iamsanjay commented on code in PR #13319:
URL: https://github.com/apache/lucene/pull/13319#discussion_r1587404169
##
lucene/core/src/java/org/apache/lucene/search/FilterWeight.java:
##
@@ -58,11 +58,6 @@ public Explanation explain(LeafReaderContext context, int
doc) throws IOEx
iamsanjay commented on PR #13319:
URL: https://github.com/apache/lucene/pull/13319#issuecomment-2090140897
Instead of both get() and cost() throwing the exception, Can make
scorerSupplier throw `UnsupportedOperationException()`?
```
@Override
public ScorerSupplier scorerSupplie
iamsanjay commented on code in PR #13319:
URL: https://github.com/apache/lucene/pull/13319#discussion_r1587434132
##
lucene/queries/src/java/org/apache/lucene/queries/spans/SpanWeight.java:
##
@@ -135,16 +135,6 @@ private Similarity.SimScorer buildSimWeight(
public abstract S
benwtrent closed issue #13326:
TestHnswBitVectorsFormat.testIndexAndSearchBitVectors fails intermittently
URL: https://github.com/apache/lucene/issues/13326
--
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 t
benwtrent merged PR #1:
URL: https://github.com/apache/lucene/pull/1
--
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.a
ChrisHegarty commented on PR #13336:
URL: https://github.com/apache/lucene/pull/13336#issuecomment-2090397935
Our scorer implementations should be able to take advantage of the ranges
being in:
* int7: 0 - 127 (inclusive)
* int4: 0 - 15 ( inclusive)
--
This is an automated message f
benwtrent merged PR #13336:
URL: https://github.com/apache/lucene/pull/13336
--
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.a
msokolov commented on code in PR #13181:
URL: https://github.com/apache/lucene/pull/13181#discussion_r1587804525
##
lucene/core/src/java/org/apache/lucene/util/quantization/QuantizedByteVectorValues.java:
##
@@ -18,13 +18,40 @@
import java.io.IOException;
import org.apache.l
msokolov commented on issue #9721:
URL: https://github.com/apache/lucene/issues/9721#issuecomment-2090843987
One thing came up during my testing / messing around that I think could
significantly affect the API we provide which is whether we want to bake in the
algorithm for computing leaves
jpountz opened a new pull request, #13337:
URL: https://github.com/apache/lucene/pull/13337
This adds `IndexInput#prefetch`, which is an optional operation that
instructs the `IndexInput` to start fetching bytes from storage in the
background. These bytes will be picked up by follow-up call
stefanvodita commented on issue #9721:
URL: https://github.com/apache/lucene/issues/9721#issuecomment-2090878680
> we may want to modify the amount of concurrency we apply to each Query in
response to operational conditions
I would really like it if we could do this. It could be a ver
jpountz commented on PR #13337:
URL: https://github.com/apache/lucene/pull/13337#issuecomment-2090882979
I created the following benchmark to simulate lookups in a terms
dictionary that cannot fit in the page cache.
```java
import java.io.IOException;
import java.nio
jpountz merged PR #13331:
URL: https://github.com/apache/lucene/pull/13331
--
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
benwtrent commented on code in PR #13181:
URL: https://github.com/apache/lucene/pull/13181#discussion_r1587945011
##
lucene/core/src/java/org/apache/lucene/search/VectorScorer.java:
##
@@ -18,64 +18,39 @@
import java.io.IOException;
import org.apache.lucene.index.ByteVectorV
jpountz commented on code in PR #13330:
URL: https://github.com/apache/lucene/pull/13330#discussion_r1587950557
##
lucene/codecs/src/java/org/apache/lucene/codecs/blockterms/VariableGapTermsIndexReader.java:
##
@@ -168,22 +165,6 @@ public FieldIndexEnum getFieldEnum(FieldInfo fi
jpountz commented on code in PR #13330:
URL: https://github.com/apache/lucene/pull/13330#discussion_r1587951466
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/blocktree/SegmentTermsEnum.java:
##
@@ -121,9 +121,6 @@ void initIndexInput() {
public Stats computeBlockS
benwtrent commented on code in PR #13181:
URL: https://github.com/apache/lucene/pull/13181#discussion_r1587957013
##
lucene/core/src/java/org/apache/lucene/util/quantization/QuantizedByteVectorValues.java:
##
@@ -18,13 +18,40 @@
import java.io.IOException;
import org.apache.
jpountz commented on PR #13319:
URL: https://github.com/apache/lucene/pull/13319#issuecomment-2091034684
> Instead of both get() and cost() throwing the exception, Can make
scorerSupplier throw UnsupportedOperationException()?
In general yes, there may be a few exceptions though like
jpountz commented on issue #9721:
URL: https://github.com/apache/lucene/issues/9721#issuecomment-2091045497
> I guess one alternative is to maintain multiple IndexSearchers with
different characteristics
Since IndexSearcher is very cheap to create, you could create a new
`IndexSearch
SaiSatwik opened a new issue, #13338:
URL: https://github.com/apache/lucene/issues/13338
### Description
We are seeing docs out of order error multiple times on Opensearch 1.2.3.
Seems issue is coming from lucene, but no clue what could be happening under
the hood. No much significan
mkhludnev commented on issue #13338:
URL: https://github.com/apache/lucene/issues/13338#issuecomment-2091183402
Hi @SaiSatwik
Are you running a some sort of test?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use th
msokolov commented on issue #9721:
URL: https://github.com/apache/lucene/issues/9721#issuecomment-2091192189
I don't know our IndexSearcher looks a little heavy; I think some of that is
our own doing and we could tease it apart, but isn't EG the query cache tied to
the IndexSearcher? And we
SaiSatwik commented on issue #13338:
URL: https://github.com/apache/lucene/issues/13338#issuecomment-2091193420
Hi @mkhludnev
No. We have seen this issue in a VM where single node opensearch deployment
is running. Due to this issue opensearch index went into RED state, leading to
fai
mkhludnev commented on issue #13338:
URL: https://github.com/apache/lucene/issues/13338#issuecomment-2091200687
Pardon. I had a wrong clue about a change in test framework occurring later.
Now, looking into the version you mention, I realized, it's wrong. Have no
idea. Don't you have index
SaiSatwik commented on issue #13338:
URL: https://github.com/apache/lucene/issues/13338#issuecomment-2091255775
@mkhludnev , we do not have index sorting configured. But, could you please
help me understand how can this issue could be related to index sorting
configuration?
--
This is an
mkhludnev commented on issue #13338:
URL: https://github.com/apache/lucene/issues/13338#issuecomment-2091305382
honestly, have no idea. How many docs you have in this index? May it exceed
2bns?
--
This is an automated message from the Apache Git Service.
To respond to the message, please
ChrisHegarty commented on code in PR #13339:
URL: https://github.com/apache/lucene/pull/13339#discussion_r1588276295
##
lucene/core/src/java/org/apache/lucene/internal/vectorization/VectorizationProvider.java:
##
@@ -198,6 +201,11 @@ private static void ensureCaller() {
priva
ChrisHegarty commented on code in PR #13339:
URL: https://github.com/apache/lucene/pull/13339#discussion_r1588276752
##
lucene/core/src/java21/org/apache/lucene/internal/vectorization/MemorySegmentByteVectorScorerSupplier.java:
##
@@ -0,0 +1,129 @@
+/*
+ * Licensed to the Apache
ChrisHegarty commented on code in PR #13339:
URL: https://github.com/apache/lucene/pull/13339#discussion_r1588277534
##
lucene/core/src/java21/org/apache/lucene/internal/vectorization/MemorySegmentByteVectorScorerSupplier.java:
##
@@ -0,0 +1,129 @@
+/*
+ * Licensed to the Apache
msokolov opened a new issue, #13340:
URL: https://github.com/apache/lucene/issues/13340
### Description
This test fails with ` java.lang.IllegalArgumentException: can't add a
parent field to an already existing index without a parent field`. If you index
any documents in the index, u
msokolov commented on issue #13340:
URL: https://github.com/apache/lucene/issues/13340#issuecomment-2091741574
I hope someone who is familiar with this will quickly see the problem - I'm
not sure if we (1) maybe fail to write any FieldInfos when the index is empty,
but now we must, or (2) f
msokolov commented on issue #13340:
URL: https://github.com/apache/lucene/issues/13340#issuecomment-2091762332
I tried removing
```
if (leaves.isEmpty()) {
uschindler commented on PR #13328:
URL: https://github.com/apache/lucene/pull/13328#issuecomment-2091805740
Thanks, looks fine. I have no time to do another closer review, please give
me some time to proceed.
--
This is an automated message from the Apache Git Service.
To respond to the m
msokolov commented on issue #13340:
URL: https://github.com/apache/lucene/issues/13340#issuecomment-2091822754
OK, now I see that field infos is part of the segment so we would not have
written it. I guess in this case we should explicitly recognize and work around
the empty case.
--
Thi
uschindler commented on code in PR #13328:
URL: https://github.com/apache/lucene/pull/13328#discussion_r1588461460
##
lucene/analysis/common/src/java/org/apache/lucene/analysis/hunspell/GeneratingSuggester.java:
##
@@ -447,14 +446,8 @@ private static int commonCharacterPositionS
msokolov commented on PR #13341:
URL: https://github.com/apache/lucene/pull/13341#issuecomment-2091836416
hm maybe this is not safe? If one creates a doc block filled with empty
documents? I'm not sure ...
--
This is an automated message from the Apache Git Service.
To respond to the mess
iamsanjay commented on PR #13319:
URL: https://github.com/apache/lucene/pull/13319#issuecomment-2092322936
> In general yes, there may be a few exceptions though like
`JustCompileSearch` where we want to keep these methods since it's about
detecting API changes.
Okay in that case, le
iamsanjay commented on PR #13319:
URL: https://github.com/apache/lucene/pull/13319#issuecomment-2092337036
I also changed scorerSupplier in few classes to delegate it to
scorerSupplier instead of scorer.
--
This is an automated message from the Apache Git Service.
To respond to the messa
bharath-techie opened a new pull request, #13342:
URL: https://github.com/apache/lucene/pull/13342
### Description
Draft PR for supporting extensions for DataCubes that mainly includes new
format , changes in SegmentInfo and in indexingchain flush , SegmentMerger
--
This is an
39 matches
Mail list logo
