예제 #1
0
  @Override
  public void write(
      Directory directory,
      SegmentInfo segmentInfo,
      String segmentSuffix,
      FieldInfos infos,
      IOContext context)
      throws IOException {
    final String fileName =
        IndexFileNames.segmentFileName(
            segmentInfo.name, segmentSuffix, Lucene46FieldInfosFormat.EXTENSION);
    try (IndexOutput output = directory.createOutput(fileName, context)) {
      CodecUtil.writeHeader(
          output, Lucene46FieldInfosFormat.CODEC_NAME, Lucene46FieldInfosFormat.FORMAT_CURRENT);
      output.writeVInt(infos.size());
      for (FieldInfo fi : infos) {
        IndexOptions indexOptions = fi.getIndexOptions();
        byte bits = 0x0;
        if (fi.hasVectors()) bits |= Lucene46FieldInfosFormat.STORE_TERMVECTOR;
        if (fi.omitsNorms()) bits |= Lucene46FieldInfosFormat.OMIT_NORMS;
        if (fi.hasPayloads()) bits |= Lucene46FieldInfosFormat.STORE_PAYLOADS;
        if (fi.getIndexOptions() != IndexOptions.NONE) {
          bits |= Lucene46FieldInfosFormat.IS_INDEXED;
          assert indexOptions.compareTo(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS) >= 0
              || !fi.hasPayloads();
          if (indexOptions == IndexOptions.DOCS) {
            bits |= Lucene46FieldInfosFormat.OMIT_TERM_FREQ_AND_POSITIONS;
          } else if (indexOptions == IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS) {
            bits |= Lucene46FieldInfosFormat.STORE_OFFSETS_IN_POSTINGS;
          } else if (indexOptions == IndexOptions.DOCS_AND_FREQS) {
            bits |= Lucene46FieldInfosFormat.OMIT_POSITIONS;
          }
        }
        output.writeString(fi.name);
        output.writeVInt(fi.number);
        output.writeByte(bits);

        // pack the DV types in one byte
        final byte dv = docValuesByte(fi.getDocValuesType());
        final byte nrm = docValuesByte(fi.hasNorms() ? DocValuesType.NUMERIC : DocValuesType.NONE);
        assert (dv & (~0xF)) == 0 && (nrm & (~0x0F)) == 0;
        byte val = (byte) (0xff & ((nrm << 4) | dv));
        output.writeByte(val);
        output.writeLong(fi.getDocValuesGen());
        output.writeStringStringMap(fi.attributes());
      }
      CodecUtil.writeFooter(output);
    }
  }
 public void testFieldNames() throws Exception {
   Directory dir1 = getDir1(random());
   Directory dir2 = getDir2(random());
   ParallelLeafReader pr =
       new ParallelLeafReader(
           SlowCompositeReaderWrapper.wrap(DirectoryReader.open(dir1)),
           SlowCompositeReaderWrapper.wrap(DirectoryReader.open(dir2)));
   FieldInfos fieldInfos = pr.getFieldInfos();
   assertEquals(4, fieldInfos.size());
   assertNotNull(fieldInfos.fieldInfo("f1"));
   assertNotNull(fieldInfos.fieldInfo("f2"));
   assertNotNull(fieldInfos.fieldInfo("f3"));
   assertNotNull(fieldInfos.fieldInfo("f4"));
   pr.close();
   dir1.close();
   dir2.close();
 }
  void createDocumentNode(final DocumentDescriptor inDescriptor) throws IOException {
    try {
      _document = DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument();
      _rootNode = _document.createElement("document");
    } catch (ParserConfigurationException e) {
      e.printStackTrace();
      System.exit(1);
    }
    AtomicReader segmentReader = _indexReader.leaves().get(inDescriptor.segmentNumber).reader();
    _rootNode.setAttribute("id", DocumentIdOperations.documentDescriptorToId(inDescriptor));
    // TODO: implement the proper way of building a title from the production report
    _rootNode.setAttribute("title", buildDocumentTitle(segmentReader, inDescriptor));
    _rootNode.setAttribute("path", "ruscorpora.ru");
    _rootNode.setAttribute(
        "tagging", Attributes.getDocAttribute(segmentReader, inDescriptor.localId, "tagging"));
    _rootNode.setAttribute("snippets", "0");
    Element attributesNode = _document.createElement("attributes");

    _rootNode.appendChild(attributesNode);

    FieldInfos fields = segmentReader.getFieldInfos();
    for (int fieldIndex = 0; fieldIndex != fields.size(); ++fieldIndex) {
      FieldInfo field = fields.fieldInfo(fieldIndex);
      // TODO: understand why field may turn into null
      if (field == null) {
        continue;
      }
      String name = field.name;
      if (Attributes.ATTRIBUTES.contains(name)
          || Attributes.ATTRIBUTES_FOR_REPORT.contains(name)
          || Attributes.ATTRIBUTES_FOR_WORD_INFO.contains(name)
          || !field.hasDocValues()) {
        // it's a word attribute
        continue;
      }
      Element attrNode = _document.createElement("attr");
      attrNode.setAttribute("name", name);
      attrNode.setAttribute(
          "value", Attributes.getDocAttribute(segmentReader, inDescriptor.localId, name));
      attributesNode.appendChild(attrNode);
    }
  }