termvectorsreader.java

lucene-2.4.0 是一个全文收索的工具包

package org.apache.lucene.index;

/**
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

import org.apache.lucene.store.BufferedIndexInput;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.IndexInput;

import java.io.IOException;
import java.util.Arrays;

/**
 * @version $Id: TermVectorsReader.java 687046 2008-08-19 13:01:11Z mikemccand $
 */
class TermVectorsReader implements Cloneable {

  // NOTE: if you make a new format, it must be larger than
  // the current format
  static final int FORMAT_VERSION = 2;

  // Changes to speed up bulk merging of term vectors:
  static final int FORMAT_VERSION2 = 3;

  // Changed strings to UTF8 with length-in-bytes not length-in-chars
  static final int FORMAT_UTF8_LENGTH_IN_BYTES = 4;

  // NOTE: always change this if you switch to a new format!
  static final int FORMAT_CURRENT = FORMAT_UTF8_LENGTH_IN_BYTES;

  //The size in bytes that the FORMAT_VERSION will take up at the beginning of each file 
  static final int FORMAT_SIZE = 4;

  static final byte STORE_POSITIONS_WITH_TERMVECTOR = 0x1;
  static final byte STORE_OFFSET_WITH_TERMVECTOR = 0x2;
  
  private FieldInfos fieldInfos;

  private IndexInput tvx;
  private IndexInput tvd;
  private IndexInput tvf;
  private int size;
  private int numTotalDocs;

  // The docID offset where our docs begin in the index
  // file.  This will be 0 if we have our own private file.
  private int docStoreOffset;
  
  private final int format;

  TermVectorsReader(Directory d, String segment, FieldInfos fieldInfos)
    throws CorruptIndexException, IOException {
    this(d, segment, fieldInfos, BufferedIndexInput.BUFFER_SIZE);
  }

  TermVectorsReader(Directory d, String segment, FieldInfos fieldInfos, int readBufferSize)
    throws CorruptIndexException, IOException {
    this(d, segment, fieldInfos, readBufferSize, -1, 0);
  }
    
  TermVectorsReader(Directory d, String segment, FieldInfos fieldInfos, int readBufferSize, int docStoreOffset, int size)
    throws CorruptIndexException, IOException {
    boolean success = false;

    try {
      if (d.fileExists(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION)) {
        tvx = d.openInput(segment + "." + IndexFileNames.VECTORS_INDEX_EXTENSION, readBufferSize);
        format = checkValidFormat(tvx);
        tvd = d.openInput(segment + "." + IndexFileNames.VECTORS_DOCUMENTS_EXTENSION, readBufferSize);
        final int tvdFormat = checkValidFormat(tvd);
        tvf = d.openInput(segment + "." + IndexFileNames.VECTORS_FIELDS_EXTENSION, readBufferSize);
        final int tvfFormat = checkValidFormat(tvf);

        assert format == tvdFormat;
        assert format == tvfFormat;

        if (format >= FORMAT_VERSION2) {
          assert (tvx.length()-FORMAT_SIZE) % 16 == 0;
          numTotalDocs = (int) (tvx.length() >> 4);
        } else {
          assert (tvx.length()-FORMAT_SIZE) % 8 == 0;
          numTotalDocs = (int) (tvx.length() >> 3);
        }

        if (-1 == docStoreOffset) {
          this.docStoreOffset = 0;
          this.size = numTotalDocs;
          assert size == 0 || numTotalDocs == size;
        } else {
          this.docStoreOffset = docStoreOffset;
          this.size = size;
          // Verify the file is long enough to hold all of our
          // docs
          assert numTotalDocs >= size + docStoreOffset: "numTotalDocs=" + numTotalDocs + " size=" + size + " docStoreOffset=" + docStoreOffset;
        }
      } else
        format = 0;

      this.fieldInfos = fieldInfos;
      success = true;
    } finally {
      // With lock-less commits, it's entirely possible (and
      // fine) to hit a FileNotFound exception above. In
      // this case, we want to explicitly close any subset
      // of things that were opened so that we don't have to
      // wait for a GC to do so.
      if (!success) {
        close();
      }
    }
  }

  // Used for bulk copy when merging
  IndexInput getTvdStream() {
    return tvd;
  }

  // Used for bulk copy when merging
  IndexInput getTvfStream() {
    return tvf;
  }

  final private void seekTvx(final int docNum) throws IOException {
    if (format < FORMAT_VERSION2)
      tvx.seek((docNum + docStoreOffset) * 8L + FORMAT_SIZE);
    else
      tvx.seek((docNum + docStoreOffset) * 16L + FORMAT_SIZE);
  }

  boolean canReadRawDocs() {
    return format >= FORMAT_UTF8_LENGTH_IN_BYTES;
  }

  /** Retrieve the length (in bytes) of the tvd and tvf
   *  entries for the next numDocs starting with
   *  startDocID.  This is used for bulk copying when
   *  merging segments, if the field numbers are
   *  congruent.  Once this returns, the tvf & tvd streams
   *  are seeked to the startDocID. */
  final void rawDocs(int[] tvdLengths, int[] tvfLengths, int startDocID, int numDocs) throws IOException {

    if (tvx == null) {
      Arrays.fill(tvdLengths, 0);
      Arrays.fill(tvfLengths, 0);
      return;
    }

    // SegmentMerger calls canReadRawDocs() first and should
    // not call us if that returns false.
    if (format < FORMAT_VERSION2)
      throw new IllegalStateException("cannot read raw docs with older term vector formats");

    seekTvx(startDocID);

    long tvdPosition = tvx.readLong();
    tvd.seek(tvdPosition);

    long tvfPosition = tvx.readLong();
    tvf.seek(tvfPosition);

    long lastTvdPosition = tvdPosition;
    long lastTvfPosition = tvfPosition;

    int count = 0;
    while (count < numDocs) {
      final int docID = docStoreOffset + startDocID + count + 1;
      assert docID <= numTotalDocs;
      if (docID < numTotalDocs)  {
        tvdPosition = tvx.readLong();
        tvfPosition = tvx.readLong();
      } else {
        tvdPosition = tvd.length();
        tvfPosition = tvf.length();
        assert count == numDocs-1;
      }
      tvdLengths[count] = (int) (tvdPosition-lastTvdPosition);
      tvfLengths[count] = (int) (tvfPosition-lastTvfPosition);
      count++;
      lastTvdPosition = tvdPosition;
      lastTvfPosition = tvfPosition;
    }
  }

  private int checkValidFormat(IndexInput in) throws CorruptIndexException, IOException
  {
    int format = in.readInt();
    if (format > FORMAT_CURRENT) {
      throw new CorruptIndexException("Incompatible format version: " + format + " expected " 
                                      + FORMAT_CURRENT + " or less");
    }
    return format;
  }

  void close() throws IOException {
    // make all effort to close up. Keep the first exception
    // and throw it as a new one.
    IOException keep = null;
    if (tvx != null) try { tvx.close(); } catch (IOException e) { if (keep == null) keep = e; }
    if (tvd != null) try { tvd.close(); } catch (IOException e) { if (keep == null) keep = e; }
    if (tvf  != null) try {  tvf.close(); } catch (IOException e) { if (keep == null) keep = e; }
    if (keep != null) throw (IOException) keep.fillInStackTrace();
  }

  /**
   * 
   * @return The number of documents in the reader
   */
  int size() {
    return size;
  }

  public void get(int docNum, String field, TermVectorMapper mapper) throws IOException {
    if (tvx != null) {
      int fieldNumber = fieldInfos.fieldNumber(field);
      //We need to account for the FORMAT_SIZE at when seeking in the tvx
      //We don't need to do this in other seeks because we already have the
      // file pointer
      //that was written in another file
      seekTvx(docNum);
      //System.out.println("TVX Pointer: " + tvx.getFilePointer());
      long tvdPosition = tvx.readLong();

      tvd.seek(tvdPosition);
      int fieldCount = tvd.readVInt();
      //System.out.println("Num Fields: " + fieldCount);
      // There are only a few fields per document. We opt for a full scan
      // rather then requiring that they be ordered. We need to read through
      // all of the fields anyway to get to the tvf pointers.
      int number = 0;
      int found = -1;
      for (int i = 0; i < fieldCount; i++) {
        if (format >= FORMAT_VERSION)
          number = tvd.readVInt();
        else
          number += tvd.readVInt();

        if (number == fieldNumber)
          found = i;
      }

      // This field, although valid in the segment, was not found in this
      // document
      if (found != -1) {
        // Compute position in the tvf file
        long position;
        if (format >= FORMAT_VERSION2)
          position = tvx.readLong();
        else
          position = tvd.readVLong();
        for (int i = 1; i <= found; i++)
          position += tvd.readVLong();

        mapper.setDocumentNumber(docNum);
        readTermVector(field, position, mapper);
      } else {
        //System.out.println("Fieldable not found");
      }
    } else {
      //System.out.println("No tvx file");
    }
  }



  /**
   * Retrieve the term vector for the given document and field
   * @param docNum The document number to retrieve the vector for
   * @param field The field within the document to retrieve
   * @return The TermFreqVector for the document and field or null if there is no termVector for this field.
   * @throws IOException if there is an error reading the term vector files
   */ 
  TermFreqVector get(int docNum, String field) throws IOException {
    // Check if no term vectors are available for this segment at all
    ParallelArrayTermVectorMapper mapper = new ParallelArrayTermVectorMapper();
    get(docNum, field, mapper);

    return mapper.materializeVector();
  }

  // Reads the String[] fields; you have to pre-seek tvd to
  // the right point
  final private String[] readFields(int fieldCount) throws IOException {
    int number = 0;
    String[] fields = new String[fieldCount];

    for (int i = 0; i < fieldCount; i++) {