stanfordnlp/CoreNLP · error · SemgrexParseException

Duplicate attribute found in semgrex expression

Error message

Duplicate attribute  found in semgrex expression

What it means

NodePattern.setAttribute (used by NodeAttributes) tracks positive (non-negated) attribute keys; setting the same key twice in one semgrex node is ambiguous, so it throws SemgrexParseException. Negated attributes are exempt from the duplicate check.

Solutions

1. Remove the duplicate attribute from the node expression, merging values if possible
2. Use a coordination/disanjunctive form if two values for one key were intended
3. Deduplicate keys programmatically before calling AddAttribute/setAttribute

Example fix

// before
node.addAttribute("word", "foo", false); node.addAttribute("word", "bar", false);
// after
node.addAttribute("word", "foo", false); node.addAttribute("tag", "bar", false);

Set<String> seen = new HashSet<>(); for (Triple<String,String,Boolean> a : attrs) if (!a.third() && !seen.add(a.first())) throw new IllegalArgumentException("Duplicate attribute " + a.first());

try { nodeAttributes.setAttribute(key, value, negated); } catch (SemgrexParseException e) { log.error(e.getMessage()); }

When it happens

Trigger: Calling setAttribute twice with the same key and negated=false on the same node's attribute builder, or writing a semgrex node with a repeated positive attribute (e.g. '[word foo word bar]').

Common situations: Hand-written semgrex patterns repeating an attribute, generated patterns from templates where a key is substituted twice, or programmatic AddAttribute calls in a loop without deduplication.

Understand the failure class

Background: Conflicting config options: "cannot be used together" — configuration validation errors across open-source libraries — this error's family across 162 libraries.

Related errors

• Semgrex pattern asked for uniq of node which does not…
• We need at least 2 extractors for ExtractorMerger to make…
• Too many columns: / (offset: )
• Too few columns: / (offset: )
• Sentence.toSentence: lengths differ

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/446b93c490a22d7c. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/semgraph/semgrex/NodeAttributes.java:65

    this.root = root;
  }

  public boolean root() {
    return root;
  }

  public void setEmpty(boolean empty) {
    this.empty = empty;
  }

  public boolean empty() {
    return empty;
  }

  public void setAttribute(String key, String value, boolean negated) {
    if (!negated) {
      if (positiveAttributes.contains(key)) {
        throw new SemgrexParseException("Duplicate attribute " + key + " found in semgrex expression");
      }
      positiveAttributes.add(key);
    }
    attributes.add(new Triple(key, value, negated));
  }

  public void addContains(String annotation, String key, String value, Boolean negated) {
    contains.add(new Quadruple(annotation, key, value, negated));
  }

  public List<Triple<String, String, Boolean>> attributes() {
    return Collections.unmodifiableList(attributes);
  }

  public List<Quadruple<String, String, String, Boolean>> contains() {
    return Collections.unmodifiableList(contains);
  }
}

View on GitHub (pinned to 1b7edd19c4)