ICUTokenizer loses a supplementary character (an emoji, a CJK Extension B ideograph) whose surrogate pair straddles the end of its 4096-char read chunk in text with no whitespace, because refill() cuts the pair in half

Finding LFV-445Area: analysis / ICU (ICUTokenizer chunked buffer refill; ICUTokenizerFactory 'icu')Priority: lowImpact: silent wrong resultsNovelty: new

Finding page for LFV-445

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

What it shows

You index a long Chinese paragraph with no spaces or line breaks through the stock ICUTokenizer, and it contains an emoji or a rare ideograph such as U+20000; you expect that character to be a token wherever it falls, as it is one position earlier or later. Instead, when its first UTF-16 half is the last char of the tokenizer's 4096-char read chunk, refill() gives the word breaker a chunk that ends in a lone high surrogate and starts the next chunk with the lone low surrogate; both halves get rule status 0 and are skipped, so the character is in no token. A TermQuery for it finds 0 documents instead of 1, with no error.

The test shows that ICUTokenizer drops U+20000 when its high surrogate is the last char (char 4095) of a whitespace-free 4096-char read chunk, because refill() splits the surrogate pair.

Expected vs actual

Expected U+20000 to be 1 token (as at any other position); actual 0 tokens.

Files

src/org/apache/lucene/analysis/icu/segmentation/TestICUTokenizerRefillSurrogatePair.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=$CP:$LUCENE/lucene/analysis/common/build/libs/lucene-analysis-common-$V.jar
CP=$CP:$LUCENE/lucene/analysis/icu/build/libs/lucene-analysis-icu-$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.analysis.icu.segmentation.TestICUTokenizerRefillSurrogatePair

The module(s) analysis-icu also need their third-party runtime jars; ./gradlew -p lucene/<module> dependencies lists them.

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.