WillMaster Possibillites Logo EzineSeek Award
Personal Search Engine
For Your Site

by
William Bontrager

Permission is granted to reprint the article below in its entirety provided no changes are made to the article text and the author's name, signature lines, and copyright line are printed with the article. However, you may change the article's title.

You've got yourself a nice site. And you want to make it nicer with a search engine.

This is the right article for you.

To clarify, "search engine" and "directory" are different concepts.

A search engine, by definition, searches. Type in a word or phrase and click the button. It does the searching for you.

With a directory, visitors do their own searching. You click on subjects until you get to the section of the directory which contains a link to the information you are looking for.

This article is about search engines.

The search engine familiar to most folks is the internet search engine. Internet search engines, as the name implies, can search all the public sites on the internet.

(Well, that's not exactly true. The size and growth rate of the internet, and the limitations imposed by physical computers and the hardware connecting the net, preclude any one search engine searching the entire internet.)

Internet search engines use http, ftp, or other protocol to fetch files from remote sites. Some well known (webcrawler.com, hotbot.com, etc.) and many relatively obscure sites make internet search engines available to netizens and our seemingly insatiable quest for information.

A local search engine, on the other hand, is limited to one site (or portion of a site). It is unable to search other sites on the internet. It just scans the files on the local server. When you use a search engine like this, your search results will pertain to one site.

No need to spend the big money for (and distract your visitors with) an "internet" search engine when all you want is help your visitors find information on your site. Large or diverse sites can keep visitors longer with a well-implemented local search engine.

Suppose you have a business opportunity site with its own ezine. A local search engine lets your visitors find otherwise obscure references to their areas of interest. An easy to use site containing information valuable to the visitor is likely to be remembered and bookmarked.

Local search engines search your site in one of two ways.

  1. Real-time: A real-time search is a search of your site as it currently exists. When you update a webpage or other searchable file, it is immediately available to the search engine.

    Anywhere But In the Kitchen, Mari's popular kitchen advice/tips and recipe request/exchange site, utilizes a local search engine to search not only the site's HTML pages but also the database of recipes. Thus, a person can get a list of recipes with specific ingredients. It is a real-time search engine. Whenever someone requests or submits a recipe, their contribution is immediately available for searching.

    Go to http://willmaster.com/kitchen/ and play with it for a bit. Get a sense of what a simple local search engine can do. Such at tool might be exactly what you're looking for.

    For local, real-time searches, I recommend the free "Simple Search." This one is so easy to use that I refer people to it rather than writing my own free engine. And it is quite modifiable for those proficient with Perl.

    You can download Matt Wright's "Simple Search" from Matt's Script Archive at http://worldwidemart.com/scripts/search.shtml

  2. Pre-compiled: The search engine, or its helper, scans your site every night (or other frequency which you determine). During that scan, a list of keywords and/or a word list of your pages are compiled and stored in a database.

    When your visitor uses the search engine, the engine searches the compiled database and renders the results page from those matches.

    The pre-compiled search method is best for large sites. Although the database is always outdated by the amount of time since the most recent compilation, it is the preferred solution for searches exceeding a search file size total of 10-15mg. (Beyond that point, a real-time search becomes unacceptably slow.)

As mentioned above, I recommend "Simple Search" for local searches, especially if the total size of your HTML files is less than 10mg. However, I have investigated only a few scripts of the many available at CGI download sites. Start at the "Links and Resources" section of http://willmaster.com/possibilities/examples.html and surf around a bit; see what is available which may meet your site requirements.

If you don't have CGI for your site, you may wish to give http://www.freefind.com/ a try. They advertise "less than 10 minutes" setup. And it's free.

If you need help let me know.

Copyright 1999 by William Bontrager

William Bontrager, Programmer and Publisher
"Screaming Hot CGI" programs
"WillMaster Possibilities" e-Newsletter
http://willmaster.com/possibilities/
subscribe-possibilities@willmaster.com