What causes "Error creating ChineseMaxentLexicon{e}" in stanfordnlp/CoreNLP?

Verify the feature directory exists and contains .gb files

stanfordnlp/CoreNLP · error · RuntimeException

Error creating ChineseMaxentLexicon

Error message

Error creating ChineseMaxentLexicon{e}

What it means

ChineseMorphFeatureSets' constructor scans a feature directory for .gb files and calls getFeatures on each; an IOException while reading any file is rethrown as a RuntimeException 'Error creating ChineseMaxentLexicon...'. It signals the morphological feature data could not be loaded from disk.

Solutions

Verify the feature directory exists and contains .gb files
Check the wrapped IOException message for the exact failing file
Fix file permissions or copy the .gb feature files into the expected location
Use the correct path to the resources bundled with the model distribution

Example fix

// before
new ChineseMorphFeatureSets("/models/chinese-features"); // dir missing
// after
File dir = new File("/models/chinese-features");
if (!dir.isDirectory() || dir.listFiles((d,n)->n.endsWith(".gb")) == null)
  throw new IllegalArgumentException("Feature dir missing or no .gb files: " + dir);
new ChineseMorphFeatureSets(dir.getPath());

Defensive patterns

Strategy: try-catch

Validate before calling

File dir = new File(featureDir);
if (!dir.isDirectory()) throw new IllegalStateException("Missing feature dir: " + featureDir);
File[] gb = dir.listFiles((d, n) -> n.endsWith(".gb"));
if (gb == null || gb.length == 0) throw new IllegalStateException("No .gb feature files in " + featureDir);

Try / catch

try { fs = new ChineseMorphFeatureSets(featureDir); }
catch (RuntimeException e) { log.error("Feature load failed: {}", e.getMessage(), e); throw new IllegalStateException("Fix chinese feature data path", e); }

Prevention

Package .gb feature files with the model and resolve paths from the classpath
Check directory readability at startup
Pin the exact feature-data layout the library expects

When it happens

Trigger: Constructing ChineseMorphFeatureSets with a featureDir that does not exist, is unreadable, or whose .gb files fail I/O while being read (also note listFiles may return null and cause an NPE in the same try block).

Common situations: Wrong path to the Chinese morphological feature dictionary files; data files missing from deployment (not in classpath/resources); permission problems reading the directory.

Understand the failure class

Background: "failed to read file", EACCES, ENOENT and "could not read <path>" errors: when a program can't read a file from disk — this error's family across 49 libraries.

Related errors

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/32965baa5188ac60. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/ie/ChineseMorphFeatureSets.java:43

  private Map<String, Pair<Set<Character>, Set<Character>>> affixFeatures = Generics.newHashMap();

  public Map<String, Set<Character>> getSingletonFeatures() {
    return singletonFeatures;
  }

  public Map<String, Pair<Set<Character>, Set<Character>>> getAffixFeatures() {
    return affixFeatures;
  }

  public ChineseMorphFeatureSets(String featureDir) {
    try {
      File dir = new File(featureDir);
      File[] files = dir.listFiles((dir1, name) -> name.endsWith(".gb"));
      for (File file : files) {
        getFeatures(file);
      }
    } catch (IOException e) {
      throw new RuntimeException("Error creating ChineseMaxentLexicon" + e);
    }
  }

  private enum FeatType {
    PREFIX, SUFFIX, SINGLETON
  }

  private void getFeatures(File file) throws IOException {
    BufferedReader in = new BufferedReader(new InputStreamReader(new FileInputStream(file), "GB18030"));
    String filename = file.getName();
    String singleFeatName = filename;
    if (singleFeatName.indexOf('.') >= 0) {
      singleFeatName = singleFeatName.substring(0, filename.lastIndexOf('.'));
    }

    FeatType featType = null;
    for (FeatType ft : FeatType.values()) {
      if (filename.contains(ft.toString().toLowerCase())) {

View on GitHub (pinned to 1b7edd19c4)