89 Commits

Author SHA1 Message Date
Jean-Francois Dockes
7b5a891ee3 idx: make Doc parameter to addOrUpdate non const to avoid extra copy 2012-03-07 08:34:25 +01:00
Jean-Francois Dockes
9bc2fc8958 Experimented with multithreading the indexing pipeline. Left undef'd as 15%-30% improvement of indexing time does not seem worth the complexity 2012-02-21 17:09:02 +01:00
Jean-Francois Dockes
4ced9bee49 add termDocCnt method 2011-10-04 08:04:17 +02:00
Jean-Francois Dockes
38e0957962 const string cleanup 2011-10-01 16:39:38 +02:00
Jean-Francois Dockes
ee0d602ab3 Implement anchored searches: terms to be found at a maximum distance of the start or end of the text 2011-09-20 16:42:56 +02:00
Jean-Francois Dockes
c5ff0cdf52 Control memory usage when deleting documents: use idxflushmb as when adding/updating 2011-09-07 19:11:11 +02:00
"Jean-Francois Dockes ext:(%22)
ebbcc115a8 Allow setting a weight increase for field terms 2011-07-22 16:43:39 +02:00
Jean-Francois Dockes
469c544915 GUI: allow setting the snippet separator inside abstract (now a real html ellipsis by default) 2011-07-07 11:11:02 +02:00
Jean-Francois Dockes
08a65f5cfc experiment with xapian spell support (not ready yet) + take care of some static init issues showing up on the mac 2011-05-10 10:15:15 +02:00
Jean-Francois Dockes
84d59f18a0 GUI: when opening the index, discriminate errors on the main index from errors on external ones, to avoid starting the initial indexing dialog in the latter case 2011-04-29 16:16:04 +02:00
Jean-Francois Dockes
01f24fa5fd cleaning up static variables 2011-04-27 09:09:01 +02:00
Jean-Francois Dockes
b28eaf23fb Got rid of all the old RCS id strings 2011-04-27 08:22:17 +02:00
Jean-Francois Dockes
21c6025ba7 Use a xapian phrase search on the split path for filtering on directory location (much faster than the current method) 2010-12-16 15:53:40 +01:00
Jean-Francois Dockes
ceb996c8fb Implement date: date range filter/searches. Remove restriction on pure negative queries 2010-09-11 12:07:53 +02:00
"Jean-Francois Dockes ext:(%22)
846bec8a73 fix english indexation -> indexing 2010-07-20 09:48:20 +02:00
Jean-Francois Dockes
4006825961 display more complete stats in spell window 2010-05-08 10:38:13 +02:00
dockes
bab030f846 Term expansion: handle field issues inside rcldb::termmatch, ensuring that we take the field name into account for all expansions. Ensures that File Name searches and filename: query language searches work the same, + overall better consistency 2009-12-07 13:27:57 +00:00
dockes
6ef7b546f2 beaglequeue indexFiles 2009-11-14 08:21:45 +00:00
dockes
645af1aa55 removed the useless keep_updated flag 2009-11-10 17:32:13 +00:00
dockes
0bd1b1a674 renamed fields for clarity 2009-10-24 11:00:00 +00:00
dockes
a27ddc32c4 rclversion.h must not include xapian.h. Replace with Rcl::version_string() 2009-10-17 06:38:04 +00:00
dockes
7dcc7c61c8 modified the time at which we unaccent so that we can do the Capitalized->nostemming test on single words (this had been broken by the change of noac/split order done earlier to get japanese to work) 2009-01-26 18:30:48 +00:00
dockes
98dcf8c9ca minor visibility cleanup 2009-01-13 08:02:29 +00:00
dockes
f57d4a91f9 compute md5 checksums for all docs and optionally collapse duplicates in results 2009-01-09 14:56:36 +00:00
dockes
d9b9b41a9d getMainConfig not actually needed and possibly harmful 2008-12-19 09:55:36 +00:00
dockes
b9cc5fb1e0 simplified db open by getting rid of the illusion that we could have several writeable dbs per config 2008-12-17 08:01:40 +00:00
dockes
828dff3bb1 doc.pc now only place where relevancy is stored 2008-09-29 08:59:20 +00:00
dockes
7d30485f87 general field name handling cleanup + sort facility in rclquery 2008-09-16 08:18:30 +00:00
dockes
54a7b69f41 copy author back from data record to Doc 2008-08-26 07:38:29 +00:00
dockes
269efce55e comments 2008-07-29 08:25:43 +00:00
dockes
24ac62eb86 use explicit parent udi term instead of Qterm structure to express parent-child relationship 2008-07-29 06:25:29 +00:00
dockes
3109a33f4a replaced path|ipath with unique doc id in rcldb i/f. Still depends on udi structure for parent/child 2008-07-28 12:24:15 +00:00
dockes
ca4a4e65b0 begin i/f cleanup: opacify doc uptodate sig (size+mtime) 2008-07-28 08:42:52 +00:00
dockes
5166fc38cf comments 2008-07-01 08:28:45 +00:00
dockes
0e7a78d688 separated rcldb and rclquery 2008-06-13 18:22:47 +00:00
dockes
7368d14d2d recollindex -l 2007-07-10 09:23:28 +00:00
dockes
2777fedb78 slightly reorganized Db::close/~Db code 2007-06-21 11:56:28 +00:00
dockes
0c74bd6e36 added open-ended field name handling 2007-06-19 08:36:24 +00:00
dockes
18b3573358 implement dynamic field name to prefix translation, query side 2007-06-18 13:04:15 +00:00
dockes
6d8040c2f0 added file system usage check 2007-06-08 16:47:19 +00:00
dockes
7edcbba55e comments and ordering 2007-06-08 16:05:25 +00:00
dockes
0f1b917b7b minimal experimental stopword functionality 2007-06-02 08:30:42 +00:00
dockes
e04fbf3fb4 implemented adjustable indexing flush threshold 2007-05-22 07:40:00 +00:00
dockes
68b25b750c sort and uniquify termMatch results out of stem expansion 2007-02-02 10:10:53 +00:00
dockes
be05eaa6e0 merge stemExpand into termExpand. return term frequencies from there and display in spellW 2006-12-19 12:11:21 +00:00
dockes
1973c06346 split rcldb.h -> rcldoc.h 2006-12-14 14:54:13 +00:00
dockes
aff98f7fc9 expose abstract synthesis to let users decide when they want it done 2006-12-05 15:17:59 +00:00
dockes
7cc20a8f78 added dynamic clauses to adv search. Still needs work 2006-11-14 13:55:43 +00:00
dockes
cdbf026738 make searchdata a more flexible struct 2006-11-13 08:50:07 +00:00
dockes
1ab0a31c41 Turn spell tool into multimode spell/wild/regexp 2006-10-30 12:59:44 +00:00