Thunderstone Search Appliance Manual

Word Definition

Syntax: one or more regular expressions (REX), each on a separate line

Sets the word matching expression(s). Each line is a regular expression defining what is considered a word within the textual content of the retrieved documents during the index process. The default expressions index normal words and some special items such as domain names.

You may supply multiple expressions, one per line, if you can't define your idea of all possible words in one expression.

For example, >>\alpha=\alnum{1,20} will index "words" beginning with an alphabetic character followed by 1 to 20 alphabetic or numeric characters.

If Word Definition is changed, the Language Characters setting (here) should generally be updated to reflect any new characters added.

Changing the word definition with Update instead of Update and GO will cause the existing search index on the data to be dropped and rebuilt. The database will not be searchable during the time that the index is being rebuilt; this may take several minutes or more for large profiles.