Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
37,179 Commits 29 Branches 469 Tags 768 MiB
Commit Graph

37179 Commits

Author SHA1 Message Date
Zhang Chao 55df3e02f3
Fix too many open files Exception for some tests (#13035) 
* init

* fix review

* fix review

* iter
 2024-02-19 05:35:24 -05:00
Dmitry Cherniachenko 9a88d8ad25
Remove redundant field initializers (#13060) 
Make protected fields in final classes private.
 2024-02-19 09:52:43 +00:00
Benjamin Trent 39f681812b
Adding a getter to FeatureField to allow value access (#13111) 
Getting the access of a FeatureField#value is useful for deduplicating. If you have a sparse vector model and you want to handle multiple inputs from them, you want flexibility in how you de-duplicate the feature dimensions.
 2024-02-17 09:03:08 -05:00
Benjamin Trent af0d0edf5e
Fix test failure in #13057 (#13102) 
The original tests assume particular document orders & scores. To make the test more resilient to random flushes & merges, I adjusted the assertion conditions. Particularly, we verify matching ids -> score instead of relying on docIds.

closes: https://github.com/apache/lucene/issues/13057
 2024-02-16 09:33:34 -05:00
Dawid Weiss cc25fd9deb React to mod-analysis-common workflow path (regression after workflow renamed). #13110 2024-02-16 12:40:49 +01:00
Dawid Weiss aa7d035ac2
github workflow cleanup and minor improvements (#13110) 2024-02-16 12:38:21 +01:00
Dmitry Cherniachenko 9206bdeb06
Move `brToString(BytesRef)` to `ToStringUtils` (#13068) 2024-02-15 18:16:44 +01:00
Zhang Chao b16d7117a3
Fix NPE for TestRandomChains.testRandomChainsWithLargeStrings (#13104) 2024-02-15 16:20:53 +01:00
Dmitry Cherniachenko f24b1de351
Cleanup TokenizedPhraseQueryNode code (#13041) 2024-02-15 10:04:58 +00:00
Simon Willnauer c9e4434cc3 Fix addBackcompatIndexes.py to properly generate missing versions (#13095) 
In #13046 several changes broke the addBackcompatIndexes.py script
to properly add and test the unreleased version. This updates the
script to again properly add the new version.

Closes #13094

Co-authored-by: Dawid Weiss <dawid.weiss@carrotsearch.com>
 2024-02-14 21:58:30 +01:00
Dawid Weiss 444d816622 Moved changes entry to 9.11 #13092 2024-02-14 11:32:55 +01:00
Dmitry Cherniachenko 3768b2e1e0
Make `static final Map` constants immutable (#13092) 2024-02-13 19:02:55 +01:00
Dawid Weiss a270acae01
This reverts the addition of spotless:on/off regions and shows just one possible alternative that is formatter fool-proof. (#13098) 2024-02-13 19:00:11 +01:00
Uwe Schindler 2639aea5dd
Change the links to Lucene's homepage in the DOAP file to https:// (#13097) 2024-02-13 14:19:33 +01:00
Simon Willnauer c42380cbd0 Enable parent field in sorted bwc tests (#13067) 
This enables the optional parent field in BWC tests from 9.10 on.
This will need to be forward ported to main branch where the
parent field is required to these tests since they add document blocks
during tests.
 2024-02-13 12:00:29 +01:00
Adrien Grand 6f026d7c36 Fix logic to always exclude latest minor from previous major. 2024-02-13 10:59:16 +01:00
Adrien Grand d7e0a7f0e4
Only track released versions in `oldVersions`. (#13096) 2024-02-13 10:48:30 +01:00
Zhang Chao f6e065dc31
Fix test failure for TestDocumentsWriterDeleteQueue.testUpdateDeleteSlices (#13089) 2024-02-13 10:10:44 +08:00
Adrien Grand 027240f522 Tidy after version bump. 2024-02-12 17:01:47 +01:00
Adrien Grand 314c553bdc Add next major version 9.11.0 2024-02-12 14:36:17 +01:00
Christine Poerschke 8bf10130e9 lucene/MIGRATE.md update: remove empty 9.9 to 9.10 section 2024-02-12 10:17:41 +00:00
Dawid Weiss 7d99e6aafb Move changes and migration entries to Lucene 10.x since it seems there'll be no way to apply this patch in 9x? #12674 2024-02-11 11:14:58 +01:00
Uwe Schindler 7e921e4ed9 fix typo in CHANGES.txt 2024-02-09 23:15:05 +01:00
Uwe Schindler 178f5a7a7e
Enable MemorySegment in MMapDirectory for Java 22+ and Vectorization (incubation) for exact Java 22 (#12706) 2024-02-09 23:02:42 +01:00
Robert Muir d7a16dc10a
Merge branch 'main' into java_21 2024-02-09 15:22:45 -05:00
Robert Muir ec04490bd2
java 17 -> java 21 2024-02-09 15:18:30 -05:00
Robert Muir 23ecac438a
java 17 -> java 21 2024-02-09 15:17:46 -05:00
Robert Muir fefde0f721
java 17 -> java 21 2024-02-09 15:16:41 -05:00
Robert Muir 1f9545e830
remove java < 21 2024-02-09 15:15:52 -05:00
Robert Muir 784c331b68
java 17 -> java 21 2024-02-09 15:14:55 -05:00
Robert Muir a221b1f677
java 17 -> java 21 2024-02-09 15:14:20 -05:00
Robert Muir d75b869d21
java 17 -> java 21 2024-02-09 15:13:37 -05:00
Zhang Chao a1faa6398b
Add necessary assertion in CheckHits#doCheckMaxScores (#13088) 2024-02-09 18:28:18 +01:00
Christine Poerschke 563fafd8ac
upgrade to OpenNLP 2.3.2 (#12674) 2024-02-09 11:21:41 +00:00
Houston Putman 521c26e4fb Add back-compat indices for 8.11.3 2024-02-08 19:25:40 -06:00
Houston Putman efdbce412a Add bugfix version 8.11.3 2024-02-08 18:16:59 -06:00
Houston Putman 01ed004e94
Fix positioning of 8.11.3 DOAP RDF entry 2024-02-08 18:49:11 -05:00
Houston Putman 06101aa85d
Add 8.11.3 release to DOAP RDF file 2024-02-08 18:43:48 -05:00
Benjamin Trent 7da509b708
Prevent humongous allocations when calculating scalar quantiles (#13090) 
The initial release of scalar quantization would periodically create a humongous allocation, which can put unwarranted pressure on the GC & on the heap usage as a whole.

This commit adjusts this by only allocating a float array of 20*dimensions and averaging the discovered quantiles from there. 

Why does this work?

 - Quantiles based on confidence intervals are (generally) unbiased and doing an average gives statistically good results
 - The selector algorithm scales linearly, so the cost is just about the same
 - We need to do more than `1` vector at a time to prevent extreme confidence intervals interacting strangely with edge cases
 2024-02-08 15:56:37 -05:00
Benjamin Trent 860a097ab3
Moving quantization logic to make future quantizer work simpler (#13091) 2024-02-08 09:38:07 -05:00
Stefan Vodita f339e24e8e
Index arbitrary fields in taxonomy docs (#12337) 2024-02-08 11:02:12 +00:00
Christine Poerschke bff5ac0ed0
in BytesRefHash constructor avoid duplicate BytesStartArray.bytesUsed() call (#13032) 2024-02-08 09:02:40 +01:00
Benjamin Trent e5bceb4086
Fix TestTopFieldCollector.testTotalHits #13080 (#13081) 2024-02-07 12:55:55 -05:00
Benjamin Trent 810e43c58a
Fix test failure TestParentBlockJoinFloatKnnVectorQuery.testSkewedIndex (#13082) 2024-02-07 12:55:43 -05:00
Zhang Chao 5e3e7465c3
Do not use mock merge policy for TestFuzzyQuery#testFuzziness (#13070) 2024-02-07 18:05:32 +01:00
Zhang Chao ce0fa2a6a6
Fix NPE in TestReqOptSumScorer.testFilterRandomRareOpt #13069 2024-02-07 13:14:01 +01:00
Dawid Weiss 8c2c276c6c Modify getEnWikiRandomLines to fetch and decompress the zstd resource #13083 2024-02-06 22:08:09 +01:00
Benjamin Trent 681fa21665
Fix knn vector visit limit fence post error (#13058) 
I noticed while experimenting with brute-force search that our visitation limit is EXACTLY the number of filtered docs to hit. Consequently, if we happen to do brute force search and visit that exact number of vectors, we will fall back again to do brute-force a second time. This struck me as weird.

This commit adjusts the visit limit threshold for approximate search to account for this.
 2024-02-06 12:37:12 -05:00
Mayya Sharipova d095ed02a2
Speedup concurrent multi-segment HNWS graph search (#12962) 
Speedup concurrent multi-segment HNWS graph search by exchanging 
the global top candidated collected so far across segments. These global top 
candidates set the minimum threshold that new candidates need to pass
 to be considered. This allows earlier stopping for segments that don't have 
good candidates.
 2024-02-06 09:16:06 -05:00
Dawid Weiss 635d09001a
Make date parsing more flexible for linedocsfile (europarl, enwiki) (#13075) 2024-02-05 19:04:07 +01:00