public void testInputFormat() {

    try {
      JobConf conf = new JobConf();
      String TMP_DIR = System.getProperty("test.build.data", "/tmp");
      Path filename = new Path("file:///" + TMP_DIR + "/tmpSeqFile");
      SequenceFile.Writer sfw =
          SequenceFile.createWriter(
              FileSystem.getLocal(conf),
              conf,
              filename,
              ChukwaArchiveKey.class,
              ChunkImpl.class,
              SequenceFile.CompressionType.NONE,
              Reporter.NULL);

      StringBuilder buf = new StringBuilder();
      int offsets[] = new int[lines.length];
      for (int i = 0; i < lines.length; ++i) {
        buf.append(lines[i]);
        buf.append("\n");
        offsets[i] = buf.length() - 1;
      }
      ChukwaArchiveKey key = new ChukwaArchiveKey(0, "datatype", "sname", 0);
      ChunkImpl val = new ChunkImpl("datatype", "sname", 0, buf.toString().getBytes(), null);
      val.setRecordOffsets(offsets);
      sfw.append(key, val);
      sfw.append(key, val); // write it twice
      sfw.close();

      long len = FileSystem.getLocal(conf).getFileStatus(filename).getLen();
      InputSplit split = new FileSplit(filename, 0, len, (String[]) null);
      ChukwaInputFormat in = new ChukwaInputFormat();
      RecordReader<LongWritable, Text> r = in.getRecordReader(split, conf, Reporter.NULL);

      LongWritable l = r.createKey();
      Text line = r.createValue();
      for (int i = 0; i < lines.length * 2; ++i) {
        boolean succeeded = r.next(l, line);
        assertTrue(succeeded);
        assertEquals(i, l.get());
        assertEquals(lines[i % lines.length], line.toString());
        System.out.println("read line: " + l.get() + " " + line);
      }
      boolean succeeded = r.next(l, line);
      assertFalse(succeeded);

    } catch (IOException e) {
      e.printStackTrace();
      fail("IO exception " + e);
    }
  }
Beispiel #2
0
    /**
     * Produce splits such that each is no greater than the quotient of the total size and the
     * number of splits requested.
     *
     * @param job The handle to the JobConf object
     * @param numSplits Number of splits requested
     */
    public InputSplit[] getSplits(JobConf job, int numSplits) throws IOException {
      int cnfiles = job.getInt(SRC_COUNT_LABEL, -1);
      long cbsize = job.getLong(TOTAL_SIZE_LABEL, -1);
      String srcfilelist = job.get(SRC_LIST_LABEL, "");
      if (cnfiles < 0 || cbsize < 0 || "".equals(srcfilelist)) {
        throw new RuntimeException(
            "Invalid metadata: #files("
                + cnfiles
                + ") total_size("
                + cbsize
                + ") listuri("
                + srcfilelist
                + ")");
      }
      Path src = new Path(srcfilelist);
      FileSystem fs = src.getFileSystem(job);
      FileStatus srcst = fs.getFileStatus(src);

      ArrayList<FileSplit> splits = new ArrayList<FileSplit>(numSplits);
      LongWritable key = new LongWritable();
      FilePair value = new FilePair();
      final long targetsize = cbsize / numSplits;
      long pos = 0L;
      long last = 0L;
      long acc = 0L;
      long cbrem = srcst.getLen();
      SequenceFile.Reader sl = null;
      try {
        sl = new SequenceFile.Reader(fs, src, job);
        for (; sl.next(key, value); last = sl.getPosition()) {
          // if adding this split would put this split past the target size,
          // cut the last split and put this next file in the next split.
          if (acc + key.get() > targetsize && acc != 0) {
            long splitsize = last - pos;
            splits.add(new FileSplit(src, pos, splitsize, (String[]) null));
            cbrem -= splitsize;
            pos = last;
            acc = 0L;
          }
          acc += key.get();
        }
      } finally {
        checkAndClose(sl);
      }
      if (cbrem != 0) {
        splits.add(new FileSplit(src, pos, cbrem, (String[]) null));
      }

      return splits.toArray(new FileSplit[splits.size()]);
    }