Hi,
For cedet beta2, I've got some pending modifications that formalize
the search results, and provide access to such things. At the
moment, however, you need to be clever about how you handle the
search results.
The results is a list like this:
( (table1 matcha matchb matchc)
(table2 matchA matchB matchC) ... )
If you want the file for matchA, you need to remember it was
associated with table2. Table2 contains the filename.
In semantic 1.4, it was common to pump the results through a strip
function to flatten the list, making it compatible with other table
handling routines.
Eric
>>> "Berndl, Klaus" <klaus.berndl@...> seems to think that:
>Hi Eric,
>
>Another question about semanticdb:
>
>If semanticdb-find-tags-by-name return found tags how can i
>get the full filename for a certain tag so i can open the
>associated file and load it into Emacs.
>
>I do not want to use the parameter find-file-match!
>
>Thanks Klaus
>
>P.S.
>I have scanned the shipped info-manual but have not found an
>answer to my question above....
>
>-----Original Message-----
>From: Eric M. Ludlam
>To: Berndl, Klaus
>Cc: david.ponce@...; cedet-devel@...
>Sent: 05.11.03 16:06
>Subject: Re[1]: Question about new semanticdb in cedet 1.0
>
>>>> "Berndl, Klaus" <klaus.berndl@...> seems to think that:
>>Hi David, Eric,
>>
>>some functions in semanticdb are marked as obsolete but there is not
>mentioned what to use instead.
>>
>>Example: ECB uses semanticdb-find-nonterminal-by-name. What is the way
>to get this functionality with semantic 2.0?
> [ ... ]
>
>Hi,
>
> Many of the emails you've sent will take me some time to account
>for, but this I can answer simply.
>
> The `semanticdb-find-nonterminal-*' functions all take 8 arguments
>or so, and this bothered me. The new functions remove several
>arguments that weren't used much, and in some cases there will be two
>functions to account for a Boolean argument. The new functions are
>`semanticdb-find-tags-*'
>
> Another important difference is that the original functions scanned
>all nearby files associated with a 'project'. Many of the new
>functions use the include statements to expand the search, and will
>seek out and find sources to search, not just the files you've
>already loaded. The old behavior is found with the routines that say
>"brute" in the name. I found these routines were restricted in use,
>so there are fewer of them than the other types of searches.
>
> There are still holes in the coverage of old routines to new
>routines, many of which are on purpose as I tried to guess which ones
>were never used. If I was wrong, I will put them back in.
>
>Eric
>
--
Eric Ludlam: zappo@..., eric@...
Home: http://www.ludlam.net Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net GNU: http://www.gnu.org
|