Web Crossing


Introduction

Installation & Upgrade

Web Crossing Features

Conferencing

Live Events

Email Services & Mailing List Mirroring

News Services

WAP Interface

Calendar

XML-RPC

Customizing & Scripting

User & Access Issues

Data Organization & Management

Performance Issues

Appendix

Site Map

Enabling Searches of Your Site

Searches
Enabling Indexed Searches
Search Rules and Interface
Editing the Default Stopwords List
Troubleshooting
Resources

Searches

In the Control Panel > Customizing > General Settings > Search, you can enable searches of the content on your site, and/or searches of the web via Forum One, which lists and categorizes community sites.

You can choose to put the search button on just the top-level page or on all submerges as well. Users can search:

  • Just the current folder or discussion
  • Your whole site
  • Forum One

Users are only be able to search areas to which they have access. Posts from areas where they don't have access will not show up in their search results.

To turn off searches altogether, uncheck the boxes to remove the search buttons from the top level and nested pages.

Enabling Indexed Searches

Web Crossing uses a mini-database called a search index to help make the searches go quickly, even if your database grows very large.

Indexed searching is on, by default, at installation. To turn it off, go to: Control Panel > Customizing > General Settings > Search.

The master switch for creating (or re-creating) a search index is at the bottom of the Control Panel at Site management > Create search index. Web Crossing will create (or re-create) the search index when you click this link. Search index creation takes place in the background, meaning Web Crossing will still serve pages while it's indexing. When the index is done, the server will reload the page and tell you the index is ready to use. You might want to periodically recreate your search index because it makes a smaller file as it gets rid of information about deleted posts.

If a user clicks Search before you've created a search index, Web Crossing will begin indexing, informing the user that indexing is proceeding and s/he is currently unable to search the site. Search index files can be quite large - 50% the size of your database file or more.

Search Rules and Interface

Web Crossing search uses a number of rules in creating the index and performing searches:

  1. The word must be greater than 2 characters in length to be indexed.
  2. The first two characters must be letters or the word won't be indexed.
  3. Words with all caps are indexed separately from mixed/lower case words.
  4. All words are converted to lower case except for words that are all uppercase.
  5. HTML tags aren't indexed.
  6. Words that are longer than 48 characters are truncated.
  7. The following suffixes are stripped: ers, ing, ied, ies, ed, es, 's, er, e, s, y, '.
  8. In order to save index space, a number of common words are not indexed. The complete list is in a file called stopwords.default in your webx directory. You can Edit this list if you wish.

To search for keywords or userIDs within your Web Crossing content, type in one or more words separated by spaces. Web Crossing will ignore upper or lower case and match all similar words with suffixes and plurals, such as "s" and "ed." Only items matching all the keywords will be returned.

Web Crossing will return the title, author and date/time of each post along with the first few words. The first 15 items will be displayed, and if there are more, clicking a button at the bottom will display the next 15.

If you put in a keyword you know is there and Web Crossing returns no matches, it might be too common a word or a word shorter than two characters. In order to save indexing space a large number of common words are not indexed. Try again, using a more specific or uncommon term, or a longer word.

Editing the Default Stopwords List

Web Crossing omits a number of common words in order to minimize the size of the search index file and increase search speed. The list is found in a text file called stopwords.default in your webx directory on the web server. You can edit this list if you wish. Here are the steps:

  1. Rename the stopwords.default file to stopwords.
  2. Edit it as you desire.
  3. Stop and restart Web Crossing and log in as sysop.
  4. Go to Control Panel > Site management > Create Search Index to rebuild the search index using your new edited file.

Troubleshooting

I'm searching for a word I know is in the database, but it's not showing up on search.

  • Try a more specific, less common word
  • Be sure your word is longer than two characters
  • Edit the stopwords file and re-index the site, and then try again

Resources

Sysop Documentation


Web Crossing FAQ


Sysop Control Panel

  • Customizing
    • General Settings
      • Search

  • Site Management
    • Create search index


A Non-Programmer's Guide to Web Crossing
by Sue Boettcher and Doug Lerner

© 2000 Web Crossing, Inc.