stanfordnlp/CoreNLP · critical · RuntimeException

CANNOT EVEN CREATE ARRAYS OF ORIGINAL SIZE!!!

Error message

CANNOT EVEN CREATE ARRAYS OF ORIGINAL SIZE!!! 

What it means

After a failed attempt to grow chart arrays for a longer sentence, ExhaustiveDependencyParser.parse tries to recreate arrays at the previous working size (arraySize). If even that allocation fails with OutOfMemoryError, it throws RuntimeException 'CANNOT EVEN CREATE ARRAYS OF ORIGINAL SIZE!!! <arraySize>' — the JVM heap is too exhausted to restore the parser's prior state.

Solutions

1. Increase JVM heap (-Xmx), e.g. -Xmx4g or more depending on grammar size and maxLength
2. Reduce testOptions.maxLength to keep chart arrays smaller
3. Run parsing in a fresh JVM or ensure prior parses/objects are garbage-collectable

Example fix

// before
java -Xmx1g ... Parser -maxLength 100 ...
// after
java -Xmx8g ... Parser -maxLength 60 ...

long maxHeap = Runtime.getRuntime().maxMemory(); if (maxHeap < requiredHeapForGrammar) throw new IllegalStateException("Increase -Xmx; heap " + maxHeap + " too small for parser chart");

try { parser.parse(sentence); } catch (RuntimeException e) { if (e.getMessage().startsWith("CANNOT EVEN CREATE ARRAYS")) { reduceMaxLengthAndRetryOrRestartJvm(); } }

When it happens

Trigger: parse() called with a long sentence; growing arrays OOMed; the fallback createArrays(arraySize) also threw OutOfMemoryError — i.e., heap is nearly full at the original chart size.

Common situations: Parsing many long sentences in one JVM without releasing memory; -Xmx too small for the grammar/chart; memory leaks from retaining previous parses.

Related errors

• Refusal to create such large arrays.
• CANNOT EVEN CREATE ARRAYS OF ORIGINAL SIZE!!
• Parsing of sentence failed, possibly because of out of…
• Attempt to create ChineseSimWordAvgDepGrammar before…
• Attribute already defined:

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/74149c935a798f77. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/parser/lexparser/ExhaustiveDependencyParser.java:189

  public boolean parse(List<? extends HasWord> sentence) {
    if (op.testOptions.verbose) {
      Timing.tick("Starting dependency parse.");
    }
    this.sentence = sentence;
    int length = sentence.size();
    if (length > arraySize) {
      if (length > op.testOptions.maxLength + 1 || length >= myMaxLength) {
        throw new OutOfMemoryError("Refusal to create such large arrays.");
      } else {
        try {
          createArrays(length + 1);
        } catch (OutOfMemoryError e) {
          myMaxLength = length;
          if (arraySize > 0) {
            try {
              createArrays(arraySize);
            } catch (OutOfMemoryError e2) {
              throw new RuntimeException("CANNOT EVEN CREATE ARRAYS OF ORIGINAL SIZE!!! " + arraySize);
            }
          }
          throw e;
        }
        arraySize = length + 1;
        if (op.testOptions.verbose) {
          log.info("Created dparser arrays of size " + arraySize);
        }
      }
    }
    if (op.testOptions.verbose) {
      log.info("Initializing...");
    }

    // map to words
    words = new int[length];
    int numTags = dg.numTagBins();//tagIndex.size();
    //System.out.println("\nNumTags: "+numTags);

View on GitHub (pinned to 1b7edd19c4)