Skip to content

Compute the doc range more efficiently when flushing doc block #14447

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
Apr 23, 2025
Merged

Compute the doc range more efficiently when flushing doc block #14447

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
9d82782
compute the doc range more efficiently when flushing doc block
bugmakerrrrrr Apr 7, 2025
d57925c
update
bugmakerrrrrr Apr 21, 2025
c111583
Merge remote-tracking branch 'upstream/main' into postings
bugmakerrrrrr Apr 23, 2025
30a16a2
changelog
bugmakerrrrrr Apr 23, 2025
92d1ba1
apply comment
bugmakerrrrrr Apr 23, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions lucene/CHANGES.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,6 +84,8 @@ Optimizations

* GITHUB#14333: Introduce a specialized trie for block tree index, instead of FST. (Guo Feng)

* GITHUB#14447: Compute the doc range more efficiently when flushing doc block. (Pan Guixin)

Bug Fixes
---------------------
(No changes)
Expand Down
9 changes: 5 additions & 4 deletions lucene/core/src/java/org/apache/lucene/codecs/lucene103/Lucene103PostingsWriter.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -429,12 +429,13 @@ private void flushDocBlock(boolean finishTerm) throws IOException {
// storage-efficient than the next number of bits per value (which effectively slightly biases
// towards the bit set approach).
int bitsPerValue = forDeltaUtil.bitsRequired(docDeltaBuffer);
int sum = Math.toIntExact(Arrays.stream(docDeltaBuffer).sum());
int numBitSetLongs = FixedBitSet.bits2words(sum);
int docRange = lastDocID - level0LastDocID;
assert docRange == Arrays.stream(docDeltaBuffer).sum();
int numBitSetLongs = FixedBitSet.bits2words(docRange);
int numBitsNextBitsPerValue = Math.min(Integer.SIZE, bitsPerValue + 1) * BLOCK_SIZE;
if (sum == BLOCK_SIZE) {
if (docRange == BLOCK_SIZE) {
level0Output.writeByte((byte) 0);
} else if (numBitsNextBitsPerValue <= sum) {
} else if (numBitsNextBitsPerValue <= docRange) {
level0Output.writeByte((byte) bitsPerValue);
forDeltaUtil.encodeDeltas(bitsPerValue, docDeltaBuffer, level0Output);
} else {
Expand Down