stanfordnlp/CoreNLP · error · RuntimeException

Problem reading similar words file!

Error message

Problem reading similar words file!

What it means

ChineseSimWordAvgDepGrammar.getMap parses a 'similar words' file into a map of similar-word triples used by the dependency grammar. Any IOException while reading that file is converted into RuntimeException 'Problem reading similar words file!', losing the filename in the message.

Solutions

  1. Confirm the similar words file exists at the expected (usually relative to working directory) location and is readable
  2. Fix file permissions or supply the file that ships with the Chinese parser model
  3. Add debugging by wrapping the read with your own IO handling to surface the actual path being opened

Example fix

// before
new ChineseSimWordAvgDepGrammar(...); // similar-words file missing
// after
// ensure the file referenced by the model (e.g. chineseSimilarWords.txt) is present in the working directory
new File(similarWordsPath).canRead() check before constructing
Defensive patterns

Strategy: try-catch

Validate before calling

if (!new File(similarWordsFile).canRead()) throw new IllegalStateException("Similar words file not readable: " + similarWordsFile);

Try / catch

try { new ChineseSimWordAvgDepGrammar(...); } catch (RuntimeException e) { if (e.getMessage().equals("Problem reading similar words file!")) { log.fatal("Check similar-words data file presence/permissions"); } }

Prevention

When it happens

Trigger: Constructing ChineseSimWordAvgDepGrammar when the similar-words data file cannot be opened or read: missing file, wrong path, permission problem, or I/O error during line parsing (readLine/regex processing loop).

Common situations: Distributing models without the required similar-words data file; running from a different working directory so the relative path breaks; the file failing mid-read due to encoding/corruption.

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/b0dabbf981d502eb. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/parser/lexparser/ChineseSimWordAvgDepGrammar.java:79

        Matcher m = linePattern.matcher(wordMapLine);
        if (!m.matches()) {
          log.info("Ill-formed line in similar word map file: " + wordMapLine);
          continue;
        }

        Pair<Integer, String> iTW = new Pair<>(wordIndex.addToIndex(m.group(1)), m.group(2));
        double score = Double.parseDouble(m.group(5));

        List<Triple<Integer, String, Double>> tripleList = hashMap.get(iTW);
        if (tripleList == null) {
          tripleList = new ArrayList<>();
          hashMap.put(iTW, tripleList);
        }

        tripleList.add(new Triple<>(wordIndex.addToIndex(m.group(3)), m.group(4), score));
      }
    } catch (IOException e) {
      throw new RuntimeException("Problem reading similar words file!");
    }

    return hashMap;
  }

  @Override
  public double scoreTB(IntDependency dependency) {
    //return op.testOptions.depWeight * Math.log(probSimilarWordAvg(dependency));
    return op.testOptions.depWeight * Math.log(probTBwithSimWords(dependency));
  }

  public void setLex(Lexicon lex) {
    this.lex = lex;
  }

  private ClassicCounter<String> statsCounter = new ClassicCounter<>();

  public void dumpSimWordAvgStats() {

View on GitHub (pinned to 1b7edd19c4)