What causes "Invalid null annotation key" in stanfordnlp/CoreNLP?

Audit the annotationKeys list passed to setAnnotations and remove/replace null entries

stanfordnlp/CoreNLP · error · java.lang.RuntimeException

Invalid null annotation key

Error message

Invalid null annotation key

What it means

MatchedExpression.setAnnotations maps a list of values onto a parallel list of annotation keys (Classes). When it encounters a null entry in annotationKeys it cannot store the value, so it throws 'Invalid null annotation key'. A null key means the caller supplied a malformed key list.

Solutions

Audit the annotationKeys list passed to setAnnotations and remove/replace null entries
Ensure the key lookup (e.g. CoreAnnotations subclass constant) is not null before calling
Guard with a null-check loop before invoking setAnnotations

Example fix

// before
annotationKeys.set(1, null);
expr.setAnnotations(...);
// after
if (annotationKeys.contains(null)) throw new IllegalStateException("null annotation key");
expr.setAnnotations(...);

Defensive patterns

Strategy: validation

Validate before calling

if (annotationKeys == null || annotationKeys.stream().anyMatch(Objects::isNull))
    throw new IllegalArgumentException("annotationKeys must not contain null");

Type guard

static boolean validKeys(List<Class<?>> keys) { return keys != null && keys.stream().allMatch(Objects::nonNull); }

Try / catch

try { expr.setAnnotations(cm, values, annotationKeys, ...); } catch (RuntimeException e) { if (e.getMessage().contains("null annotation key")) { /* fix key list */ } else throw e; }

Prevention

Verify Class lookups for annotation keys never return null
Use static CoreAnnotations class constants instead of reflection
Assert key list non-null before annotate calls

When it happens

Trigger: Calling annotate/setAnnotations with an annotationKeys list that contains a null element while the value object is a List whose size covers that index.

Common situations: Programmatically building annotation key arrays where a Class lookup returned null (e.g. reflection miss or uninitialized key constant).

Related errors

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/e4ec3e669a0d2428. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/ling/tokensregex/MatchedExpression.java:84

    public Function<MatchedExpression, Value> expressionToValue;
    public Function<MatchedExpression,?> resultAnnotationExtractor;
    public CoreMapAggregator tokensAggregator;

    @Override
    public Value apply(CoreMap in) {
      return valueExtractor.apply(in);
    }

    private static void setAnnotations(CoreMap cm, List<Class> annotationKeys, Object obj) {
      if (annotationKeys.size() > 1 && obj instanceof List) {
        // List of annotationKeys, obj also list, we should try to match the objects to annotationKeys
        List list = (List) obj;
        int n = Math.min(list.size(), annotationKeys.size());
        for (int i = 0; i < n; i++) {
          Object v = list.get(i);
          Class key = annotationKeys.get(i);
          if (key == null) {
            throw new RuntimeException("Invalid null annotation key");
          }
          if (v instanceof Value) {
            cm.set(key, ((Value) v).get());
          } else {
            cm.set(key, v);
          }
        }
      } else {
        // Only a single object, set all annotationKeys to that obj
        for (Class key:annotationKeys) {
          if (key == null) {
            throw new RuntimeException("Invalid null annotation key");
          }
          cm.set(key, obj);
        }
      }
    }

View on GitHub (pinned to 1b7edd19c4)