Java Partitioner.getPartition方法代码示例

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
private void assertData(int totalShardCount) throws IOException {
  Partitioner<IntWritable, IntWritable> partitioner = new HashPartitioner<IntWritable, IntWritable>();
  for (int i = 0; i < totalShardCount; i++) {
    HdfsDirectory directory = new HdfsDirectory(configuration, new Path(path, ShardUtil.getShardName(i)));
    DirectoryReader reader = DirectoryReader.open(directory);
    int numDocs = reader.numDocs();
    for (int d = 0; d < numDocs; d++) {
      Document document = reader.document(d);
      IndexableField field = document.getField("id");
      Integer id = (Integer) field.numericValue();
      int partition = partitioner.getPartition(new IntWritable(id), null, totalShardCount);
      assertEquals(i, partition);
    }
    reader.close();
  }
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
private static void createShard(Configuration configuration, int i, Path path, int totalShardCount)
    throws IOException {
  HdfsDirectory hdfsDirectory = new HdfsDirectory(configuration, path);
  IndexWriterConfig conf = new IndexWriterConfig(Version.LUCENE_43, new KeywordAnalyzer());
  TieredMergePolicy mergePolicy = (TieredMergePolicy) conf.getMergePolicy();
  mergePolicy.setUseCompoundFile(false);
  IndexWriter indexWriter = new IndexWriter(hdfsDirectory, conf);

  Partitioner<IntWritable, IntWritable> partitioner = new HashPartitioner<IntWritable, IntWritable>();
  int partition = partitioner.getPartition(new IntWritable(i), null, totalShardCount);
  assertEquals(i, partition);

  Document doc = getDoc(i);
  indexWriter.addDocument(doc);
  indexWriter.close();
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
/** Get an entry from output generated by this class. */
public static <K extends WritableComparable<?>, V extends Writable>
    Writable getEntry(MapFile.Reader[] readers, 
    Partitioner<K, V> partitioner, K key, V value) throws IOException {
  int part = partitioner.getPartition(key, value, readers.length);
  return readers[part].get(key, value);
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
/** Get an entry from output generated by this class. */
public static <K extends WritableComparable<?>, V extends Writable>
    Writable getEntry(MapFile.Reader[] readers, 
    Partitioner<K, V> partitioner, K key, V value) throws IOException {
  int readerLength = readers.length;
  int part;
  if (readerLength <= 1) {
    part = 0;
  } else {
    part = partitioner.getPartition(key, value, readers.length);
  }
  return readers[part].get(key, value);
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
/**
 * Validates the first non-empty partition hfile has right partitioning function.
 * It reads several keys, then calculates the partition according to the partitioning function
 * client offering. If the calculated partition number is different with actual partition number
 * an exception is thrown. If all partition hfiles are empty, an exception is thrown.
 *
 * @param parts full absolute path for all partitions
 * @param partitionerType type of paritioning function
 * @param numShards total number of partitions
 * @throws IOException if something goes wrong when reading the hfiles
 * @throws IllegalArgumentException if the partitioner type is wrong or all partitions are empty
 */
public void validate(List<Path> parts, PartitionerType partitionerType, int numShards)
    throws IOException {
  boolean hasNonEmptyPartition = false;
  HColumnDescriptor columnDescriptor = new HColumnDescriptor();
  // Disable block cache to ensure it reads the actual file content.
  columnDescriptor.setBlockCacheEnabled(false);
  for (int shardIndex = 0; shardIndex < parts.size(); shardIndex++) {
    Path fileToBeValidated = parts.get(shardIndex);
    HFile.Reader reader = null;
    try {
      FileSystem fs = FileSystem.newInstance(fileToBeValidated.toUri(), conf);
      CacheConfig cc = new CacheConfig(conf, columnDescriptor);
      reader = HFile.createReader(fs, fileToBeValidated, cc);
      Partitioner partitioner = PartitionerFactory.getPartitioner(partitionerType);
      byte[] rowKey = reader.getFirstRowKey();
      if (rowKey == null) {
        LOG.warn(String.format("empty partition %s", fileToBeValidated.toString()));
        reader.close();
        continue;
      }
      hasNonEmptyPartition = true;
      BytesWritable key = new BytesWritable(rowKey);
      int partition = partitioner.getPartition(key, null,  numShards);
      if (partition != shardIndex) {
        throw new IllegalArgumentException(
            String.format("wrong partition type %s for key %s in partition %d, expected %d",
                partitionerType.toString(), new String(key.getBytes()), shardIndex, partition)
        );
      }
    } finally {
      if (reader != null) {
        reader.close();
      }
    }
  }
  if (!hasNonEmptyPartition) {
    throw new IllegalArgumentException("all partitions are empty");
  }
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
/**
 * Generate hfiles for testing purpose
 *
 * @param sourceFileSystem source file system
 * @param conf configuration for hfile
 * @param outputFolder output folder for generated hfiles
 * @param partitionerType partitioner type
 * @param numOfPartitions number of partitions
 * @param numOfKeys number of keys
 * @return list of generated hfiles
 * @throws IOException if hfile creation goes wrong
 */
public static List<Path> generateHFiles(FileSystem sourceFileSystem, Configuration conf,
                                        File outputFolder, PartitionerType partitionerType,
                                        int numOfPartitions, int numOfKeys)
    throws IOException {
  StoreFile.Writer[] writers = new StoreFile.Writer[numOfPartitions];
  for (int i = 0; i < numOfPartitions; i++) {
    writers[i] = new StoreFile.WriterBuilder(conf, new CacheConfig(conf), sourceFileSystem, 4096)
        .withFilePath(new Path(String.format("%s/%s", outputFolder.getAbsoluteFile(),
            TerrapinUtil.formatPartitionName(i))))
        .withCompression(Compression.Algorithm.NONE)
        .build();
  }
  Partitioner partitioner = PartitionerFactory.getPartitioner(partitionerType);
  for (int i = 0; i < numOfKeys; i++) {
    byte[] key = String.format("%06d", i).getBytes();
    byte[] value;
    if (i <= 1) {
      value = "".getBytes();
    } else {
      value = ("v" + (i + 1)).getBytes();
    }
    KeyValue kv = new KeyValue(key, Bytes.toBytes("cf"), Bytes.toBytes(""), value);
    int partition = partitioner.getPartition(new BytesWritable(key), new BytesWritable(value),
        numOfPartitions);
    writers[partition].append(kv);
  }
  for (int i = 0; i < numOfPartitions; i++) {
    writers[i].close();
  }
  return Lists.transform(Lists.newArrayList(writers), new Function<StoreFile.Writer, Path>() {
    @Override
    public Path apply(StoreFile.Writer writer) {
      return writer.getPath();
    }
  });
}
 

import org.apache.hadoop.mapreduce.Partitioner; //导入方法依赖的package包/类
@Override
public int run(String[] args) throws Exception {
	if (args.length != 2) {
		JobBuilder.printUsage(this, "<path> <key>");
		return -1;
	}
	Path path = new Path(args[0]);
	IntWritable key = new IntWritable(Integer.parseInt(args[1]));

	Reader[] readers = MapFileOutputFormat.getReaders(path, getConf());
	Partitioner<IntWritable, Text> partitioner = new HashPartitioner<IntWritable, Text>();
	Text val = new Text();

	Reader reader = readers[partitioner.getPartition(key, val,
			readers.length)];
	Writable entry = reader.get(key, val);
	if (entry == null) {
		System.err.println("Key not found: " + key);
		return -1;
	}
	NcdcRecordParser parser = new NcdcRecordParser();
	IntWritable nextKey = new IntWritable();
	do {
		parser.parse(val.toString());
		System.out.printf("%s\t%s\n", parser.getStationId(),
				parser.getYear());
	} while (reader.next(nextKey, val) && key.equals(nextKey));
	return 0;
}