javanna commented on PR #14284:
URL: https://github.com/apache/lucene/pull/14284#issuecomment-2681021215
> FWIW it looks like there are other cleanups to be done in this class, e.g.
removing DocIdSet#all(int) and DocIdSet#bits() (another remainder from the
query/filter merge).
Thanks
javanna opened a new pull request, #14288:
URL: https://github.com/apache/lucene/pull/14288
DocIdSet#all is no longer relevant since Query and Filter were merged. We
can remove it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to Git
javanna commented on PR #14284:
URL: https://github.com/apache/lucene/pull/14284#issuecomment-2680997183
> pulling an iterator shouldn't throw an I/O exception
agreed, that was also my thinking when I made the change.
> Should this be a 11.0 change rather than 10.2?
I do
gf2121 commented on PR #14277:
URL: https://github.com/apache/lucene/pull/14277#issuecomment-2680655769
> (Longer-term, I'm thinking of removing the query cache
https://github.com/apache/lucene/pull/14187)
Thanks for explanation, I understand the motivation that we are focusing on
sk
gf2121 closed pull request #14277: Introduce DocIdStream#intoBitset to speed up
cache
URL: https://github.com/apache/lucene/pull/14277
--
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 co
dungba88 commented on PR #14226:
URL: https://github.com/apache/lucene/pull/14226#issuecomment-2680433274
@msokolov I didn't see the change to cap `perLeafTopK` in the [latest
commit](https://github.com/apache/lucene/pull/14226/commits/5b06e168c683e3edf36987379091357c298f0f28#diff-6bf79d1f0e
asfgit merged PR #14286:
URL: https://github.com/apache/lucene/pull/14286
--
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 unsubscribe, e-mail: issues-unsubscr...@lucene.apac
rmuir commented on PR #14286:
URL: https://github.com/apache/lucene/pull/14286#issuecomment-2680113100
Thank you @msfroh !
--
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-actions[bot] commented on PR #14194:
URL: https://github.com/apache/lucene/pull/14194#issuecomment-2680017293
This PR has not had activity in the past 2 weeks, labeling it as stale. If
the PR is waiting for review, notify the d...@lucene.apache.org list. Thank you
for your contributi
rmuir opened a new pull request, #14287:
URL: https://github.com/apache/lucene/pull/14287
See background by @mkhludnev on the dev list:
https://lists.apache.org/thread/pm1szr9og6qhmjzp371xwk0mvwxxkd1l
In some cases: "invalid reference" is generated, passes through Xdoclint and
broken
rmuir commented on PR #14286:
URL: https://github.com/apache/lucene/pull/14286#issuecomment-2679880554
This crazy SleepingLockWrapper still lives in main/, but its
`DEFAULT_POLL_INTERVAL` is `static final` there. I wonder if this got fixed in
another commit or something.
--
This is an au
msfroh commented on PR #14286:
URL: https://github.com/apache/lucene/pull/14286#issuecomment-2679883061
> This crazy SleepingLockWrapper still lives in main/, but its
`DEFAULT_POLL_INTERVAL` is `static final` there. I wonder if this got fixed in
another commit or something.
Yeah, I f
msfroh commented on PR #14286:
URL: https://github.com/apache/lucene/pull/14286#issuecomment-2679879073
> Looks like CI fails in a check on core/. there might be more, if it gets
past core/:
>
Oh... dang, I forgot to run locally with error-prone enabled. I'll do that
and clean up
rmuir commented on PR #14286:
URL: https://github.com/apache/lucene/pull/14286#issuecomment-2679876336
Looks like CI fails in a check on core/. there might be more, if it gets
past core/:
```
/home/runner/work/lucene/lucene/lucene/core/src/java/org/apache/lucene/store/SleepingLock
msfroh opened a new pull request, #14286:
URL: https://github.com/apache/lucene/pull/14286
### Description
This change backports the following PRs to 10.x:
*
https://github.com/apache/lucene/commit/4cd4f8e2f99b8e300fdd996b54920625a257acf7
* https://github.com/apache/lucene/p
jpountz commented on PR #14277:
URL: https://github.com/apache/lucene/pull/14277#issuecomment-2679834513
To be honest, I'm a bit on the fence about introducing specialization for
caching. This is why I was wondering if faceting could benefit from it too,
though I'm not a fan of the fact tha
jpountz commented on PR #14187:
URL: https://github.com/apache/lucene/pull/14187#issuecomment-2679832873
For the record, another downside of the query cache: the fact that it caches
per segment doesn't play nicely with intra segment concurrency. E.g. if you
have a single large segment that
jpountz commented on PR #14277:
URL: https://github.com/apache/lucene/pull/14277#issuecomment-2679821035
I had thought of something in-between your previous PR and this one. E.g.
adding `BulkScorer#intoBitSet` with a similar signature and contract as
`DocIdSetIterator#intoBitSet`.
--
Thi
rmuir commented on issue #14285:
URL: https://github.com/apache/lucene/issues/14285#issuecomment-2679794279
--
This is an automated message from the Apache Git Service.
To respond to the message, ple
rmuir opened a new issue, #14285:
URL: https://github.com/apache/lucene/issues/14285
### Description
See background by @mkhludnev on the dev list:
https://lists.apache.org/thread/pm1szr9og6qhmjzp371xwk0mvwxxkd1l
In some cases: "invalid reference" is generated, passes through Xd
sgup432 commented on issue #14183:
URL: https://github.com/apache/lucene/issues/14183#issuecomment-2679778282
@jpountz Just checking if you’ve had a chance to look into this. As
mentioned, I believe dynamically adjusting `skip_factor` would be beneficial.
Additionally, we can also introduc
dweiss commented on PR #14228:
URL: https://github.com/apache/lucene/pull/14228#issuecomment-2679643415
Thanks, @msfroh ! I think it should be possible to cherry pick a series of
commits from main - perhaps with minor adjustments. When you try to cherry pick
this commit from main, you'll se
javanna opened a new pull request, #14284:
URL: https://github.com/apache/lucene/pull/14284
There is no implementation of DocIdSet#iterator that throws IOException.
This commit proposes removing throwing IOException from its signature.
--
This is an automated message from the Apache Git S
msfroh commented on PR #14228:
URL: https://github.com/apache/lucene/pull/14228#issuecomment-2679564727
> I think a backport to 10x would be nice but perhaps it's worth another
issue since it's not trivial.
I can take care of that.
After going down the rabbit hole once with t
rmuir commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2679383631
Yes, I'm just suggesting to split it. We can add this new parameter here,
backport to minor release 10.2.0, no breaking changes. Separately we can
default it to `true` for 11.0?
--
Thi
sgup432 commented on issue #14222:
URL: https://github.com/apache/lucene/issues/14222#issuecomment-2679321128
>But in a typical workload we expect to be spending most of our time
executing queries rather than caching them, which will reduce the amount of
time spent acquiring locks, and the
pseudo-nymous commented on PR #14279:
URL: https://github.com/apache/lucene/pull/14279#issuecomment-2679232219
There were small fixes which I missed in previous commit. This didn't had
any history attached to it and I kept the commit which had comments.
Also, I'll keep this in check
rmuir commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2679239148
for changing defaults, my goto would be, if we could do that as a followup
PR, for a major release.
We can expose this parameter in a minor release without hurting anyone, but
if w
renatoh commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2679257613
I would argue, at least in German, nothing but longestMatch=true and
skipping forward does make any sense. Without skipping forward the filter
extracts a lot of nonsense and in my opinio
rmuir commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2679218267
I'm not really opinionated on it, was just brainstorming because I had to
look at the source code to figure out what the parameter was doing.
And I agree, it is surprising behavior
renatoh commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2679176831
> looks good to me. I wonder about the name of the parameter, maybe "greedy"
would be more intuitive as a way to describe what it is doing?
not saying "consumeChars" is a good name
benwtrent commented on PR #14078:
URL: https://github.com/apache/lucene/pull/14078#issuecomment-2679008833
@gaoj0017
> The OSQ method (introduced in this PR) has its major idea similar to our
extended RaBitQ method and our extended RaBitQ method is a prior art which
achieves good ac
benwtrent commented on PR #14226:
URL: https://github.com/apache/lucene/pull/14226#issuecomment-2679029410
> so I rebased and removed the reuse-scores part of this since it was
conflicting with other changes and doesn't seem worth preserving
@msokolov thanks for confirming and digging
benwtrent commented on PR #14078:
URL: https://github.com/apache/lucene/pull/14078#issuecomment-2679022162
> I wonder where can I find the code for the benchmarks that you are
mentioning in the description? Thanks!
@lpld I patched a version of Lucene util, sort of like this:
https://
javanna merged PR #14282:
URL: https://github.com/apache/lucene/pull/14282
--
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 unsubscribe, e-mail: issues-unsubscr...@lucene.apa
msokolov commented on PR #14226:
URL: https://github.com/apache/lucene/pull/14226#issuecomment-2678966117
There were some conflicts with other recent changes, so I rebased and
removed the reuse-scores part of this since it was conflicting with other
changes and doesn't seem worth preserving
javanna commented on PR #14270:
URL: https://github.com/apache/lucene/pull/14270#issuecomment-2678961874
@jpountz do you have opinions on this? Who else should I ping otherwise?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
javanna commented on PR #14282:
URL: https://github.com/apache/lucene/pull/14282#issuecomment-2678958683
thanks for the reference, I found that in the history but I did not have all
the context, it's good to know. I was also wondering if this could be perceived
as a breaking change and I th
jpountz commented on PR #14273:
URL: https://github.com/apache/lucene/pull/14273#issuecomment-2678921217
> Looks like we can do similar trick for range facets and long values facets?
This is right.
--
This is an automated message from the Apache Git Service.
To respond to the messag
jpountz commented on issue #14281:
URL: https://github.com/apache/lucene/issues/14281#issuecomment-2678895153
This is interesting, the linked change is expected to replace some
potentially expensive calls to madvise (since they need to iterate all pages)
with cheaper checks, so I wouldn't h
jpountz opened a new issue, #14283:
URL: https://github.com/apache/lucene/issues/14283
`DenseConjunctionBulkScorer` has good test coverage, but we don't test that
it correctly reacts to the min competitive score being set to a higher value
than the constant score, or to a competitive iterat
rmuir commented on code in PR #14278:
URL: https://github.com/apache/lucene/pull/14278#discussion_r1967791155
##
lucene/analysis/common/src/test/org/apache/lucene/analysis/compound/TestCompoundWordTokenFilter.java:
##
@@ -682,4 +687,41 @@ protected TokenStreamComponents createCo
rmuir commented on PR #14278:
URL: https://github.com/apache/lucene/pull/14278#issuecomment-2678643974
looks good to me. I wonder about the name of the parameter, maybe "greedy"
would be more intuitive as a way to describe what it is doing?
--
This is an automated message from the Apache
msokolov commented on PR #14226:
URL: https://github.com/apache/lucene/pull/14226#issuecomment-2678496890
I pushed a version that re-uses scores *and* limits per-leaf topK to global
topK. The former didn't make very much difference, but the latter change did
improve things quite a bit. He
javanna opened a new pull request, #14282:
URL: https://github.com/apache/lucene/pull/14282
The argument is unused, its callers can stop providing it.
--
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
epotyom commented on code in PR #14273:
URL: https://github.com/apache/lucene/pull/14273#discussion_r1967691866
##
lucene/core/src/java/org/apache/lucene/search/DocIdStream.java:
##
@@ -34,12 +33,35 @@ protected DocIdStream() {}
* Iterate over doc IDs contained in this strea
dweiss commented on PR #14279:
URL: https://github.com/apache/lucene/pull/14279#issuecomment-2678495642
Please do not force-push your changes. Keep prior commits so that the
history is preserved and comments are attached where they should be. Thanks!
--
This is an automated message from t
msokolov commented on issue #14222:
URL: https://github.com/apache/lucene/issues/14222#issuecomment-2678412268
This shows a nice improvement on the microbenchmark! But in a typical
workload we expect to be spending most of our time executing queries rather
than caching them, which will redu
benwtrent commented on PR #14256:
URL: https://github.com/apache/lucene/pull/14256#issuecomment-2678310512
> it might suffice to create a wrapping RandomVectorScorer that would
supply the cached scores, while delegating the others to the underlying scorer?
🤔 Maybe we could wrap the
benwtrent commented on code in PR #14256:
URL: https://github.com/apache/lucene/pull/14256#discussion_r1967569988
##
lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraphBuilder.java:
##
@@ -229,25 +230,32 @@ public void addGraphNode(int node,
UpdateableRandomVectorScorer
pseudo-nymous commented on code in PR #14279:
URL: https://github.com/apache/lucene/pull/14279#discussion_r1967566449
##
.github/workflows/verify-changelog-and-set-milestone.yml:
##
@@ -0,0 +1,100 @@
+name: "Change Log Entry Verifier and Milestone Setter"
+run-name: Change log e
dweiss commented on PR #14228:
URL: https://github.com/apache/lucene/pull/14228#issuecomment-2678269055
I've merged this. There are conflicts when backporting to branch_10x and
they sort of depend on previous fixes to spotless, so I leave this patch on
main only. I think a backport to 10x w
pseudo-nymous commented on code in PR #14279:
URL: https://github.com/apache/lucene/pull/14279#discussion_r1967542895
##
.github/workflows/verify-changelog-and-set-milestone.yml:
##
@@ -0,0 +1,100 @@
+name: "Change Log Entry Verifier and Milestone Setter"
+run-name: Change log e
pseudo-nymous commented on code in PR #14279:
URL: https://github.com/apache/lucene/pull/14279#discussion_r1967542895
##
.github/workflows/verify-changelog-and-set-milestone.yml:
##
@@ -0,0 +1,100 @@
+name: "Change Log Entry Verifier and Milestone Setter"
+run-name: Change log e
dweiss merged PR #14228:
URL: https://github.com/apache/lucene/pull/14228
--
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 unsubscribe, e-mail: issues-unsubscr...@lucene.apac
pseudo-nymous commented on code in PR #14279:
URL: https://github.com/apache/lucene/pull/14279#discussion_r1967537809
##
.github/workflows/verify-changelog-and-set-milestone.yml:
##
@@ -0,0 +1,100 @@
+name: "Change Log Entry Verifier and Milestone Setter"
+run-name: Change log e
pseudo-nymous commented on code in PR #14279:
URL: https://github.com/apache/lucene/pull/14279#discussion_r1967532868
##
.github/workflows/verify-changelog-and-set-milestone.yml:
##
@@ -0,0 +1,100 @@
+name: "Change Log Entry Verifier and Milestone Setter"
+run-name: Change log e
lpld commented on PR #14078:
URL: https://github.com/apache/lucene/pull/14078#issuecomment-2678033882
Hi @benwtrent, that's an amazing amount of work.
I wonder where can I find the code for the benchmarks that you are
mentioning in the description? Thanks!
--
This is an automated m
imario42 opened a new pull request, #14280:
URL: https://github.com/apache/lucene/pull/14280
lazy initialize the ScorerUtil DEFAULT_IMPACTS_ENUM_CLASS to prevent
initialization issues with this class if the thread gets interrupted.
### Description
The ScorerUtils fails to initi
javanna commented on code in PR #13470:
URL: https://github.com/apache/lucene/pull/13470#discussion_r1967234360
##
lucene/core/src/java/org/apache/lucene/search/TopDocs.java:
##
@@ -350,4 +354,89 @@ private static TopDocs mergeAux(
return new TopFieldDocs(totalHits, hits,
jpountz commented on PR #13470:
URL: https://github.com/apache/lucene/pull/13470#issuecomment-2677780646
Thanks for taking a look. I have a bias for the latter, as I was planning on
improving the docs of the oal.search package as a follow-up to provide guidance
wrt how to do hybrid search b
javanna commented on PR #13470:
URL: https://github.com/apache/lucene/pull/13470#issuecomment-263959
This looks good to me. Perhaps we could mark the new static method
experimental, especially if we think we are going to want to support more ways
of combining topdocs soon enough. I don'
62 matches
Mail list logo
