languagetool-org/languagetool · error · IllegalArgumentException

AdvancedSynthesizerFilter: Index out of bounds in ${match.ge

Error message

AdvancedSynthesizerFilter: Index out of bounds in ${match.getRule().getFullId()}, value: ${lemmaFromStr}

What it means

Same filter, same validation, but for the 'lemmaFrom' argument: acceptRuleMatch() requires lemma_from (absolute or marker-based) to fall within 1..patternTokens.length, because the token at that index supplies the lemma for synthesis. Out of range means the rule references a nonexistent token.

Source

Thrown at languagetool-core/src/main/java/org/languagetool/rules/AbstractAdvancedSynthesizerFilter.java:90

    }
    if (postagFrom < 1 || postagFrom > patternTokens.length) {
      throw new IllegalArgumentException("AdvancedSynthesizerFilter: Index out of bounds in "
          + match.getRule().getFullId() + ", value: " + postagFromStr);
    }
    int lemmaFrom = 0;
    if (lemmaFromStr.startsWith("marker")) {
      while (lemmaFrom < patternTokens.length && patternTokens[lemmaFrom].getStartPos() < match.getFromPos()) {
        lemmaFrom++;
      }
      lemmaFrom++;
      if (lemmaFromStr.length()>6) {
        lemmaFrom += Integer.parseInt(lemmaFromStr.replace("marker", ""));
      }
    } else {
      lemmaFrom = Integer.parseInt(lemmaFromStr);
    }
    if (lemmaFrom < 1 || lemmaFrom > patternTokens.length) {
      throw new IllegalArgumentException("AdvancedSynthesizerFilter: Index out of bounds in "
          + match.getRule().getFullId() + ", value: " + lemmaFromStr);
    }

    String postagReplace = getOptional("postagReplace", arguments);

    String desiredLemma = getAnalyzedToken(patternTokens[lemmaFrom - 1], lemmaSelect).getLemma();
    String originalPostag = getAnalyzedToken(patternTokens[lemmaFrom - 1], lemmaSelect).getPOSTag();
    String desiredPostag = getAnalyzedToken(patternTokens[postagFrom - 1], postagSelect).getPOSTag();
    if (!newLemma.isEmpty()) {
      if (newLemma.startsWith("_")) {
        desiredLemma = getNewLemma(desiredLemma, newLemma);
      } else {
        desiredLemma = newLemma;
      }
    }
    if (desiredLemma == null) {
      return null;
    }

View on GitHub (pinned to 2e990059ce)

Solutions

Correct lemma_from in the rule XML to a valid 1-based token index in the pattern.
After changing a pattern, revalidate all filter argument indices.
Verify marker arithmetic stays within the matched token count.
Run the rule's test sentences to exercise acceptRuleMatch and confirm the fix.

Example fix

// before (3-token pattern, lemma_from:7)
<filter class="...AdvancedSynthesizerFilter" args="lemma_from:7 postag_from:2"/>
// after
<filter class="...AdvancedSynthesizerFilter" args="lemma_from:2 postag_from:2"/>

Defensive patterns

Strategy: validation

Validate before calling

int lemmaFrom = Integer.parseInt(args.get("lemma_from").replace("marker", ""));
if (lemmaFrom < 1 || lemmaFrom > patternTokens.length)
  throw new RuleFormatException("lemma_from out of range");

Prevention

Validate lemma_from against the actual pattern token count
Avoid copy-pasting filter args between rules with different patterns
Add rule test cases that exercise the filter

When it happens

Trigger: A rule's lemma_from value (e.g. lemma_from="3" or marker arithmetic) resolves to <1 or beyond the number of matched pattern tokens; patternTokens[lemmaFrom-1] would otherwise be accessed.

Common situations: Rule pattern edited/shortened without updating lemma_from; copy-paste of filter args between rules with different token counts; off-by-one (1-based indices).

Related errors

AI-assisted analysis of languagetool-org/languagetool@2e990059ce (2026-09-06). Data as JSON: /api/errors/1a2a14c49081dad0. Report an issue: GitHub.