Using Fluffy Search

Fluffy Search has been designed to be able to search most web sites with little or no modification to the existing pages. In fact, unless you use multiple framesets, you will be unlikely to need to do anything.

Simply install the scripts on your server and edit the configuration file to point it to the right files. Create a 'template' page which gives the header and footer for the search form and search results. Finally, run the indexer, and link the search form to your existing pages. And you now have a search engine!

Every time you modify the site, you need to re-index the site. This can be done from a web browser, using a password you specify in the configuration file.

Controlling what is indexed

You can control what is indexed on your site. If you include a file called .fcs_exclude in a directory, then all the files within that, and the directories below it, are excluded from the index, and won't be found in a search. (The .fcs_exclude should be a zero length file. Fluffy Search only looks for it's existence to exclude the directory, not the content.)

Within the configuration file, you can also specify rules on what files will be indexed, so for example, you could index .htm and .shtml files, but not .html files if you so wished.

You can even specify exceptions to the above rules, to, for example, exclude files with particular names. You might want to do this to exclude files called 'toc.html' if these are tables of contents used to display navigation in framesets.

Finally, you can enclose text in <fcs_ni></fcs_ni> tag pairs to exclude parts of a page from the index. This might be useful to exclude constant navigational elements from pages.

Following the UNIX convention for hiding files, directories with names beginning with '.' are ignored. To avoid indexing the extra data FrontPage extensions add into the document directories, directories with names beginning with '_' are also ignored.

Framesets

If your site uses multiple framesets, for example, one for each section, Fluffy Search can display found pages in the correct frameset when you click on a link on the search results.

It does this by finding the frameset for that section, and building a new frameset 'on the fly' to enclose the found page.

You need to tell Fluffy Search which pages go in which frameset, and the frame within that frameset which is supposed to contain the content. You can either do this by a general rule (for example, the frameset is always called 'index.html' in the directory above the page) or by a list of directories and the location of the frameset for each.

Alternatively, you can use a combination of the two methods. The general rule specifies the majority of the framesets, and the list gives the exceptions to the rule.

There is a slight restriction. All pages within a directory must use the same frameset. Reorganising your site to meet this requirement should be the only modification you need make to your site.

Password protected directories

A word of warning: The page highlighter script does not do any password protection. If you index a password protected directory on your server but do not apply the same password protection to the cgi-bin directory Fluffy Search is in, then you will be able to use the search to retrieve the protected files without a password.

Make sure the script directory has the same password protection as the pages you're searching!

If there are some password protected areas on your site, and you want to search them as well as the non-protected areas, you can either turn off highlighting, in which case all files will be retrieved from the web server as normal, or create two search scripts, one which searches the non-protected pages only and the second which searches everything, and is also password protected.

< Contents | Requirements >