stanfordnlp/CoreNLP · error · RuntimeException

Invalid match group for entry

Error message

Invalid match group for entry ${entry}

What it means

During createPatternMatcher, each mapping entry's declared 'group' column is validated against the compiled TokenSequencePattern's total group count; if annotateGroup < 0 or exceeds pattern.getTotalGroups(), a RuntimeException naming the entry is thrown. It means the mapping file asks to annotate a capture group the pattern never defines.

Solutions

  1. Reduce the 'group' column value on the offending entry to a group that exists in the pattern (0 or 1 for single-group patterns).
  2. Add a capture group to the pattern so the requested group index exists.
  3. Remove the 'group' column from the row to fall back to the default group 0.
  4. Re-run and read the entry name in the message to locate the exact mapping-file line.

Example fix

// mapping file before (pattern has no group 2)
[ { word:/CEO/ } ]	group=2
// after
[ { word:/CEO/ } ]	group=0
Defensive patterns

Strategy: validation

Validate before calling

// Before loading: verify each group value fits its pattern's group count
int groups = countCaptureGroups(patternText); // e.g. count unescaped '(' opening groups
if (groupValue < 0 || groupValue > groups) throw new IllegalArgumentException("group " + groupValue + " invalid for pattern: " + patternText);

Try / catch

try { annotator = new TokensRegexNERAnnotator(name, props); } catch (RuntimeException e) { if (e.getMessage().startsWith("Invalid match group")) { fixMappingFile(e.getMessage()); } else throw e; }

Prevention

When it happens

Trigger: A mapping file row has a 'group' value of e.g. 2 while its pattern contains only one parenthesized group; or a negative group value is written in the group column.

Common situations: Copying a pattern from another entry and forgetting to update the group column; editing a pattern to remove a capture group while leaving group=2; hand-written regexes in TokensRegex mapping files where counting groups off by one.

Understand the failure class

Background: "value must be between 0 and 1" / "out of range" / "must not be negative" errors: fixing range-validation failures across open-source libraries — this error's family across 42 libraries.

Related errors

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/cfa3fc85073f59cb. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/pipeline/TokensRegexNERAnnotator.java:391

      if (entry.tokensRegex != null) {
        // TODO: posTagPatterns...
        pattern = TokenSequencePattern.compile(env, entry.tokensRegex);
      } else {
        List<SequencePattern.PatternExpr> nodePatterns = new ArrayList<>(entry.regex.length);
        for (String p:entry.regex) {
          CoreMapNodePattern c = CoreMapNodePattern.valueOf(p, patternFlags);
          if (posTagPattern != null) {
            c.add(CoreAnnotations.PartOfSpeechAnnotation.class, posTagPattern);
          }
          nodePatterns.add(new SequencePattern.NodePatternExpr(c));
        }
        if (nodePatterns.size() == 1) {
          nodePatterns = Collections.singletonList(nodePatterns.get(0));
        }
        pattern = TokenSequencePattern.compile(new SequencePattern.SequencePatternExpr(nodePatterns));
      }
      if (entry.annotateGroup < 0 || entry.annotateGroup > pattern.getTotalGroups()) {
        throw new RuntimeException("Invalid match group for entry " + entry);
      }
      pattern.setPriority(entry.priority);
      pattern.setWeight(entry.weight);
      patterns.add(pattern);
      patternToEntry.put(pattern, entry);
    }
    return TokenSequencePattern.getMultiPatternMatcher(patterns);
  }

  private void annotateMatched(List<CoreLabel> tokens) {
    List<SequenceMatchResult<CoreMap>> matched = multiPatternMatcher.findNonOverlapping(tokens);
    for (SequenceMatchResult<CoreMap> m:matched) {
      Entry entry = patternToEntry.get(m.pattern());

      // Check if we will overwrite the existing annotation with this annotation
      int g = entry.annotateGroup;
      int start = m.start(g);
      int end = m.end(g);

View on GitHub (pinned to 1b7edd19c4)