stanfordnlp/CoreNLP · error · NoSuchElementException

DocumentIterator exhausted.

Error message

DocumentIterator exhausted.

What it means

CoNLLDocumentReader.DocumentIterator.next() implements java.util.Iterator semantics: when the internal lookahead document (nextDoc) is null because all underlying files/lines have been consumed, it throws NoSuchElementException instead of returning null. Callers are expected to check hasNext() before calling next(); this throw signals the iterator was advanced past its end.

Solutions

1. Guard every next() call with while (it.hasNext()) { ... }
2. If you need a reusable pass, create a new DocumentIterator from the reader instead of reusing the exhausted one
3. If collecting all documents, use a for-each loop over the Iterable rather than manual next() calls

Example fix

// before
while (true) {
  CoNLLDocument doc = docIter.next(); // throws at end
  process(doc);
}
// after
while (docIter.hasNext()) {
  CoNLLDocument doc = docIter.next();
  process(doc);
}

if (docIter.hasNext()) { CoNLLDocument d = docIter.next(); }

function nextOrNull(Iterator<CoNLLDocument> it) { return it.hasNext() ? it.next() : null; }

try { doc = docIter.next(); } catch (NoSuchElementException e) { doc = null; // treat as end of iteration }

When it happens

Trigger: Calling next() on the reader's DocumentIterator after hasNext() returned false, or calling next() more times than there are documents without checking hasNext().

Common situations: Manual iteration loops over CoNLL corpus files that use a do-while or fixed-count loop instead of while(hasNext()); reusing an exhausted iterator after a first pass; off-by-one when counting documents in the CoNLL-2012 corpus.

Understand the failure class

Background: EmptyResultError / "no results found": when an API or scraper succeeds but returns zero rows — this error's family across 9 libraries.

Related errors

• Error running hybrid coref system
• No more elements from
• DocumentIterator exhausted.
• Coreference is not implemented for Spanish
• Processed trees, but there are more trees and text is empty

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/ee1fc979c5a5e38c. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/coref/docreader/CoNLLDocumentReader.java:342

    int lineCnt = 0;
    int docCnt = 0;

    public DocumentIterator(String filename, Options options) throws IOException {
      this.options = options;
      this.filename = filename;
      this.br = IOUtils.readerFromString(filename);
      nextDoc = readNextDocument();
    }

    @Override
    public boolean hasNext() {
      return nextDoc != null;
    }

    @Override
    public CoNLLDocument next() {
      if (nextDoc == null) {
        throw new NoSuchElementException("DocumentIterator exhausted.");
      }
      CoNLLDocument curDoc = nextDoc;
      nextDoc = readNextDocument();
      return curDoc;
    }

    private static final Pattern starPattern = Pattern.compile("\\*");

    private static Tree wordsToParse(List<String[]> sentWords) {
      StringBuilder sb = new StringBuilder();
      for (String[] fields:sentWords) {
        if (sb.length() > 0) {
          sb.append(' ');
        }

        String str = fields[FIELD_PARSE_BIT].replace("NOPARSE", "X");
        String tagword = "(" + fields[FIELD_POS_TAG] + " " + fields[FIELD_WORD] + ")";
        // Replace stars

View on GitHub (pinned to 1b7edd19c4)