public NameSample read()

in opennlp-tools/src/main/java/opennlp/tools/formats/BioNLP2004NameSampleStream.java [90:186]
72 lines of code
25 McCabe index (conditional complexity)

  public NameSample read() throws IOException {

    List<String> sentence = new ArrayList<>();
    List<String> tags = new ArrayList<>();

    boolean isClearAdaptiveData = false;

    // Empty line indicates end of sentence
    String line;
    while ((line = lineStream.read()) != null && !StringUtil.isEmpty(line.trim())) {

      if (line.startsWith("###MEDLINE:")) {
        isClearAdaptiveData = true;
        lineStream.read();
        continue;
      }

      if (line.contains("ABSTRACT TRUNCATED"))
        continue;

      String[] fields = line.split("\t");

      if (fields.length == 2) {
        sentence.add(fields[0]);
        tags.add(fields[1]);
      }
      else {
        throw new IOException("Expected two fields per line in training data, got " +
            fields.length + " for line '" + line + "'!");
      }
    }

    if (!sentence.isEmpty()) {

      // convert name tags into spans
      List<Span> names = new ArrayList<>();

      int beginIndex = -1;
      int endIndex = -1;
      for (int i = 0; i < tags.size(); i++) {

        String tag = tags.get(i);

        if (tag.endsWith("DNA") && (types & GENERATE_DNA_ENTITIES) == 0)
          tag = CODEC_TAG_O;

        if (tag.endsWith("protein") && (types & GENERATE_PROTEIN_ENTITIES) == 0)
          tag = CODEC_TAG_O;

        if (tag.endsWith("cell_type") && (types & GENERATE_CELLTYPE_ENTITIES) == 0)
          tag = CODEC_TAG_O;

        if (tag.endsWith("cell_line") && (types & GENERATE_CELLTYPE_ENTITIES) == 0)
          tag = CODEC_TAG_O;
        if (tag.endsWith("RNA") && (types & GENERATE_RNA_ENTITIES) == 0)
          tag = CODEC_TAG_O;

        if (tag.startsWith(CODEC_TAG_B)) {

          if (beginIndex != -1) {
            names.add(new Span(beginIndex, endIndex, tags.get(beginIndex).substring(2)));
          }

          beginIndex = i;
          endIndex = i + 1;
        }
        else if (tag.startsWith(CODEC_TAG_I)) {
          endIndex++;
        }
        else if (tag.equals(CODEC_TAG_O)) {
          if (beginIndex != -1) {
            names.add(new Span(beginIndex, endIndex, tags.get(beginIndex).substring(2)));
            beginIndex = -1;
            endIndex = -1;
          }
        }
        else {
          throw new IOException("Invalid tag: " + tag);
        }
      }

      // if one span remains, create it here
      if (beginIndex != -1)
        names.add(new Span(beginIndex, endIndex, tags.get(beginIndex).substring(2)));

      return new NameSample(sentence.toArray(new String[0]),
          names.toArray(new Span[0]), isClearAdaptiveData);
    }
    else if (line != null) {
      // Just filter out empty events, if two lines in a row are empty
      return read();
    }
    else {
      // source stream is not returning anymore lines
      return null;
    }
  }