[Kbabel] new searchengine

[Andrea Rizzi]

Alle 10:41, martedì 3 giugno 2003, Stanislav Visnovsky ha scritto:
> Dňa Pondelok 02 Jún 2003 19:16 Andrea Rizzi napísal:
> > Hi all,
> > I finally finished my university studies so I have time for KDE and i'm
> > going to finish the new search engine  for kbabel.
> >
> > How can I fit the release schedule for kbabel, or better what is the
> > kbabel release schedule?
>
> Hmmm, there is no schedule ATM. I'd like to push out 1.2 after fixing the
> reported problems. But there is no timetable.
>
> IMHO we should not include the new search engine in 1.2, but postpone it
> for the 1.3 (still in KDE 3.2 timeframe).
Ok.

> > I do not plan to write an import for old database, is it a problem?
>
> Maybe yes - a lot of users have a database several MB big. The diff feature
> works very nicely then.

I know, ok I'll try to implement something but it may be a problem.  It is not 
easy and this need to write a reader for the old database.

BTW is is possible to download an old CVS, launch the "scan recur", update the 
CVS and launch the scan again.

> Question: do you plan a possibility to have multiple instances of the
> search engine? For example to use one for glossary, another one for all
> translations etc?
I do not understand here.
There is another way to interface to kbabel different from 
kbabeldict/SearchEngine ?

BTW, transactions protection is not yet implemented, this is not a must at the 
moment. I still have to implement Catalog Information database, CVS/pofile  
scan,fuzzy sentence archive, a setup GUI.
Bye
Andrea
P.S.
Can we change kbmailer.cpp so that filenames in archives are provided with 
path?  (I implemented it on my local copy of kbabel but it use absolute paths 
and I think it is a security problem, i.e. contains info about fs hieracy. It 
should be relative to catalogs-root-dir but I have no idea about how can I 
access it)