001/*
002 * Licensed to the Apache Software Foundation (ASF) under one
003 * or more contributor license agreements.  See the NOTICE file
004 * distributed with this work for additional information
005 * regarding copyright ownership.  The ASF licenses this file
006 * to you under the Apache License, Version 2.0 (the
007 * "License"); you may not use this file except in compliance
008 * with the License.  You may obtain a copy of the License at
009 *
010 *     http://www.apache.org/licenses/LICENSE-2.0
011 *
012 * Unless required by applicable law or agreed to in writing, software
013 * distributed under the License is distributed on an "AS IS" BASIS,
014 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
015 * See the License for the specific language governing permissions and
016 * limitations under the License.
017 */
018package org.apache.hadoop.hbase.io.encoding;
019
020import java.io.IOException;
021import java.io.OutputStream;
022import org.apache.hadoop.hbase.util.Bytes;
023import org.apache.yetus.audience.InterfaceAudience;
024
025/**
026 * Provide access to all data block encoding algorithms. All of the algorithms are required to have
027 * unique id which should <b>NEVER</b> be changed. If you want to add a new algorithm/version,
028 * assign it a new id. Announce the new id in the HBase mailing list to prevent collisions.
029 */
030@InterfaceAudience.Public
031public enum DataBlockEncoding {
032
033  /** Disable data block encoding. */
034  NONE(0, null),
035  // id 1 is reserved for the BITSET algorithm to be added later
036  PREFIX(2, "org.apache.hadoop.hbase.io.encoding.PrefixKeyDeltaEncoder"),
037  DIFF(3, "org.apache.hadoop.hbase.io.encoding.DiffKeyDeltaEncoder"),
038  FAST_DIFF(4, "org.apache.hadoop.hbase.io.encoding.FastDiffDeltaEncoder"),
039  // id 5 is reserved for the COPY_KEY algorithm for benchmarking
040  // COPY_KEY(5, "org.apache.hadoop.hbase.io.encoding.CopyKeyDataBlockEncoder"),
041  // PREFIX_TREE(6, "org.apache.hadoop.hbase.codec.prefixtree.PrefixTreeCodec"),
042  ROW_INDEX_V1(7, "org.apache.hadoop.hbase.io.encoding.RowIndexCodecV1");
043
044  private final short id;
045  private final byte[] idInBytes;
046  private DataBlockEncoder encoder;
047  private final String encoderCls;
048
049  public static final int ID_SIZE = Bytes.SIZEOF_SHORT;
050
051  /** Maps data block encoding ids to enum instances. */
052  private static DataBlockEncoding[] idArray = new DataBlockEncoding[Byte.MAX_VALUE + 1];
053
054  static {
055    for (DataBlockEncoding algo : values()) {
056      if (idArray[algo.id] != null) {
057        throw new RuntimeException(
058          String.format("Two data block encoder algorithms '%s' and '%s' have " + "the same id %d",
059            idArray[algo.id].toString(), algo.toString(), (int) algo.id));
060      }
061      idArray[algo.id] = algo;
062    }
063  }
064
065  private DataBlockEncoding(int id, String encoderClsName) {
066    if (id < 0 || id > Byte.MAX_VALUE) {
067      throw new AssertionError("Data block encoding algorithm id is out of range: " + id);
068    }
069    this.id = (short) id;
070    this.idInBytes = Bytes.toBytes(this.id);
071    if (idInBytes.length != ID_SIZE) {
072      // White this may seem redundant, if we accidentally serialize
073      // the id as e.g. an int instead of a short, all encoders will break.
074      throw new RuntimeException("Unexpected length of encoder ID byte " + "representation: "
075        + Bytes.toStringBinary(idInBytes));
076    }
077    this.encoderCls = encoderClsName;
078  }
079
080  /** Returns name converted to bytes. */
081  public byte[] getNameInBytes() {
082    return Bytes.toBytes(toString());
083  }
084
085  /** Returns The id of a data block encoder. */
086  public short getId() {
087    return id;
088  }
089
090  /**
091   * Writes id in bytes.
092   * @param stream where the id should be written.
093   */
094  public void writeIdInBytes(OutputStream stream) throws IOException {
095    stream.write(idInBytes);
096  }
097
098  /**
099   * Writes id bytes to the given array starting from offset.
100   * @param dest   output array
101   * @param offset starting offset of the output array
102   */
103  public void writeIdInBytes(byte[] dest, int offset) throws IOException {
104    System.arraycopy(idInBytes, 0, dest, offset, ID_SIZE);
105  }
106
107  /**
108   * Return new data block encoder for given algorithm type.
109   * @return data block encoder if algorithm is specified, null if none is selected.
110   */
111  public DataBlockEncoder getEncoder() {
112    if (encoder == null && id != 0) {
113      // lazily create the encoder
114      encoder = createEncoder(encoderCls);
115    }
116    return encoder;
117  }
118
119  /**
120   * Find and create data block encoder for given id;
121   * @param encoderId id of data block encoder.
122   * @return Newly created data block encoder.
123   */
124  public static DataBlockEncoder getDataBlockEncoderById(short encoderId) {
125    return getEncodingById(encoderId).getEncoder();
126  }
127
128  /**
129   * Find and return the name of data block encoder for the given id.
130   * @param encoderId id of data block encoder
131   * @return name, same as used in options in column family
132   */
133  public static String getNameFromId(short encoderId) {
134    return getEncodingById(encoderId).toString();
135  }
136
137  /**
138   * Check if given encoder has this id.
139   * @param encoder   encoder which id will be checked
140   * @param encoderId id which we except
141   * @return true if id is right for given encoder, false otherwise
142   * @exception IllegalArgumentException thrown when there is no matching data block encoder
143   */
144  public static boolean isCorrectEncoder(DataBlockEncoder encoder, short encoderId) {
145    DataBlockEncoding algorithm = getEncodingById(encoderId);
146    String encoderCls = encoder.getClass().getName();
147    return encoderCls.equals(algorithm.encoderCls);
148  }
149
150  public static DataBlockEncoding getEncodingById(short dataBlockEncodingId) {
151    DataBlockEncoding algorithm = null;
152    if (dataBlockEncodingId >= 0 && dataBlockEncodingId <= Byte.MAX_VALUE) {
153      algorithm = idArray[dataBlockEncodingId];
154    }
155    if (algorithm == null) {
156      throw new IllegalArgumentException(String
157        .format("There is no data block encoder for given id '%d'", (int) dataBlockEncodingId));
158    }
159    return algorithm;
160  }
161
162  protected static DataBlockEncoder createEncoder(String fullyQualifiedClassName) {
163    try {
164      return (DataBlockEncoder) Class.forName(fullyQualifiedClassName).getDeclaredConstructor()
165        .newInstance();
166    } catch (Exception e) {
167      throw new RuntimeException(e);
168    }
169  }
170
171}