The phrase matchers add a document position and a query position offset in int: with an offset near 2^31 an exact PhraseQuery or MultiPhraseQuery loops forever inside one document (IndexSearcher#setTimeout does not stop it) and a sloppy one returns false hits

Finding LFV-196Area: search / phrase matching (ExactPhraseMatcher, SloppyPhraseMatcher and PhrasePositions position arithmetic; PhraseQuery.Builder and MultiPhraseQuery.Builder accept the positions)Priority: lowImpact: hangNovelty: new

Finding page for LFV-196

AI-assisted finding, confirmed by automated checks, not yet triaged by a human Lucene committer. This site files nothing upstream.

What it shows

You build a PhraseQuery with term a at position 0 and term b at position Integer.MAX_VALUE (the builder accepts it, because it only checks that positions are >= 0 and in order) and count it on a one-document index whose text is 'x a b'. You expect 0 hits, because no indexed position can be that far from a. Instead the exact matcher adds a's document position 1 and b's offset Integer.MAX_VALUE in int, the sum wraps to a negative number, both of its advance steps then leave its state unchanged, and IndexSearcher#count never returns, even with a 1 s IndexSearcher#setTimeout; with slop 1 the same kind of overflow makes the sloppy matcher return the document 'b x x a' as a hit.

The test shows that an exact PhraseQuery with a@0 and b@Integer.MAX_VALUE, counted on the one-document index "x a b", never returns: ExactPhraseMatcher.nextMatch computes position 1 + Integer.MAX_VALUE in int, the sum wraps, and the matcher loops forever inside the document.

Expected vs actual

Expected: count(q) == 0, returned within the 5 s @Timeout; actual: Test timeout exceeded (>= 5000 msec), and the leaked test thread is still RUNNABLE at ExactPhraseMatcher.nextMatch(ExactPhraseMatcher.java:171).

Files

• src/org/apache/lucene/search/TestPhraseQueryPositionOverflow.java (JUnit test)

Run it

• Tested on Lucene main e1faeba08bde (2026-11-02): still reproduces.
• Tested on Lucene main 645d4fe14a20 (2026-10-28): reproduced.
• Recorded outcome of run 1 on Lucene main 645d4fe14a20: 1 test(s) run, 2 failed.
• How to read the outcome: a failing test is the finding reproducing (the failure message states expected against actual).
• Seeds: a rerun with the same seed replays the same random choices. Assertions off (-da -dsa -Dtests.asserts=false) is the production-like setting.

# Lucene checkout at the tested commit, built with: ./gradlew --no-daemon assemble   (JDK 25)
LUCENE=/path/to/lucene; V=12.0.0-SNAPSHOT
CP=$LUCENE/lucene/core/build/libs/lucene-core-$V.jar
CP=$CP:$LUCENE/lucene/codecs/build/libs/lucene-codecs-$V.jar
CP=$CP:$LUCENE/lucene/test-framework/build/libs/lucene-test-framework-$V.jar
CP=$CP:randomizedtesting-runner-2.9.2.jar:junit-4.13.2.jar:hamcrest-3.0.jar   # test-framework dependencies (Maven Central)
CP=$CP:junit-platform-console-standalone-6.1.3.jar   # JUnit Platform console launcher (Maven Central)

javac -d classes -cp "$CP" $(find src -name '*.java')

# run 1: assertions on, seed DEADBEEF
java -ea -esa --add-modules jdk.incubator.vector --enable-native-access=ALL-UNNAMED -Dtests.seed=DEADBEEF \
     -cp "classes:$CP" org.junit.platform.console.ConsoleLauncher \
     execute --select-class org.apache.lucene.search.TestPhraseQueryPositionOverflow

Gradle alternative: copy the test into the matching module's src/test/ tree of the checkout and run ./gradlew -p lucene/<module> test --tests <class> -Ptests.seed=<seed> (add -Ptests.asserts=false for a run without assertions).

Feedback

Open an issue in lucene-ai-bug-hunt. Please do not open issues or pull requests in apache/lucene about this finding unless a Lucene committer agrees it should be filed.

Snapshot of 2026-11-04. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.