StableTflSimilarity scores SynonymQuery and CombinedFieldQuery with the length of a placeholder term string ('synonym pseudo-term', 19 code points, or 'pseudo_term', 11), not the length of the query terms

Finding LFV-261Area: search / similarities: StableTflSimilarity with SynonymQuery and CombinedFieldQueryPriority: mediumImpact: silent wrong resultsNovelty: new

Finding page for LFV-261

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

What it shows

You set StableTflSimilarity on the searcher, whose javadoc derives a term's rarity from the length |t| of the query term, and you search with a SynonymQuery (for example the one QueryBuilder builds when a synonym filter stacks 'dog' and 'dogs' at one position) or with a CombinedFieldQuery. You expect the real query terms' length to be used, so that SynonymQuery{ab, cd} on a document that contains 'ab' once scores like TermQuery(ab), and a one-field CombinedFieldQuery scores like a TermQuery on a copy field. Instead both queries hand the similarity a fixed placeholder term ('synonym pseudo-term', 19 code points; 'pseudo_term', 11 code points), so every synonym group is scored as an almost maximally rare 19-character term and every combined-field term as an 11-character one: SynonymQuery{ab, cd} scores 1.4050181 instead of 1.3490043, and in 200-token documents the group {dog, dogs} scores 2.13x TermQuery(dog) and outranks a document that matches 'elephant'.

The test shows that with StableTflSimilarity, SynonymQuery passes the placeholder term "synonym pseudo-term" (19 code points) to the similarity instead of its real terms, so a synonym group gets the rarity of a 19-character term.

Expected vs actual

Expected: SynonymQuery{ab, cd} on a document with the single token "ab" scores like TermQuery(ab), 1.3855287. Actual (-ea and -da): 1.4050188.

Files

• src/org/apache/lucene/search/TestStableTflSynonymQuery.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, 1 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.TestStableTflSynonymQuery

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 2026-11-03. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.