Re: Apropos commands and regexps

[Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann)]

"Eli Zaretskii" <eliz@is.elta.co.il> writes:

>> From: Kai.Grossjohann@CS.Uni-Dortmund.DE
>> Date: Sun, 19 May 2002 14:02:24 +0200
>> 
>> > C-h i m emacs i paste RET
>> > will make you jump to the glossary where it says "see killing and yanking",
>> > so the user should then know to search for `yank'.
>> 
>> This won't work for more complicated queries, only for the
>> single-word case.  What happens when the user searches for "cut
>> line"?  The fact that "cut" is mentioned in the glossary won't help
>> so much.
>
> The same problem will happen if the user searches for "yank line".
> So I'm not sure what is your point; it is quite clear that some of
> the searches will fail to find all the keywords, and will have to do
> something smart about that.

When the user enters a query `foo bar' and Emacs knows that quux is a
synonym for foo, then the system should search for `foo bar quux'.
(The preceding sentence assumes that the system interprets the query
in a best-match fashion.  It will have to be changed accordingly for
other query interpretations.)

The fact that the system already finds the glossary entry for quux
when the user enters foo doesn't provide the functionality that's
described in my first sentence in this message.

Does that make it clearer what my point is?

kai
-- 
Silence is foo!