Ok, last post.

In config.h it says this for WORDCHARS

** Note that if you omit "0123456789&#;" you will not be able to
** index HTML entities. 

Why should WordCharacters have anything to do with HTML Entities?
Shouldn't HTML entities be converted *before* extracting words from the
source with WordCharacters, BeginChars, EndChars, IgnoreLast, IgnoreFirst?

Thanks,

Bill Moseley
mailto:moseley@hank.org