-----BEGIN PGP SIGNED MESSAGE-----
Ok, well the properties indexed and most of that are pretty standard,
they are using like an rdf style format as opposed to a dublin:core
thing like us. Conceptually, should be simple, I might start playing
with it tonight, but most of the raw lucene stuff is pretty alien to me,
so I don't expect much. (meaning if someone else would like to give it a
shot, feel free ;) )
Joe Shaw wrote:
> On Thu, 2006-07-06 at 13:27 -0400, Kevin Kubasik wrote:
>> As of the 0.7 or 'Cardnial' release series, flock uses Clucene 1.9. If
>> you wanna check it out, your index is in the folder 'historysearch' in
>> the flock profile directory.
> I am pretty sure the CLucene file format is on-disk compatible with the
> original Java one (and our DotLucene). As long as that's the case,
> writing a driver for it should be trivial.
>> Whats really the best thing about it is that because the indexing is all
>> in C and part of the browser, they have the scheduling etc. to make it
>> not interfere with the browsing experience at all.
> Sounds familiar. ;)
>> Is there some command line tool to just dump the contents of lucene
>> index into text form or something? I can't seem to find much, but again,
>> its not something I'm exactly experienced with.
> There's no real command-line tool for it. You could maybe extract some
> code from beagle-dump-index to do it, or try to decypher the incredibly
> hard-to-use but still useful tool Luke: http://www.getopt.org/luke/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v18.104.22.168 (GNU/Linux)
-----END PGP SIGNATURE-----
Dashboard-hackers mailing list