[GitHub] [lucene] gsmiller commented on pull request #12055: Better skipping for multi-term queries with a FILTER rewrite.

2023-03-02 Thread via GitHub
gsmiller commented on PR #12055: URL: https://github.com/apache/lucene/pull/12055#issuecomment-1451951125 Nightly bench runs have refreshed and show some nice improvements. I'll try to add some annotations soon. * https://home.apache.org/~mikemccand/lucenebench/Prefix3.html * https

[GitHub] [lucene] jpountz commented on pull request #12055: Better skipping for multi-term queries with a FILTER rewrite.

2023-03-02 Thread via GitHub
jpountz commented on PR #12055: URL: https://github.com/apache/lucene/pull/12055#issuecomment-1451952360 Woohoo! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscr

[GitHub] [lucene] gsmiller opened a new pull request, #12179: Better PostingsEnum reuse in MultiTermQueryConstantScoreBlendedWrapper

2023-03-02 Thread via GitHub
gsmiller opened a new pull request, #12179: URL: https://github.com/apache/lucene/pull/12179 ### Description We can be a bit better about PostingsEnum reuse when backfilling collected terms (i.e., reuse the same postings if we consume them into the bitset). -- This is an automated

[GitHub] [lucene] benwtrent commented on pull request #12160: Concurrent rewrite for KnnVectorQuery

2023-03-02 Thread via GitHub
benwtrent commented on PR #12160: URL: https://github.com/apache/lucene/pull/12160#issuecomment-1452062267 I foresee backporting this to 9.6. Unless @kaivalnp @jpountz or @zhaih object. -- This is an automated message from the Apache Git Service. To respond to the message, please log on t

[GitHub] [lucene] jpountz commented on pull request #12160: Concurrent rewrite for KnnVectorQuery

2023-03-02 Thread via GitHub
jpountz commented on PR #12160: URL: https://github.com/apache/lucene/pull/12160#issuecomment-1452084626 +1 to backporting -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. T

[GitHub] [lucene] zhaih commented on pull request #12160: Concurrent rewrite for KnnVectorQuery

2023-03-02 Thread via GitHub
zhaih commented on PR #12160: URL: https://github.com/apache/lucene/pull/12160#issuecomment-1452243542 @benwtrent I think the [concurrent rewrite API change](https://github.com/apache/lucene/pull/11840) is only at Lucene10 right now, it is a bit tricky to backport it to 9x and not yet done

[GitHub] [lucene] benwtrent commented on pull request #12160: Concurrent rewrite for KnnVectorQuery

2023-03-02 Thread via GitHub
benwtrent commented on PR #12160: URL: https://github.com/apache/lucene/pull/12160#issuecomment-1452311392 Thank you for the context @zhaih, Do you mind then guiding this PR through the merging and potential backporting process? Backporting would indeed be blocked until your change c

[GitHub] [lucene] gsmiller opened a new issue, #12180: Add bulk path -> ordinal lookup for taxonomy faceting

2023-03-02 Thread via GitHub
gsmiller opened a new issue, #12180: URL: https://github.com/apache/lucene/issues/12180 ### Description [LUCENE-9476](https://issues.apache.org/jira/browse/LUCENE-9476) added the ability to do bulk ordinal -> path lookups, but we have no bulk lookup in the other direction. I think we

[GitHub] [lucene] zhaih commented on pull request #12160: Concurrent rewrite for KnnVectorQuery

2023-03-02 Thread via GitHub
zhaih commented on PR #12160: URL: https://github.com/apache/lucene/pull/12160#issuecomment-1452384680 > Do you mind then guiding this PR through the merging and potential backporting process? Yeah I can help with that, I would suggest let's keep it in Lucene 10 for now, and if the p

[GitHub] [lucene] Trey314159 commented on pull request #12172: Add Romanian stopwords with s&t with comma

2023-03-02 Thread via GitHub
Trey314159 commented on PR #12172: URL: https://github.com/apache/lucene/pull/12172#issuecomment-1452625462 > After reading up on the history of these characters, I think we should treat them "the same" for Romanian always. Yeah, I agree. > I think a filter may be worthwhile as

[GitHub] [lucene] rmuir commented on pull request #12172: Add Romanian stopwords with s&t with comma

2023-03-02 Thread via GitHub
rmuir commented on PR #12172: URL: https://github.com/apache/lucene/pull/12172#issuecomment-1452764003 I think its actually quite easy to fix the stemmer if we want to just send them a pull request. I can help if you don't want to do it, I don't want to steal your thunder though :) T

[GitHub] [lucene] david-sitsky commented on issue #10025: Make NativeUnixDirectory pure java now that direct IO is possible [LUCENE-8982]

2023-03-02 Thread via GitHub
david-sitsky commented on issue #10025: URL: https://github.com/apache/lucene/issues/10025#issuecomment-1453005213 > Michael McCandless (@mikemccand) ([migrated from JIRA](https://issues.apache.org/jira/browse/LUCENE-8982?focusedCommentId=17223693&page=com.atlassian.jira.plugin.system.issuet