><hr size=1></td></tr>'
/DISPLAY
This code displays the order number and the name of the contact associated with the order. Below it,
the Edit, Print, and View options display. Each option is a link to a different Web page. Below that
LAYOUT tag displays a series of links, each connecting to the changestate on-click Overlay. When
the user clicks a link, the sc_orderdata variable is set to the order number followed by a vertical
bar and the name of the desired order state. The changestate on-click Overlay contains the
following code:
sc_newstate=CUT(CHOPRIGHT(sc_orderdata,'|'),'|')
sc_ordernum=CUT(CHOPLEFT(sc_orderdata,'|'),'|')
sstr='ORDER_NUMBER="'+sc_ordernum+'"'
stat=DBFILL('/apps/sbs/data/orders',sstr,'ORDER_STATUS',sc_newst
ate)
GOTO PAGE
Using the CUT tag and CHOPRIGHT tags the new order state and order number are extracted from
the sc_orderstate variable. Then a search string is set to select the records belonging to the
specified order number and used in the DBFILL tag to change the ORDER_STATUS variable to its
new order state. The DBFILL tag is a DB tag used to change a particular field across a selection of
many records. It is described in the knowledge base on the Aestiva Web site at
http://www.aestiva.com/support.
Other Potential SBS Options
You can expand the SBS back-office application to provide additional functionality. For example, you
can limit access to the pages with login pages like those described in the Chapter 9, Building Login
Pages. By adding a login page, you can limit the tasks office staff can perform.
Order options may be split between a shipping department, which can complete an order, place it on
hold, or send it back to the input queue, but not edit the order—a task reserved for order takers or
managers. Tasks for the computer department can be added so they can issue login IDs and access
rights to different Web pages (applications). Additional options to manage additional features on the
Web site may also be needed.
Another option that you can add to the SBS system is an e-mail broadcast page. The option can give
individuals in the marketing department the ability to send mass e-mails to people who ordered
before. The Web pages to send a broadcast e-mail are divided into the following three tasks:
ƒ
Database selection— Database selection involves setting up query pages that pull
records from databases containing customer or prospective customer information.
The purpose of these pages is to create a list of e-mail addresses and related
-269-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
ƒ
ƒ
information for use in an e-mail merge broadcast system. These pages are built using
the same techniques described in Chapter 12, Building Query Pages.
E-mail template management— E-mail template management is a matter of giving
the user the ability to set up a directory of text files containing the kinds of e-mails you
wish to broadcast. That directory is displayed on the screen in a pull-down menu. The
files can contain placeholders as described in the "Using an E-Mail Template" section
of Chapter 9, Building Login Pages, so they can be used with an e-mail broadcast
system.
E-mail merge broadcast— E-mail broadcast systems use a specified e-mail
template file and a variable or file containing e-mail addresses and information that
need to be merged into the e-mails (such as customer name) and broadcasts it to the
e-mail addresses on the list.
As an example of e-mail broadcasting, suppose the name of your template file is stored in the
mytemplate variable and your e-mail addresses are stored in a two-column variable called
myemails with the e-mail addresses in the first column and customer names in the second. Then
the page to broadcast e-mails can be as follows:
<< # Ten-Line E-Mail Broadcast Page
# Incoming Variables:
# Email variable: myemails
# Template file: mytemplate
# Subject: mysubject
/#
IF ISINTEGER(email_row)="FALSE" THEN email_row=1 /IF
IF myemails[1,email_rowno] != "" THEN
temp=myemail[1, email_rowno]
mymetalink='<META HTTP-EQUIV="REFRESH"' +
'CONTENT="1;URL='+PAGE+'">'
myletter=REPLACEALL(myletter,"[name]",
myemails[2,email_rowno])
MAIL myletter TO ADDRESS=temp SUBJECT=mysubject /MAIL
msg="Message #"+email_row+" sent."
email_rowno = email_rowno + 1
ELSE
mymetalink=""
email_row="ERROR"
msg="Broadcast Complete"
/IF >>
<html>
<title>Ten-Line E-Mail Broadcast Page</title>
<<mymetalink>>
<<msg>>
</html>
-270-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
This page sends one e-mail message and then redisplays itself, each time sending the next e-mail in
the myemails variable, until all e-mails have been sent. The tricky part of this page is knowing you
cannot simply send mass e-mails using a FOR loop since this places too much load on an outgoing
mail server and, further, may take so long the Web page may time out. Instead, e-mails should be
sent one at a time with a couple of seconds of a pause between them. Sending out an e-mail,
redisplaying the page, and repeating the process with the next e-mail is a good way to broadcast
multiple e-mails.
The code for this page begins by initializing email_row, if necessary. This is the row in the
myemails variable (which contains a list of e-mail addresses) to send. Then you see whether the email in the row is empty. If it's not empty, another e-mail needs to be sent: In this case the
mymetalink variable is filled with the HTML needed to redisplay the page. The HTML command to
redisplay a Web page is written as follows, where URL is the name of the Web page to display and
seconds is the number of seconds to wait before bouncing to the specified URL. The command can
be placed just below the <TITLE>...</TITLE> tags in the HTML document.
<META HTTP-EQUIV="REFRESH" CONTENT="seconds;URL=URL">
Summary
AM
FL
Y
After the mymetalink variable is set, an e-mail message is composed and the message is emailed
with the MAIL tag. Note the use of the REPLACEALL tag for swapping [name] with the name of the
recipient. You can extend this e-mail merge feature by adding additional REPLACEALL instructions.
TE
In this chapter we covered the nature of back-end management systems and viewed a sample
system for managing an e-commerce site. You learned that back-end systems and Web-based
offices are built from the same components already discussed in this book. Indeed, an understanding
of the techniques used in this book is what you need to build both advanced Web sites and Webbased offices.
In Appendix C, The Next Generation: Web-Based Products, you look beyond advanced Web
development—to the next generation of Web-based computing. You learn that Web-based
computing is also about building products that can be sold and installed in hosting accounts all
across the Web.
Exercises
In the exercises that follow you apply the concepts described in this chapter. Answers to these
exercises are provided on this book's companion Web site as described in the book's Preface.
Exercise 1
The "Other Potential SBS Options" section of this chapter included a ten-line e-mail broadcast page.
Complete the application by setting up a page that allows you to select templates stored in a
directory, define a subject line for your e-mails, and launch the e-mail broadcast application.
Exercise 2
Add a login page to the SBS system discussed in this chapter. Have the page divide marketing staff
from order processing staff. Replace menu.html with one that works with your login page to give
different users different sets of options.
-271Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Exercise 3
Add a download option to the SBS system so orders being viewed can be downloaded. Hint: Add the
option to the top of orderqueue.html page. When a user clicks the link, create a delimited text file
with the orders and use the FILEPUSHLINK tag to push the file to the user. Information on setting up
download options and the FILEPUSHLINK tag is available in Aestiva knowledge base at
http://www.aestiva.com/support.
-272-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Appendix A: HTML/OS Resources
Aestiva Web site
http://www.aestiva.com/
The home page for Aestiva. Features product and sales information on HTML/OS and other Webbased products and information on Aestiva.
Companion Web Site for Advanced Web Sites Made Easy
http://www.topfloor.com/advanced/
Information of interest to readers of this book. Includes answers to exercises, sample code, book
notes, and reviews.
30-Day Free Trial Sign-Up for Advanced Web Sites Made Easy
http://dev.aestiva.com/freetrial/
The sign-up page for your reserved copy of HTML/OS.
User Forum for Advanced Web Sites Made Easy
http://dev.aestiva.com/advanced/forum/
An interactive user forum for getting answers to development questions. Reserved for readers of this
book.
Aestiva Freeware Library
http://www.aestiva.com/freeware/
A code library containing dozens of freeware applications. Includes sample e-commerce sites, Webbased utilities and back-office applications.
Aestiva Knowledge Base
http://www.aestiva.com/support/
A database of answers to development questions. Includes answers to questions on more topics,
sample code, and reference pages on tags not described in this book.
Aestiva Clips Library
http://www.aestiva.com/freeware/
A code library containing dozens of document sets for performing specific tasks. Includes login
pages, sample HTML forms processing pages, database pages, HTML editors, and more.
-273-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Appendix B: Major HTML Tags
Overview
This appendix provides a crash minicourse in HTML. Without a knowledge of these tags you won't
become an accomplished Web developer even if you're an experienced programmer, because Web
developers must think in HTML. Producing on-the-fly documents is about visualizing the end result—
documents containing HTML tags rendered by the browser.
Luckily, almost anyone can get comfortable with HTML in a few hours. Most HTML tags don't work
across different kinds of browsers, browser versions, and different hardware. Learning HTML is not
about learning all the tags available. It's about learning the limited number of HTML tags that work.
To learn HTML, take the tags listed here and play with them in an HTML document on your copy of
HTML/OS. Spend a few hours testing these tags. Test all the tags until you feel comfortable with
them. After you're done you'll have enough knowledge of HTML to use this book. The HTML tags
discussed in this appendix fall into the following four categories, which are discussed in the sections
that follow:
ƒ
ƒ
ƒ
ƒ
Text tags
Page attributes
HTML tables
Links and images
After learning these tags, you will also want to learn how to place input boxes and other input
elements on the page, which involves another set of HTML tags. They are described in Chapter 10,
HTML Forms Processing.
In this appendix you start by looking at how HTML tags are written. Then you look at the most
common tags used, presented in the four categories we've just noted. Note that the tags and
attributes presented in this appendix are not intended to be complete but serve as a quick
introduction to HTML. For example, HTML style tags and tags that work only in the latest browsers
are not listed here. For further information see the resources listed in the "Further Reading" section
at the end of this appendix.
Writing HTML Tags
By convention all HTML tags begin with a < character and end with a > character. The tags are
sequences of text placed in HTML documents, which are themselves text documents. The first word
in every HTML tag is the name of the tag. The name of the tag is then followed with one or more
name-value pairs. Most tags have a corresponding end tag. These ending tags always begin with </
characters followed by the name of the tag and the > character. They never contain name-value
pairs. As an example, consider the following HTML:
<font size=3 color=red face=helvetica>Hello World</font>
This prints "Hello World" on the page. The tag preceding these words is <font>. It has three namevalue pairs. The names in these name-value pairs are called attributes and are predefined for the
tag. In this <font> tag, the first attribute is size. It is set to 3. The second attribute is color. The
-274-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
last attribute sets the typeface of the text. The ending tag to the <font> tag is the </font> tag. The
pair of tags affects only the text between them.
As a rule, all HTML tags are built this way. To fully understand how a tag is written, you need to
know the name of the tag and the possible name-value pairs it may contain. The order of the namevalue pairs in the tag is unimportant and, as the examples in this appendix suggest, name-value
pairs are often optional—that is, you don't have to use all options. A tag's name and its attributes can
be written in either upper or lower case. HTML is not case-sensitive.
Text Tags
Text tags are the easiest tags in HTML. You use them to change how your text looks on the page.
The most important ones to know are discussed in the sections that follow.
<b>...</b>
Place text between <b> and </b> tags and your text will appear bold when the browser displays it.
Browsers also allow you to use the tags <i> and </i> to make your text italic, and <u> and </u> to
underline your text. These tags always appear in pairs.
Example: This is <b>bold</b> text.
<font size=size color=color face=typeface>...</font>
Text that you place between these tags resizes to a specified size and the color of the text changes
to the specified color. Browsers accept sizes that start at 1 (very small). The default size is 3. A very
large font would be 6. The color you specify can be certain, predefined English word such as red, or
a set of three two-digit hexadecimal numbers from 00 to FF representing the amount of red, blue,
or green in the color. White is #FFFFFF and black is #000000. The font tag can also contain a
typeface. Popular typefaces are Arial and Courier. The default font (if not specified) is Times Roman.
These tags always appear in pairs.
Example: This is <font color=red>red</font> text.
<center>...</center>
Place text between the <center> and </center> tags to center the line they are on. Funny thing
though, there are no <right>...</right> tags in HTML. Who knows why. In any event, we
describe right justification using HTML tables in the "HTML Tables" section, later in this appendix.
These tags appear in pairs.
Example: <center>This line is centered.</center>
<br>
You can place <br> at the end of lines or paragraphs as a page-break tag. Everything after the tag
appears on lines below the tag. This tag is needed since browsers ignore end-of-line characters in
raw HTML text, because end-of-line characters are different on the Macintosh, Unix, and Windows
systems. Since browsers work with all kinds of text files from all kinds of computers, they treat
conventional end-of-line characters (and sequences of these characters or spaces) as if they were a
single space. If you need a line break you must supply it. This tag does not need an ending tag.
Example: This is line one.<BR>This is line two.<BR>This is line three.
-275-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Page Attributes
Page attributes are HTML tags that tell the browser how to build the page. For example, you can tell
the browser what background color to use for the page, what to write in the title bar across the top of
a browser, and so on. The most common page attribute tags follow.
<html>...</html>
Put all of your text and HTML tags between these two tags. In other words, place the <html> at the
top of the document and </html> at the end of the document. Note that HTML/OS programming
can appear above and below these tags. The <html> and </html> tags surround the display
portion of the page.
<title>...</ title>
Follow the <html> tag with a line of text between the <title> and </title> tags. This is the text
you want to appear in the title bar of a visitor's browser. If you do not define a title for your document,
browsers will either leave the title line blank or use your document name as the title.
Example: <title>Home Page</title>
<body bgcolor=color> or <body background="image">
Follow the <title>...</title> line with one of these two tags if you want a background color or
image for the page. When specifying an image file, use a URL or a filename. Only .gif or .jpg
image types are supported universally on the Web. This tag does not need an ending tag.
Example 1: <body background="/images/mylogos.gif">
Example 2: <body bgcolor=white>
HTML Tables
HTML enables users to organize information in grids with rows and columns, meaning in tables. To
accomplish this, you need three kinds of tags: you use a <table> tag to surround the entire grid, a
<tr> tag to surround each row of a table, and a <td> tag to surround a cell containing data in the
table. <table> tags surround <tr> tags, which surround <td> tags, which surround text, images,
and content on the page. The following sections discuss each of these tags in more detail.
<table border=border bgcolor=color>...</table>
The <table> and </table> tags are the starting and ending tags for a whole table. The border
attribute in the tag sets the width of the border around the grid. Note that this pair of tags must
surround one or more table row and end table row tags, as described in the following section. These
tags appear in pairs.
<tr bgcolor=color>...</tr>
The <tr> and </tr> tags are the starting and ending tags for rows in a table. The bgcolor attribute
sets the background color for all of the cells in the row. It overrides any bgcolor setting specified in
the table tag. That is, a row's bgcolor attribute defines the color for the current row, while the
table's bgcolor attribute defines the color for remaining rows. Note that this pair of tags must
surround one or more table cell tags, as described in the next section. These tags appear in pairs.
-276-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<td align=value valign=value color=color>...</td>
The <td> and </td> tags are the starting and ending tags for a cell in a row in a table. The
bgcolor attribute sets the background of the cell. It overrides any bgcolor setting specified in the
table row or table tags. These tags appear in pairs.
Example:
<table border=1>
<tr><td>Cell 1</td><td>Cell 2</td></tr>
<tr><td>Cell 3</td><td>Cell 4</td></tr>
</table>
Links and Images
Perhaps the most important aspect of HTML documents is their ability to contain hypertext links and
images. After all, the acronym HTML stands for HyperText Markup Language. Here you learn how to
write links and embed images in documents.
<img border=border src="image">
You use the <img> tag to display images inside a document. The image is the name of the image
you want to display. The border sets up a border around the image. Setting it to 0 removes any
border. This tag does not need an ending tag.
Example: <img border=0 src="/images/logo.gif">
<a href="LINK">...</a>
The href tag is called a hypertext link. When you place it and its closing tag around text, a browser
makes it a link to the specified file. By default the link displays as blue, underlined text. HTML also
allows you to place images between these tags. Click the text or image between the tags and the
browser changes the document to the one specified. In HTML/OS, the link can be the name of an
HTML document, a URL to a page somewhere else on the Web, or a link to an on-click Overlay at
the bottom of the same Web page.
Example: <a href="/work/cart.html">View Shopping Cart</a>
Further Reading
Visit sites like Yahoo! (http://www.yahoo.com/) to see how other sites write their HTML pages. Also,
you may wish to look at the following books on the subject.
Chuck Musciano and Bill Kennedy. HTML and XHTML: The Definitive Guide, Second Edition (O'Reilly &
Associates, 2000) ISBN 059600026X.
Steve James and Ed Tittel. HTML for Dummies, Third Edition (Hungry Minds, Inc., 1997) ISBN
076450214X.
-277-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Appendix C: The Next Generation: Web-Based
Products
Overview
This appendix digresses from the book's main purpose, which is the development of your Webconstruction skills. Here you learn how to build commercially salable products that run across the
Web.
The ability to build a Web-based product and sell it to any business on the Web represents a huge
opportunity for developers. But the systems-integration approach to Web development, as discussed
in the "Legacy Beast" section of Chapter 1, Introduction, is once again a problem. Whereas systems
integration creates development complexities, upgrade problems, and reliability nightmares, that
pales in comparison to the complexities encountered by Web developers interested in creating
products that run not only on their own equipment, but also across the wide spectrum of
environments on the Web. Most Web developers encounter so many installation and compatibility
problems that they find they cannot sell their products at all. They literally give up and opt instead to
rent their applications rather than sell them. This model of renting software is known as the
application service providers (ASP) model. Unfortunately, this way of providing software (no matter
how financially attractive it may appear) is not attractive to customers who dislike the pay-forever
model of computing. Given a choice to buy or rent, many would rather buy. There is historical
precedent for this as mentioned in the accompanying "Is History Repeating Itself?" note.
Is History Repeating Itself?—Prior to 1985 most corporate software was rented and hosted on
large time-sharing computers. Why did companies rent software? Because, like today,
development was too costly and packaged solutions were not yet available. In 1985 the industry
suddenly collapsed when these billion dollar companies could no longer compete with the new
generation of young software developers selling vertical solutions for just about every industry
segment. Will application service providers suffer the same fate as these timer-sharing companies?
Only history can answer this question.
In this appendix you learn how to build Web-based products people can buy and install on their own
equipment or their own hosting accounts. First you look at the development precautions you need to
take. Then you are shown how to package a product using Bundle Bee, a wizard that walks you
through a dozen steps, from defining the icon for your application to defining which files should be
copy-protected. By the end of this appendix you'll know how to build Web-based products that can
be run anywhere on the Web.
-278-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Preparing Your Product
When building a Web-based software product, you follow the same steps you follow when you build
any advanced Web site except that you design your Web pages so they satisfy the following
requirements:
ƒ The Web pages must be installable into any directory.
ƒ The Web pages must be configurable by the user.
ƒ The Web pages must not contain server-specific components.
The first requirement, that the users must be able to install Web pages into any directory, is needed
since the HTML/OS installer gives users of HTML/OS the ability to set their own installation directory.
When users install your set of documents, they may place them in any directory, so you need to
design them to run from any directory. To do so, you simply need to specify directory, file, and
database names with relative paths. You should not use absolute references to filenames.
The second requirement, that the user must be able to configure the Web pages, is needed since
one product must serve many different customers. Parts of Web pages with customer-specific
information must be placed in text files or databases. You must give the user the ability to change
those settings using settings or configuration pages.
The last requirement, that Web pages must not contain server-specific components, is needed so
the product runs on any kind of platform. In general this is an easy requirement to satisfy since
HTML/OS tags, in general, do not contain server-specific parameters. Such parameters, like the
name of an outgoing mail server, are defined in the HTML/OS Control Panel, not in your application.
However, when it comes to file naming, it is important to realize that Unix and Linux file systems are
case sensitive, whereas the Windows system is not. To ensure your application works the same
across these different file systems, it's a good idea to use only lower-case filenames.
Once you have completed a set of Web pages satisfying these requirements, you can bundle the
files into a single file that users can install with the Install option in the HTML/OS Control Panel. To
do this you use an application called Bundle Bee. The product produced by Bundle Bee is a single
binary file that can be transported across the Web and installed on any copy of HTML/OS,
regardless of the server configuration or server hardware.
To understand the process of building a Web-based software product let's follow the packaging of a
Web-based game called Aestiva Landmines. The game is shown in Figure C.1.
-279-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure C.1: Landmines is a Web-Based Software Product Packaged with Bundle Bee.
The files and directories for the game can be found in the /apps/bundlebee/sample directory.
Copy them to the /apps/mygame directory. A total of seven files are used in the game. They are as
follows:
ƒ mines.html
ƒ lib/functions.lib
ƒ lib/genmines.lib
ƒ gifs/app_icon_landmines.gif
ƒ gifs/landminelogo.gif
ƒ gifs/mine.gif
ƒ gifs/minehit.gif
Note how the files are organized. These pages use the following conventions. While these
conventions are optional, they are recommended since they help make your application tidy and
easy to manage:
ƒ Place images in their own directory. Name the directory gifs or images.
ƒ Place files containing functions in a lib directory. Give each a .lib extension.
ƒ Place databases in their own directory. Name the directory dbs or data.
ƒ Place product settings files in their own directory. Name it settings or lib/settings.
ƒ Place the entry document in the highest-level directory of the product. Clear the directory
of other files, so it contains only the entry document, other subdirectories, and perhaps
a ReadMe file.
ƒ Use lowercase file and directory names.
ƒ Put the word icon in the name of the image for your product's icon image.
Using Bundle Bee
After you set up and test your files, you're ready to bundle them into a product with the Bundle Bee
program as follows:
-280-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
1. Click the Bundle Bee icon on the HTML/OS desktop.
2. Click Bundle on the Bundle Bee menu bar. This will provide you the first screen of
Bundle Bee's Application Bundling Wizard. The wizard allows you to step through the
process of creating a product.
TE
AM
FL
Y
3. First you determine the base directory for the files of your product. All product files
must be in this directory. The Landmines game files are stored in the
/apps/mygame directory, so that is your base directory. Change to that directory.
You should now see a screen like that shown in Figure C.2.
Figure C.2: When Packaging a Product You First Select Your Base Directory.
4. Click Continue. Here you select the files in /apps/mygame that you want in your
product. Since you wish to include all the files click All. This will place all the files and
directories in /apps/mygame into a Bundle List—shown on the right side of the page.
You should now see a screen like that shown in Figure C.3. Click Continue.
-281Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure C.3: Your Next Step in Packaging a Product is Selecting the Files to Place in Your
Product.
5. Now you select the image for your HTML/OS desktop icon. If you already have one,
select that. If not, use the Select From System Icons link and click Continue.
6. You will be asked to enter the desktop title for the icon. Enter Land Mines and click
Continue.
7. The wizard asks you which HTML document to use as your entry document. Select
mines.html and click Continue.
8. Next you select files that should not be overwritten at installation time. This is where
you select settings and configuration files so your users can safely install the bundle
on top of a pre-existing version of your product without disrupting the user's product
settings. All files you select here will not be installed if they already exist. The
Landmines product has no settings or configuration files, so click Continue.
9. Now the Bundle Bee Wizard asks for a product code, which Bundle Bee uses when
copy-protecting products. If you decide to restrict access to your product with a
product registration key, you must enter the product code, along with a product
registration key, in the HTML/OS Control Panel when you install the product. If you
don't correctly enter the product code and registration key, your program can detect
that within the code. See the accompanying note "Adding Product Registration
Keys." For this Landmines product, enter the product code landmines and click
Continue.
10. Adding Product Registration Keys—Registration keys are multidigit numbers a
user must enter into the Register Product option in the Control Panel to run
products that check for these keys. The registration key can be generated on a
copy of HTML/OS registered to the domain name where bundling is done, but not
elsewhere. The ISGOODKEY tag is what you use in your code to determine whether
-282-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
a user has a valid key. If ISGOODKEY is not TRUE, for example, you can limit
specific features or display an unregistered message. As an example, consider the
following Underlay:
11.
<<
12.
IF ISGOODKEY != "TRUE"
13.
THEN GOTO "register.html"
14.
/IF
15.
>>
16. You can place this code at the top of the entrance document of the application. It
directs unregistered users to register.html. Note that product registration is only
useful if the Web pages in your product are copy-protected (stopping users from
being able to view or modify the source code of your application). Copy protection is
done in Step 11 of the procedure in the main text of this section.
17. Next you set the default install directory for your application. Note that this is the
suggested install directory. It is not necessarily where your application will run, since
the user can change this at installation time. That is why it is important to design your
application so it can run from any directory. Also note that you should not necessarily
set this directory to the directory where the source files are located. Enter
/apps/landmines and click Continue. The users installing this application will see
this as their default installation directory. By convention, products are installed in a
subdirectory of the /apps directory. The word apps is short for applications.
18. Next you select the extensions of files that should be copy-protected. By default,
Bundle Bee copy protects files with .html, .htm, and .lib extensions. Copy
protection is a method of encrypting documents so they cannot be viewed. In
HTML/OS this is known as scrambling. The HTML/OS Professional engine can run
either source code or scrambled documents. The HTML/OS Runtime engine, as
discussed in the section "The Runtime Engine," later in this appendix, can run only
scrambled pages. Select the .html, .htm, and .lib extensions and click
Continue.
19. Next you select the HTML/OS desktop menu for your product's icon. You may want
your application icon to be placed in its own menu. Or, if you leave the setting blank,
your product icon will be installed in the Main Menu. Note that products can contain
other bundled products. If this occurs, the HTML/OS installer, at installation time, will
install all the icons in the menu you select here. Click Continue since you do not
need to build a custom menu.
20. You'll see the last setting page. Here the wizard asks you to enter a filename for your
product file. By convention, product files are given the .bb file extension. Enter a
filename and click Continue. You'll be placed on a summary screen that shows all
the settings you've entered for your product. To go back and make changes to the
specifications of your bundle, click the edit link next to the setting you wish to modify.
This summary screen is shown in Figure C.4.
-283-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure C.4: Before Final Bundling You Can Review Your Settings and Make Necessary
Changes.
21. To complete the bundling process, click Bundle Files. Bundle Bee will ask you
whether you wish to save the settings for the future and then bundle your files into a
single file or bundle. This usually takes a few seconds. When done, you have a
product that can be installed onto any copy of HTML/OS.
Further information on building registered copy-protected software is available in the on-line help
pop-up box in the Bundle Bee application. There you'll find complete descriptions of the previous
steps.
Installing a Product—To install a Bundle Bee file use the Install option in the HTML/OS Control
Panel. Click Control Panel from the menu bar on your HTML/OS desktop. Select Install. Then click
Install Product. You can then upload a Bundle Bee file or browse the HTML/OS file system for one.
After selecting a Bundle Bee file, click Install. When installation is complete you'll get a successful
installation message. You can now return to the HTML/OS desktop to view the icon for your new
product and run it.
The Runtime Engine
The HTML/OS engine comes in two versions: the Professional and the Runtime editions. The
Professional edition is discussed throughout this book. The Runtime edition is a version of the
engine designed specifically for running Web-based software products. It can run copy-protected
applications that have been scrambled with Bundle Bee. The Runtime engine cannot run HTML
documents with Overlays in them.
The Runtime edition of HTML/OS is generally what you want to distribute with your product. It is less
costly than the Professional edition and runs only scrambled applications. It includes the standard
HTML/OS Control Panel for configuring e-mail and security settings, and dbConsole, which is
convenient, because it gives customers open access to the databases within your product. The
ability to import and export data from databases in your product makes it more open, accessible, and
easier to interface with other Web-based products.
-284-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Appendix D: HTML/OS Tag Reference Guide
Overview
This appendix provides each HTML Overlay tag used in this book. Each tag includes a general
description of the tag, its parameters, the output returned by the tag, and values placed in
TAGRESULTS, and error status variable, if any.
Each description includes a "See Also" entry that lists related tags that may or may not appear in this
appendix. Related tags that don't appear in this book are marked with an asterisk. If further options
are available other than those described in this appendix, then the "See Also" entry will refer to other
options for the tag. Such additional options are listed in the Official HTML/OS Tag Reference. This
official reference is included with purchase of HTML/OS Professional. The additional options and
related tags are also available in the knowledge base on the Aestiva Web site at
http://www.aestiva.com/support/.
The appendix begins by listing tag names by category. Thereafter, Overlay tag descriptions are listed
alphabetically.
Tag Names by Category
The following are those HTML/OS Overlay tags that are used in this book grouped by categories. For
a complete list of Overlay tags see the HTML/OS tag Reference provided with purchase of HTML/OS.
Learn these First
This section includes the HTML/OS tags you should know like the back of your hand. Other tags can
be referenced as needed.
APPEND.../APPEND
COPY.../COPY
IF-THEN
LAYOUT
DISPLAY.../DISPLAY
FOR.../FOR
MAIL.../MAIL
PAGE
GOTO
ROW
Text Processing
Text processing tags work with sequences of text characters.
CUT
LOWER
CUTALL
MIDDLE
FORMAT
LEFT
REPLACE
REPLACEALL
LENGTH
RIGHT
LOCATE
TRIM
Table Operations
Table operations tags work with arrays—variables containing one or more rows and columns.
-285-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
APPEND.../APPEND
COLTOLIST
LAYOUT
LISTTOCOL
GETCOLBEGIN
MERGE
GETCOLEQ
ROW
GETCOLNOTEQ
SUMCOL
GETTABLE
Database
Database tags work with databases.
DBADD
DBDELETE
DBFINDSORT
DBGET
DBEDIT
DBGETREC
DBFIND
DBREMOVE
DBFINDJ
DBUNIQUE
File System
File system tags work with files and directories in the Web file system.
APPEND.../APPEND
COPY.../COPY
FILELIST
ISDIR
SYSMV
SYSRM
ISFILE
Programming
You use programming tags for conditionals and loops, and when performing tests, debugging, and
working with special characters.
COOKIEREAD
ISMOD10
COOKIEWRITE
ISNUMBER
COUNT
CR
LF
MAX
EXPAND.../EXPAND
FOR.../FOR
FUNCTION
IF-THEN
ISGOODKEY
ISINTEGER
RANDOM
ROUNDDOWN
ROUNDUP
TAB
TRACE
WHILE.../WHILE
Date and Time
Date and time tags work with dates and time.
ADDDAYS
ISWEEKDAY
GETDATE
TIMEFROM
ISDATE
TODAY
-286-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Index
Note: Boldface numbers indicate illustrations.
A
access control, 16, 27–18
Allow list for, 227, 227
in back-end management systems, 334
file, 23
in login pages and, 123
in networking, 226–228
accessing HTML/OS, 4, 14
account applications, 1
ACTION, 37, 43–44, 139
Add to Cart Button, 244, 261–262, 265, 280, 282, 284
ADDDAYS, 357
Aestiva Shell, 32, 32
Aestiva Web site, 59, 309, 309
Allow list for access control in, 227, 227
Allow Public Access setting, 16
AND, in query strings, 179, 183, 186
APPEND, 242, 245, 286, 355, 356, 358
applets, 6, 53
application service providers (ASPs), 347
applications, 19
archiving files, 111–113
assigning value to variable, 64–65
assignment files, network node, 231–232
Authorize.net credit card collection centers, 312
auto data typing, 65
automatic crash recovery in HTML/OS, 86
automatic login page, 128–129
B
<b> tag, 342
Baby Heirlooms Web site, 279, 279
Back button, 25
Back button protection, 206
back-end management systems, 315–337
access control in, 334
benefits of Web-based offices and, 316–317
browser selection for, 319
CHOPRIGHT used in, 334
collaboration in, 315
customizing, 317
CUT used in, 334
database management and, 319,/321–322
DBFIND used in, 325, 331, 333
DBGETREC used in, 326
design considerations in, 317–319
distributed control through, 318
documentation in, 318
e-mail and, 315
-287-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
editing documents in, 327–328
editing product database in, 322–327
email broadcast using, 334–336
EXPAND used in, 333
FOR loop in, 333
form in, 332–333
functions of, 320
GETCOLNOTEQ used in, 331–332
HTML documents used in, 320, 320
HTML editors and, 318–319
LAYOUT used in, 334
on-click overlays in, 325
order state in, 320
ordering system in, 328–334, 334
Quit button in, 327
Quit to Products button in, 327
REPLACEALL used in, 336
Save button in, 327
Simple Back End System (SBS) for, 319–336, 320
task limits in, 318
tasks performed in, 315
text editor for, 97
underlay in, 332
Web development and, 317–319
Web site maintenance in, 315
workflow system in, 328–334, 330
background color, 344
background themes, 17
banners, in shopping cart systems, 290
begins with (~) operator, in query strings, 179, 181
<body bgcolor> tag, 344
bold face, 342
bookmarks, 19
Boolean expressions, in query strings, 45, 179
<br> tag, 343
broadcast e-mail, in back-end management systems, 334–336
browsers
HTML and, 53
in back-end management systems, 319
overlay (Otag) processing by, 37–39
supported by HTML/OS, 8–9
Web-based offices and, 316–317
building a Web page, 26–30
Bundle Bee, 29, 30, 349, 350–354, 351–353
buttons, 6, 116–117
C
C languages, 7, 59, 62
calculations, 6
overlays containing, 55
variables in, 69
case insensitivity of HTML/OS, 54, 66
case sensitivity of Linux/Unix, 348
category lists, in product navigation pages, 259, 261, 266–272
cells, database, 47
-288-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
<center> tag, 343
centered/justified text, 343
checkbox, in forms, 139, 143
Checkout button, in shopping cart systems, 249
checkout page, 249–252, 257, 305–314
Complete Order button in, 311
Continue button in, 306, 308, 311
credit card collection in, 312
final review page for, 309–311
order confirmation in, 313
order processing in, 311–313
payment information page in, 307–308
posting transactions to database from, 313
Previous button in, 306, 308
shipping information page in, 305–307
splitting up transactions in, 305
Thank You notice in, 311–313
validation of data in, 306, 308
CHOPRIGHT, 334
clearing a variable, 69
clickable tree, in product navigation pages, 270–271
collaboration, 1
in back-end management systems, 315
in text editor program, 114–120
collections display, in product detail pages, 279–280
color code for icons, 19, 98
color text, 35–36, 343
COLTOLIST, 172, 356, 358
columns, 155
sortable, in reports, 197–199
comma-delimited files, 46, 102, 153
in Web database, 168–169, 168
command-line interface (Aestiva Shell), 32, 32
commenting out code, 81, 85–86
comments, 81, 85–86
Compare button, in product navigation pages, 265
comparison operators, 67
compatibility of HTML/OS, 8–9
Complete Order button, 252, 311
compression utilities, 31, 31
conditionals, 67
confirmation page, in e-commerce system, 252
contact management systems, 1
contained in (~ ~) operator, in query strings, 181
content, 6
Continue button, 306, 308, 311
Control Panel, 10, 15, 17, 28–30
accessing pages through, 27–28
Database in, 28, 30
Email in, 28, 29
Install option in, 28, 29, 354
network access level setting using, 229
Network in, 28, 30
network node setup using, 222–224
Security in, 28, 29–30
setting up network access using, 224, 225
-289-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
System in, 28, 29
COOKIEID, 296–298
COOKIEREAD, 102–104, 357, 358
cookies, 102–104
in login pages, 128–129
in shopping cart systems and, 296–298
tracking, 296–298
COOKIEWRITE, 102–104, 128–129, 357, 358–359
COPY, 46, 73, 75, 96, 97, 102, 112, 126, 153, 355, 356, 359
Copy button, in text editor program, 107–108
copy files, 21
copy protection, in Web-based product development, 353–354
CORBA, 3
COUNT, 146, 357, 359
Courtesy CareFree Garden Web site, 289, 289
CR, 357, 359
crash recovery in HTML/OS, 86
credit card collection, 312
cross-referencing, in product detail pages, 277
currency data, 180
in Web database, 163
CUT, 272, 334, 356, 359–360
CUTALL, 356, 360
D
DAT, 163
data types, 65–68, 163–164, 180
in Web database, 166
data validation, in forms, 144–156
database-driven E-commerce Web site, 242–252
database editor, 171–174, 203–217, 204
accessing or logging in to, 203–204
Back button protection in, 206
Copy buttons in, 204–205, 207, 208
DBADD used in, 206
DBDELETE used in, 206, 215–216
DBEDIT used in, 206, 215
DBFIND used in, 206, 215
DBFINDSORT in, 215–216, 215
DBREMOVE in, 216
Delete buttons in, 204–205, 207
Fifteen-line Database Editor example of, 204–216, 204
file upload option in, 211–212
Find buttons in, 204–205
GOTO used in, 206
group-based security in, 206
hidden fields in, 206
HTML form components in, 210
IF-THEN used in, 206, 208
login IDs for, 204
multiple database access using, primary and secondary databases, 212–216, 213
ownership of records and, 206
Save buttons in, 204–205, 207, 208
Six-line Find Page example for, 209–210
text area in, 210
-290-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
TE
AM
FL
Y
user feedback for, 207–209
validating data in, 208–209
database engine of HTML/OS, 4
Database Report example, 191–200
Database Service Providers (DSPs), 228–229
databases and database
management, 20, 30, 41–50, 41
accessing, 42
ACTION tags in, 43–44
in back-end management systems, 319, 321–322
cells in, 47
creating, using dbConsole, 42–43
dbConsole and, 30
DISPLAY data to Web page from, 46, 47
in e-commerce system, 253–254
editing, 322–327
editor for (See database editor), 203–217
error handling in, 44
fields in, 42–43
formatting/organizing, using LAYOUT, 41, 44, 46–48
IF-THEN in, 44, 45
joins in, 42, 186–189
length of fields in, 42
in login pages, 126–131
lost passwords at login, 131–135
networking in (See networking), 219–237
opening, 43
overlays (Otags) for, 43–44, 45
query pages in, 45, 177–190
records in, 42
reformatting data from, 48–49
reports from, 191–202
schemas in, 42
searching, using DBFIND, 41, 44, 45–46, 130
server processing of, 42
string data in, 43
tables in, 42, 48–49, 49
tags used in, 356
two-dimensional data in, 41
type of fields in, 43
variables in, 45, 46, 47–48
Web (See Web databases), 161–175
Web page for, 43
date and time data type, 180
tags for, 357
as variables, 68
in Web database, 164
DB tags, return status of, to isolate network failures, 234
DBADD, 42, 171–174, 206, 232, 233, 356, 360
dbConsole, 20, 30, 164–170
add fields to tables using, 166
change fields in database using, 167–168, 168
copy table using, 170
Create Database screen in, 167, 167
create table using, 166–167
creating database in, 42–43
-291Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
data types in, 166
database setup using, 164
deleting tables using, 170
Edit database using, 164–165
export records from database using, 169, 170
group fields of database using, 168, 181–183, 182
import or load records to database using, 168–169, 169
index or cross-reference database records using, 166, 167
information about database using, 170, 170
naming fields in, 166
opening database in, 43
"open" vs. "selected" in, 165
Purge records from database using, 165
recovering purged records in database using, 165
selection screen in, 164, 165
string data in, 167
DBDELETE, 171–174, 206, 215, 216, 356, 360
DBEDIT, 171–174, 42, 206, 215, 232, 233, 356, 360–361
DBFIND, 7, 41, 42, 44–46, 130, 178, 182, 186, 187, 191, 193–195, 198, 206, 215, 244, 263, 264, 267,
298, 325, 331, 333, 356, 361
parameters of, 45–46
DBFINDJ, 187–189, 361–362
DBFINDSORT, 198, 215–216, 263, 356, 362
DBGET, 356, 363
DBGETREC, 171–174, 245, 326, 356, 363
DBGETX, 165
DBREMOVE, 216, 356, 363
DBSEARCHX, 165
DBUNIQUE, 268, 272, 364
DCOM, 3
debugging (See also Tracer), 79–94
automatic crash recovery and, 86
bracketed messages in, 82
breaking code into small portions for, 83
commenting out code in, 81, 85–86
displaying intermediate values for, 85
ERROR in, 82–83
error reporting in, 80–83
full-screen reports in, 80–81, 81
Hello World example for, 90–91, 92
in page reports in, 82, 82
locating bugs in, 84–86
parse errors in, 80
tag specific results in, 82–83
TAGRESULTS in, 83
techniques for, 83–86
Tracer, 32, 81, 86–92
finding last point of execution using, 89–91
reserved word operations in, 89
trace console in, 87–88, 88
trace log in, 87
TRACE tag in, 88–89
variable changes and, 91–92
verify and repeat in, 84
viewing source in, 84
DEBUNIQUE, 356
-292-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
default variable values, 62–63
Delete button, in text editor program, 107–108
delimited text files, 73–74
delimiters, 73–74
DELROW, 75
desktop, 4
desktop applications, 20–26
Desktop Only access setting, 16
detail page to report links, 199–200
diamonds, 19
directories, 21–22
New Folder button, 26
Web File System (WFS), 23
in Web database, 163
DISPLAY, 46, 47, 55, 56, 66, 106, 146, 191, 242, 355, 364
distributed control, in back-end management systems, 318
distributed execution in HTML/OS, 10
distributed state persistence in HTML/OS, 10
document collaboration, in text editor program, 114–120
document in HTML, 53
document management systems, 1
documentation, in back-end management systems, 318
DocumentRoot, 23
DOL, 163, 167, 180
dot (.) as name character, 62
double-pass programs, 38
downloading files, 16, 21
drill-down, 186–187
dynamic Web sites, 53
E
e-commerce systems, 3, 239–258, 240
Add to Cart button in, 244
advanced site design for, 253–258, 253
APPEND used in, 242, 245
Checkout button in, 249
checkout page for, 249–252, 257, 305–314
Complete Order button in, 252
confirmation page for, 252
credit card collection in, 312
customization in, advanced, 255
database-driven designs for larger businesses, 242–252
database for, 253, 254
DBFIND in, 244
DBGETREC used in, 245
DISPLAY used in, 242
e-mail message page in, 242
error handling in, 242, 248
FOR loop in, 249
GOTO used in, 242
IF-THEN used in, 242
MAIL used in, 242
on-click overlays in, 239
order confirmation in, 313
Order database for, 254
-293-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
order processing in, 242, 311–313, 311
other pages for, 257–258
overlays in, 239
payment information page in, 307–308
posting transactions to database from, 313
product detail page in, 244–245, 256, 275–287
product information/navigation page in, 242, 243–244, 255, 259–274
Recalculate button in, 249, 302
review page, 309–311
settings data in, 253, 254
shipping formula calculator in, 246–249, 299–302
shipping information page in, 305–307
shopping cart system for, 244, 245–249, 256, 289–303
tax calculator in, 245–249, 299
Ten-line E-commerce Web Site example for, 239–242, 240
Thank You notice in, 311–313
User database for, 254
Web pages in, 253–254, 255
Edit Icon screen, 19, 19
editing files, 16, 22
editor (See Web-based text editor), 24
Eight-line Database Editor example, 171–174
ELSE, 67
e-mail, 6, 17, 28
in back-end management systems, broadcast using, 315, 334–336
in e-commerce system, 242
in login pages and, 134–135
lost password system using, 133
template to automate, 134–135
WebMail for, 33
end-of-line break, 343
ERROR, 44, 62, 71–72, 82–83, 125, 131, 145, 148
error handling, 30, 80–83
database, 44
in e-commerce system, 242, 248
ERROR as default variable in, 44, 62, 71–72, 82–83, 125, 131, 145, 148
in forms, 145, 148
in login pages, 124, 125
in networking, 226
in shopping cart systems, 248
error logs, in networking, 226
escape characters, 74
EXPAND, 333, 357, 364
explicit values for variables 72–73
export records from database, 169, 170
F
failure recovery, in networking, 233–236
FALSE, 66, 131
Famous UncleWebster Web site, 269, 269
field operators, in query strings, 180–181
fields, 42–43, 162–164
File Manager and general file
management, 1, 6, 15–17, 20–24, 20, 348
archiving and version control in, 111–113
-294-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
case sensitivity caveats in, 348
copy files using, 21
directory navigation using, 21–22
document collaboration in, 114–120
download files using, 21
editing files using, 22
editor in, Web-based text editor, 24–26, 25
exiting, 27
Find It using, 24
finding Start links using, 27, 28
Help using, 24
in networking, 220
in text editor program, 105–107, 106
in Web-based product development, 349, 350
information about files/directories using, 22, 23
internal file areas and, 23
last modified date of file using, 22
list of files using, sorting, 21, 113
move files using, 21, 118
multifile tasks in, 21
New File button for, 26
New Folder button for, 26
opening new file, using New, 26
path and pathname of file using, 22
saving work, using Save and Save As, 26–27
security settings for files using, 22
size of file using, 22
Start link access settings using, 22–23
tags used in, 356
type of document using, 22
viewing files using, 22
Web File System (WFS) and, 23
file selection, in text editor program, 105–107, 106
file transfer, 4
FILELIST, 113, 155, 356, 364–365
filenames, 348
final review page, in checkout pages, 309–311
financial reporting applications, 1
Find button
in product navigation pages, 261
in query pages and strings, 177, 178
Find It, 24
Find Page example, 209–210
firewalls, in networking, 221
Flash applications, 6
floating point data, 163, 180
FLT, 163, 167, 180
<font> tag, 343
fonts, 17, 343
color in, 35–36, 343
FOR loop, 49, 74, 75–77, 106, 191–193, 249, 333, 355, 357, 365
<form> tag, 138
FORMAT, 66, 356, 365
formatting/organizing data
FORMAT for, 66
using LAYOUT, 41, 44, 46–49
-295-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
in product detail pages, 281–282, 281
forms, forms processing, 137–157
ACTION parameters in, 139
attributes of components in, 140
in back-end management systems, 332–333
browser vs. server-side validation in, 145
checkbox in, 139, 143
components used in, 139–140, 139
content of components in, 140
data collection using, 138, 140
data validation in, 144–156
in database editor using, 210
destination of, 139
error handling in, 145, 148
Fifteen-Line Spreadsheet Editor example of, 151–153, 152
<form> tag for, 138
GOTO tags in, 139
hidden field, 139, 142
hyperlinks in, 139
IF-THEN statement in, 146
image area in, 139, 144
linking to Web pages from, using ACTION, 37
nesting unsupported in, 138
on-click overlays in, 56–57, 102, 146
password box in, 139, 142
in product detail pages, converting to HTML form components in, 282–286
radio buttons in, 139, 142–143
Select box in, 139, 141
Six-line Upload Page example of, 153–155
Submit button in, 138, 139, 144, 153
text box in, 139, 141–142
textarea in, 139, 140
Thirty-Line Guest Book example, 146–151
U.S. Tax Filing example of, 137, 137
upload button in, 139, 143, 154–155
uses for, 138
variables in, 140
fourth generation languages (4GL), HTML/OS as, 42
fractions, as variable, 66
freeware, 33
FTP, 4
full-screen error reports, 80–81, 81
FUNCTION, 357, 366
function libraries, 59
G
Get IP Address tool, 223
GETCOLBEGIN, 113, 356, 366
GETCOLEQ, 126, 155, 356, 366
GETCOLNOTEQ, 280, 296, 298, 331–332, 356, 366–367
GETDATE, 68, 357, 367
GETTABLE, 356, 367
GIF, 23
global variables, 68–69
globes, 19
-296-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Go button, 36
GOTO, 56, 96, 129, 139, 146, 206, 242, 355, 367
group-based security, in database editor, 206
group fields, in query strings, 181–183, 182
Guest Book example, 146–151
H
handles, database, 30
headers, in reports, 197
Hello World first program, 35–40, 35
color text in, 35–36
debugging, 90–91, 92
Go button in, 36
HTML document as template for, 39
overlay (Otags) tags in, 36
radio buttons in, 36
redisplaying page after SUBMIT in, 37
Help, 17, 24
hidden fields, 139, 142, 206
History button, 25, 29
hosting services, 8, 347
HTML, 20, 36, 53
browsers and, 53
computing and
programming-based on, 59
display HTML with overlays, 57–58
forms processing (See forms and forms processing), 137–157
HTML/OS tags in, 3
optional use of overlays in, 58–59
optional use of tags in, 58–59
overlays rules for, 57–59
overlays contained within HTML tags in, 58
page attribute tags in, 343–344
tag reference for, 341–344
tags in, 36
as template for HTML/OS, 39
text tags in, 342–343
U.S. Tax Filing form example of, 137, 137
underlay status of, 58–59
variables in, 71–73
writing tags in, 342
HTML editors, back-end management systems and, 318–319
<HTML> tag, 343–344
HTTPS, 30
detail pages to reports using, 199–200
in forms, 139
I
icons, 4
color code for, 19, 98
creating, 28
Edit Icon screen for, 19, 19
editing preferences in, 19
marker for, 19
menu of, 18
-297-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
New menu for, 15–16, 16
positioning, 16
row and column positioning of, 17
IF-THEN, 44, 45, 66, 67, 71–72, 97, 98, 108, 117, 125, 131, 146, 153, 173, 192, 206, 208, 242, 296, 355,
357, 367–368
image area, in forms, 139, 144
image selection page, in product navigation pages, 262–263
image zoom feature, in product detail pages, 275–276
images, 6
importing records to database, 168–169, 169
in-page error reports, 82, 82
index, in Web database, 162–163, 166, 167
information about files/directories, 22, 23
inlays, 55–56
Install option, 28, 354
installing applications, 17, 29
installing HTML/OS, 4
integer data (INT), 66, 163, 167, 180
integrated HTML/OS environment, 71
intelligent searches, 273
internal file areas, 23
intranets, 1
inventory management, in product detail pages, in real time, 278
IP addresses, networking and, 221–224
ISDATE, 357, 368
ISDIR, 356, 368
ISFILE, 66, 356, 368
ISGOODKEY, 357, 368
ISINTEGER, 357, 368
ISMOD10, 357, 369
ISNUMBER, 357, 369
ISWEEKDAY, 357, 369
JK
Java, 6, 53, 62
JavaScript, 6, 17, 53, 62
joins, 42, 186–189
JPEG, 23
justified text, 343
King Arthur Flour Web site, 260, 260
L
last-modified date of file, 22
launch windows, 17
LAYOUT, 7, 41, 44, 46–49, 191, 193, 262, 284, 334, 355, 356, 369
LEFT, 356, 370
legacy systems, 2–3, 161–162, 347
LENGTH, 146, 356, 370
LF, 357, 370
licensing, 8
line breaks, 133
linking, 6, 53, 56–57
HTML forms to Web pages, using ACTION, 37
Linux, 8, 9, 348
command-line interface (Aestiva Shell) for, 32, 32
-298-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
list of files, 21
LISTTOCOL, 270–271, 284, 285–286, 356, 370
literals, 70–71
local variables, 68–69
LOCATE, 356, 370
logging into HTML/OS, 13–14
accessing, 14
login screen for, 14
password for, 13
registration number for, 13
URL for, 13
usage limitations in, 14
logging, database, 30
logicals as variables, 66
login pages, 3, 123–135
access control and, 123
automatic login using, 128–129
cookies in, 128–129
database accessed for, 126–131
e-mail template for, 134–135
error reporting in, 124, 125
expiration date for password in, 130
functionality for, 124
loading user specific information in, 129–131
lost password handling and, 131–135
password page for, 124, 126, 129
security and, 123
Start links in, 123
Submit button in, 124
ten-line version of, 124–126, 125
in text editor program, 115–116
login screen for HTML/OS, 14
Logout, 15
loops, 74–77, 106, 153
lost passwords at login, 131–135
LOWER, 356, 371
M
MacOS, 9
MAIL, 242, 355, 371
Main Menu, 18
math convention, variables and, 64
math operators, 69
MAX, 357, 371
Menu options, 15, 97, 98
menus
Menu options for, 15
pull-down, 282–286, 283
MERGE, 286, 356, 371
MIDDLE, 356, 371–372
modular programming, 59
moving files, 21, 118
moving HTML/OS sites, 9
multidatabase searches (See also joins), 186
multifile tasks, 21
-299-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
multiselect product list, 264–266
My Apps, 18
My Websites, 18
N
NAME, 76, 77
names (See variables)
name-value pairs, variable, 71, 72
naming conventions
overlays, 58
variable, 61–62
nested HTML/OS tags, 155
nested parameters, in variables, 70–71
.NET, 3
Network, 28
network node assignment files, 231–232
networking using HTML/OS, 9–10, 17, 30, 219–237
access control for, 226–228
access level setting for, using Control Panel, 229
access to, 224, 225
accessing databases through, 230–236
Allow list for access control in, 227, 227
Control Panel setup for, 222–224
Database Service Providers (DSPs) for, 228–229
database-specific access in, 228
DBADD used in, 232, 233
DBEDIT used in, 232, 233
default access setting in, 227
distributed nature of, 220
ease of use in, 220
error logs for, 226
failure recovery in, 233–236
features of HTML/OS in, 220
file access in, 220
firewalls and, 221
Get IP Address tool for, 223
hardware requirements for, 220
IP addresses and, 221–224
naming nodes in, 223
network node assignment files for, 231–232
nodes in, 219–224
peering architecture in, 221
return status of DB tags in, to isolate failures, 234
scalability of, 220
servers for, 228–229
setting up network for, 221–224
TCP/IP in, 219
testing setup for, 225–226
tolerance of, 220
URLs and, 221–224
Web network and, 219–226
writing to database using, 232–233
New File button, 26
New Folder button, 26
New menu, 15–16, 16
-300-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
nodes, 30, 219–224
O
TE
AM
FL
Y
object-oriented programming, 62
OLE, 3
on-click overlays, 56–57, 102
in back-end management systems, 325
in e-commerce system, 239
in forms, 146
in product detail pages, 285–286
on-the-fly Web database, 164
on-the-fly Web sites, 6, 53
operators, 69, 180–181
option bars in HTML/OS, 15
OR, in query strings, 179, 183, 186
order confirmation, in checkout pages, 313
Order database, in e-commerce system, 254
order processing page, in e-commerce system, 242, 311–313
order state, in back-end management systems, 320
ordering system, in back-end management systems, 328–334
overlay (Otags), 5–6 54–55
calculation within, 55
case insensitivity in, 54
database, 45
displaying HTML text with, 57–58
in e-commerce system, 239
in Hello World first program, 36
HTML rules for, 57–59
HTML tags containing, 58
inlays vs., 55–56
instructions within, 54
names of, 58
on-click type, 56–57, 102
optional use of, 58–59
overlays within, 58
parse errors for, 80
positioning of, within document, 55–57
processing of, 37–39
security and, 98
underlays vs., 55–56
in Web database read and write, 164
P
Packit 2.0, 31, 31
PAGE, 355, 372
page attribute HTML tags, 343–344
page breaks, 196, 343
Page up/Page Down, in reports, 193–195
parameters, variable, 70–71
parse errors, 80
password box, in forms, 139, 142
password pages, in login pages, 124, 126, 129
passwords, 98, 126, 129
-301Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
database for, 126–131
expiration date for, 130
when logging into HTML/OS, 13
lost, handling of, 131–135
Web page, 97, 98
path, 22, 163
payment information page, in checkout pages, 307–308
peering architecture, networking, 221
Perl, 5
portability of HTML/OS, 8–9
portals, 1
positioning overlays, 55–57
Preferences, 15, 17, 18
Previous button, 306, 308
printing reports, 195–197
private access, 23, 98
product capsules, in product navigation pages, 259, 260
product collections display, in product detail pages, 279–280
product detail page, in e-commerce system, 244–245, 256, 275–287
Add to Cart Button in, 280, 282, 284
APPEND used in, 286
cross-referencing in, 277
formatting data in, 281–282
GETCOLNOTEQ used in, 280
HTML form components in, 282–286
image zoom feature in, 275–276
inventory management using, in real time, 278
LAYOUT used in, 284
LISTTOCOL used in, 284, 285–286
MERGE used in, 286
on-click overlays in, 285–286
options vs. add-ons for, 281–286
product collections display in, 279–280
pull-down menu in, 282–286 283
related items listing in, 276–277
SKU lists in, 277
WHILE loop in, 280, 285
product lists, in product
navigation pages, 259, 261–266, 272
in product navigation pages, 261–266, 261
product navigation pages, 242, 243–244, 259–274
Add To Cart link in, 261–262, 265
category lists in, 259, 261, 266–272
clickable tree in, 270–271
Compare button in, 265
CUT used in, 272
DBFIND used in, 263, 264, 267
DBFINDSORT used in, 263
DBUNIQUE used in, 268, 272
dynamic categories in, 268
in e-commerce system, 255
editable categories in, 267–268
Find button in, 261
hierarchical categories in, 269–272
image selection page for, 262–263
LAYOUT used in, 262
-302-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
LISTTOCOL used in, 270–271
multiselect product list in, 264–266
navigation feature selection for, 259–261
product capsules in, 259, 260
product lists in, 259, 261–266, 272
product specials listing in, 263–264
REPLACE used in, 272
REPLACEALL used in, 272
search boxes in, 260, 261, 272–273
subcategories for category list in, 271–272
use of, caveats for, 259
WHILE loop in, 262
product sharing, 8
Product Specific Promotion example, in shopping cart systems, 293–296
Professional edition of HTML/OS, 354
programming related tags, 357
promotional messages, in shopping cart systems, 290–296
public access, 23, 27, 98
pull-down menu, in product detail pages, 282–286, 283
Puritan's Pride Web site, 253, 253
Q
query pages and strings, 45, 177–190
AND and OR used in, 179, 183, 186
begins with (~) operator for, 179, 181
Boolean expressions used in, 179
check all that apply requests in, 184
complex, 184–186
contained in (~ ~) operator for, 181
data types searched for, 180–181
DBFIND in, 178, 182, 186, 187
DBFINDJ in, 187–189
drill-down and, 186–187
field operators in, 180–181
Find button and, 177, 178
Four-line Query Page example of, 178
free text requests in, 184
group fields in, 181–183, 182
joins in, on-the-fly, 186–189
LAYOUT for, 177
multidatabase searches (See also joins in), 186
multiple inputs in, 184–186
query pair requests in, 184
query string in, 178, 179–183
quotation marks in, 179–181
search boxes and, 272–273
selection requests in, 184
quotation marks
in literals, 70
in query strings, 179–181
in variables, 73
quoting characters, 73
R
radio buttons, 36, 107–108, 139, 142–143
-303-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
RANDOM, 357, 372
read database, 6
Recalculate button, in shopping cart systems, 249, 302
records, database, 42
registering programs, 29
registration keys, in Web-based product development, 352
registration number, to log in to HTML/OS, 13
related-items listing, in product detail pages, 276–277
Reload button, 25
REPLACE, 112, 118, 272, 356, 372
REPLACEALL, 70, 134, 146, 272, 336, 356, 372
reports, 191–202
database search results, 48–49
DBFIND used in, 193–195, 198
DBFINDSORT used in, 198
DISPLAY used in, 191
FOR Loop used in, 191, 192, 193
header in, 197
LAYOUT used in, 191, 193
linking to detail pages in, 199–200
page breaks in, 196
Page up/Page Down in, 193–195
printing, 195–197
ROUNDUP tag in, 197
Seven-line Database Report example for, 191–200
sortable columns in, 197–199
tables in, 48–49, 49
TAGRESULTS in, 194, 195
resources for HTML/OS, 339–340
return status of DB tags, to isolate network failures, 234
review page, in checkout pages, 309–311
RIGHT, 356, 373
ROUNDDOWN, 357, 373
ROUNDUP, 197, 357, 373
ROW, 355, 356, 373
rows and columns, 17
running a page, from Web-based text editor, 25–26
Runtime edition of HTML/OS, 354
S
sales systems, 1
Save/Save As, 26–27
Save/View button, 25
schemas, database, 42, 162
Schweitzer-linen Web site, 263, 263
scrambling, 353–354
search boxes, in product navigation pages, 260, 261, 272–273
search database, 6, 7
searches, intelligent, 273
searching (See also DBFIND), database, 44, 45
security, 28, 29–30, 98
access control and, 27–28
access control settings, 16
file, 22
group-based, 206
-304-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
in login pages and, 123
Select box, in forms, 139, 141
server jumping, 69
servers, 4, 59
directory structure of, 23
nodes and, 220–224
overlay (Otag) processing by, 37–39
supported by HTML/OS, 9
Web File System (WFS) and, 23
in Web database, 161–162
seven-line Database Report example, 191–200
shell, command-line interface (Aestiva Shell) for, 32, 32
shipping formula calculator, in shopping cart systems, 246–249, 299–302
shipping information page, in checkout pages, 305–307
Shipping Promotion example, in shopping cart systems, 290–293
shopping cart systems, 4–5, 69, 244, 245–249, 256, 289–303
Add to Cart Button in, 244, 261–262, 265, 280, 282, 284
advanced shipping options for, 299–302
APPEND used in, 245
banners in, 290
cart triggered promotions for, 290
category triggered promotions for, 290
Checkout button in, 249
checkout page for, 249–252
cookie tracking in, 296–298
DBFIND used in, 298
DBGETREC used in, 245
error handling in, 248
FOR loop in, 249
GETCOLNOTEQ used in, 296, 298
IF-THEN used in, 296
product specific promotions for, 290, 293–296
promotional messages in, 290–296
Recalculate button in, 249, 302
shipping formula calculator in, 246–249, 299–302
shipping promotion for, in twenty four-lines, 290–293
tax calculator in, 245–249, 299
shortcuts, 17
Simple Back End System (SBS) (See also back-end management systems), 319–336, 320
simplicity of HTML/OS, 7–10
SIN, 69
single pass programs, 38
size of file, 22
SKU lists, in product detail pages, 277
SOAP, 3
sockets, 7
Solaris, 10
sorting
columns in reports, 197–199
DBFINDSORT in, 198, 215–216
file lists, 21
speed of HTML/OS, 7
spell check, in text editor program, 108–110, 110
Spell Kit, 108–110, 110
Spreadsheet Editor example, 151–153, 152
spreadsheet programming and overlay tag analogy, 6
-305-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
SQL, 3
SQRT, 69
Start links, 19, 24
access setting for, 22–23
assigning, 27, 28
in login pages, 123
state persistence of variables, 69
static Web sites, 6, 53
STR, 163, 166, 180
string data, 43, 163, 167, 180
Stuffit, 31
Submit button, 56, 96, 138
in forms, 139, 144, 153
in login pages, 124
redisplay page after, 37
SUMCOL, 69, 70, 356, 373
SYSLS, 76
SYSMV, 118, 356, 374
SYSRM, 356, 374
System, 28
system information, 29
system-wide settings, 17
systems integration vs. HTML/OS, 2–3, 7, 8
T
TAB, 357, 374
table operations tags, 356
tables
add fields to, 166
copying, 170
database, 42, 48–49, 49, 48
deleting, 170
in Web database, 162–163, 166–167
tag-specific error results, 82–83
tagging HTML/OS in documents, 5–6
TAGRESULTS, 83, 194, 195
tags for HTML/OS, 355–376 nesting, 155
tags, HTML, 36, 341–344
Tar, 31
task limits, in back-end management systems, 318
tax calculator, in shopping cart systems, 245–249, 299
Tax Filing form example, 137, 137
TCP/IP, 219
Telnet, 4
template, e-mail automation, 134–135
text, as variable, 65
textarea, 139, 140
in database editor, 210
user-controlled resizing for, 99–105, 100
text box, 139, 141–142
resizing, 24–25, 24
text editors, 24–26, 25, 95–121
advanced options for, 99–120
Back button in, 25
in back-end management systems and, 327–328
-306-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Copy button in, 107–108
Delete button in, 107–108
document collaboration in, 114–120
file selection in, 105–107, 106
History button in, 25
login pages in, 115–116
in product navigation pages, editable categories list in, 267–268
quitting, 26
radio buttons in, 107–108
Reload button in, 25
running a page from, 25–26
Save/View button in, 25
six-line, 95–99, 95
spell check in, 108–110, 110
user-controlled text area resizing in, 99–105, 100
version control in, 111–113
workflow buttons in, 116–117
text processing tags, 356
text tags in HTML, 342–343
Thank You notice, in checkout pages, 311–313
TIMEFROM, 112, 357, 374
<title> tag, 344
TODAY, 357, 374
tolerance, in networking, 220
TRACE, 88–89, 357, 374–375
trace log, 87
Tracer, 32, 81, 86–92
finding last point of execution using, 89–91
reserved word operations in, 89
trace console in, 87–88, 88
trace log in, 87
TRACE tag in, 88–89
variable changes and, 91–92
tracking cookies, 296–298
transaction processing, 7–8
TRIM, 146, 356, 375
TRUE, 66
TS parameter, 74
two-dimensional data, 41, 63
type of document, 22
U
underlay, 55–56, 332
HTML as, 58–59
uninstall, 29
Unix, 8, 9, 348
command-line interface (Aestiva Shell) for, 32, 32
upload button
in database editor, 211–212
in forms, 139, 143, 154–155
Upload Page example, 153–155
uploading files, 16
URL
when logging into HTML/OS, 13
in networking and, 221–224
-307-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Start links as, 24
usage limitations, when logging into HTML/OS, 14
user-controlled text area resizing, 99–105, 100
User database, in e-commerce system, 254
user feedback, in database editor, 207–209
users, login pages, user specific information in, 129–131
Utilities, 18
V
validating data
in checkout pages, 306, 308
in database editor, 208–209
in forms data, 144–156
variables, 38, 45, 46–48
assignment of value to, 64–65
calculations and, 69
case insensitivity in, 66
clearing, using ERROR, 69
column and row location of, 64
comparison operators and, 67
conditionals and, 67
data types in, 65–68
dates and times as, 68
debugging, 91–92
default values of, 62–63
delimited text files and, 73–74
ERROR as default value for, 62–63
explicit values for, 72–73
in forms, 140
fractions as, 66
global, 68–69
HTML use of, 71–73
integers as, 66
limitations of, 66
literals and, 70–71
local, 68–69
logicals as, 66
loops and, 74–77
math convention in, 64
names, name-value pairs in, 71, 72
naming, 61–62
nesting parameters in, 70–71
operators for, 69
parameters for, 70–71
quotation marks and, for literals, 70
server jumping in, 69
size limitations of, 66
specifying, 64–77
state persistence of, 69
text type, 65
two-dimensional nature of, 45, 46, 63
version control, in text editor program, 111–113
viewing files, 22
-308-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
W
Web-based applications, 1, 59
Web-based offices, benefits of, 316–317
Web-based product development, 347–354
bundling programs for, 349, 350–354, 351–353
copy protection or scrambling in, 353–354
file organization in, 349, 350
preparing the product in, 348–350
registration keys for, 352
Web-based text editor (See text editor)
Web database (See also databases and database management), 161–175
add fields to tables, 166
architecture of, using HTML/OS, 162–164
change fields in, 167–168, 168
columns to lists in, using COLTOLIST, 172
comma or tab delimited files in, 168–169
copy table in, 170
Create Database screen for, 167, 167
create table in, 166–167
currency data in, 163
data types in, 163–164, 166
date and time data in, 164
DB overlay tags in, 171–174
dbConsole for, 164–170
deleting tables in, 170
Edit database, 164–165
Eight-line Database Editor example in, 171–174
export records from, 169, 170
fields in, 162–164
floating point data in, 163
group fields in, 168
import or load records to, 168–169, 169
index or cross-reference, 162–163, 166, 167
information about database in, 170, 170
integer data in, 163
legacy vs., 161–162
naming fields in, 166
on-the-fly construction of, 164
overlay tags in, to read and write to, 164
path and directory for, 163
Purge records from, 165
query pages in, 177–190
recovering purged records in, 165
schemas in, 162
selection screen for, 164, 165
servers and, 161–162
setup for, using dbConsole, 164
string data in, 163, 167
tables in, 162–163
Web development, 317–319
Web File System (WFS), 23
Web networking (See networking)
WebMail, 33
Webtext box resizing using, 24–25
WHILE DISPLAY, 153
WHILE loop, 74–75, 153, 262, 280, 285, 357, 375
-309-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Windows, 8, 9, 348
workflow buttons, 116–117
workflow system, 1, 114–120, 328–334, 330
write to database, 6
XYZ
Y2K convention, 68
Zip files, 31
zoom feature, in product detail pages, 275–276
-310-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
List of Figures
Chapter 2: Logging into Your Copy of HTML/OS
Figure 2.1: This is Your HTML/OS Login Screen.
Figure 2.2: The HTML/OS Desktop…
Figure 2.3: Creating a New Icon or Bookmark.
Figure 2.4: You Use the Preferences Window to Change Desktop Preferences.
Figure 2.5: In the Edit Icon Window You Can Delete or Move Icons.
Figure 2.6: From the File Manager, You Can Access Documents on Your Web Site, Edit Them,
And Run Them.
Figure 2.7: You Can View File Information From the File Manager.
Figure 2.8: Your Web-Editor is Used to Edit HTML Documents and Try Them Out.
Figure 2.9: Start-Links are Allowed using the Control Panel.
Figure 2.10: Packit Backs up Files and Directories or Transports Large Sets of Documents
between Copies of HTML/OS.
AM
FL
Y
Figure 2.11: Aestiva Shell is an Optional Command-Line Interface for those Who Want to Keep
Their Fingers on the Keyboard.
Chapter 3: Your First Program
Figure 3.1: Hello World is Your First Web-Based Computer Program.
Chapter 4: Your First Web Database Program
Figure 4.1: Creating a Web Database is Point and Click.
TE
Figure 4.2: Your Web Database Program with Search Results Will Look like This.
Figure 4.3: Your Web Database Program Looks like this with Nicely Formatted Search Results.
Chapter 6: Variables, Conditionals, and Loops
Figure 6.1: The Size of my_var Varies Depending on What You Place in It.
Figure 6.2: The View Source Option of Your Browser Allows You to View the HTML-Only
Document Rendered by HTML/OS.
Chapter 7: Debugging Techniques
Figure 7.1: Full-Screen Error Reports Occur When HTML/OS Cannot Render a Web Page.
Figure 7.2: HTML/OS Provides In-Page Error Reports When it Can Render the Web Pages Despite
the Error.
Figure 7.3: The Main Tracer Screen Includes a Trace View with Many Controls.
Figure 7.4: Here's a Trace of a Hello World Program Populated with TRACE Flags.
Figure 7.5: Here's a Trace of a Hello World Program with Variable Tracing.
Chapter 8: Building Text Editors
Figure 8.1: You Build the Six-Line Text Editor by using Six HTML/OS Instructions and Three HTML
Form Components.
Figure 8.2: The Six-Line Editor Now Has User-Controlled Text-Area Resizing.
-311Team-Fly®
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Figure 8.3: A File Selection Screen is a Useful Addition to the Six-Line Editor.
Figure 8.4: A Spell-Check Screen is Another Useful Addition to the Six-Line Editor.
Chapter 9: Building Login Pages
Figure 9.1: The Ten-Line Login Page Uses Ten HTML/OS Instructions and an HTML Form.
Chapter 10: HTML Forms Processing
Figure 10.1: This U.S. Tax Filing Service is an Application of HTML Forms Processing in an
HTML/OS-Driven Web Site.
Figure 10.2: These 10 HTML Form Components are all You Need to Build an Advanced Web Site.
Figure 10.3: This Spreadsheet Editor Gives Users the Ability to Edit Two-Dimensional Data.
Chapter 11: The Web Database
Figure 11.1: The Database Selection Screen Highlights Databases Making Them Easy to
Recognize.
Figure 11.2: You Can Expand the Database Create Screen to Accommodate Additional Fields by
Clicking the Add Fields Button.
Figure 11.3: The Change Database Screen Includes an Add Group Option for Adding Group Fields.
Figure 11.4: The Import Option in dbConsole Allows You Load Various Kinds of Delimited Files Into
HTML/OS Databases.
Figure 11.5: When Importing From a Delimited File, For Every Column in the File, You Specify the
Field in the Database it Should Fill.
Figure 11.6: The Info Screen in dbConsole Provides a Snapshot of Your Database.
Figure 11.7: This Eight-Line Database Editor is Great at Editing Small Databases.
Chapter 12: Building Query Pages
Figure 12.1: Use the Change Section in dbConsole to Add Group Fields to Databases.
Chapter 14: Building Database Editors
Figure 14.1: The Fifteen-Line Database Editor Works with User-Assigned Records.
Figure 14.2: The Project Database Edit Page Works with Two Databases.
Chapter 15: Database Networking
Figure 15.1: The HTML/OS Control Panel Network Configuration Screen Contains the Nodes You
Peer With.
Figure 15.2: To Give a Node Full Network Access to Your Node Set Their Node to Full-Access in
the Network Allow List Page of the HTML/OS Control Panel.
Figure 15.3: The Network Database Allow List Page of the HTML/OS Control Panel is Where You
Place Limitations on Other Nodes Accessing Your Databases.
Chapter 16: Designing E-Commerce Systems
Figure 16.1: The Ten-Line E-Commerce Web Site has all the Basic Capabilities of an E-Commerce
Site.
Figure 16.2: Sophisticated E-Commerce Sites Merge Navigation and Product Information Together
on the Same Page.
-312-
Advanced Web Sites Made Easy: The Simple Way to Build Databases and Other Applications with HTML/OS
Chapter 17: Building Product Navigation Pages
Figure 17.1: The Recipe Book on this Web Site Uses Multiple Navigation Tools to Ensure Recipes
are Easy to Find.
Figure 17.2: This Web Site Uses Images as a Navigation Tool.
Figure 17.3: The UncleWebster Web Site Uses a Hierarchical Tree as a Navigational Aid.
Chapter 18: Building Product Detail Pages
Figure 18.1: This Baby Store Uses Product Collections to Display Related Products on the Same
Web Page.
Figure 18.2: Product Detail Pages Can be Outfitted with Product Option Menus, as Shown Here.
Chapter 19: Building Shopping Cart Pages
Figure 19.1: This Garden Supply Store Displays Promotional Products on Its Shopping Cart Page.
Chapter 20: Building Checkout Pages
Figure 20.1: The Checkout Pages of this Software Company Include this Review Page so a User
Can Edit the Order before Finalizing It.
Chapter 21: Building Back-End Management Systems
Figure 21.1: The SBS System Automates the Tasks Needed to Maintain a Web Site.
Figure 21.2: The SBS System Uses Only the HTML Documents Shown in this Listing.
Figure 21.3: The Workflow System Used by SBS Includes a Page Where Staff Can Move Orders
among Different Order States.
Appendix C: The Next Generation: Web-Based Products
Figure C.1: Landmines is a Web-Based Software Product Packaged with Bundle Bee.
Figure C.2: When Packaging a Product You First Select Your Base Directory.
Figure C.3: Your Next Step in Packaging a Product is Selecting the Files to Place in Your Product.
Figure C.4: Before Final Bundling You Can Review Your Settings and Make Necessary Changes.
-313-
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement