How Client Search Works

In most web site search facilities the search processing is done at the Server. This is the computer where the web pages are hosted (e.g. ISP's computer). A Browser reads a HTML page and displays it on the Client computer (e.g. a home computer). The Client fills in a Form which is submitted to the Server for processing. After processing, the results are passed back to the Client for displaying in the Browser.

This mechanism assumes that the Server provides search processing software. In some cases this is not available or cannot be assumed. Examples where this software is not available are low cost ISPs where only very basic web page serving is available or CD-ROM distributions.

Client Search provides a facility where the search processing is performed on the Client computer. The software is provided in an Applet and in invoked by a HTML page in a standard Browser.

The Client Search Applet is specified in a HTML page. The Browser downloads it to the Client computer and runs it. Note: The Applet is only allowed by Java Security rules to read the files on the host web site so maintains the privacy of Client installations.

The user enters keywords into the Applet on the HTML page and initiates the search. The Applet will read HTML files from the host web site, scan them for keywords and determine search statistics. If the spider facility is enabled, the scan will also check for linked HTML files on the web site and scan those. If any of those files contain links to other files on the host site, the links will be further followed until all files have been searched. When the search is complete, a ranked report is given to the user stating the files with keyword hits together with hyperlinks.

Note: The search is performed only on HTML files for speed. Graphics files are not downloaded for scanning, making the search more efficient.