Example #1
0
  @Override
  protected Object doWork() {
    IOUtil.assertFileIsReadable(INPUT);
    IOUtil.assertFileIsWritable(OUTPUT);

    if (INPUT.getAbsolutePath().endsWith(".sam")) {
      throw new UserException("SAM files are not supported");
    }

    final SAMFileHeader samFileHeader =
        SamReaderFactory.makeDefault().referenceSequence(REFERENCE_SEQUENCE).getFileHeader(INPUT);
    for (final String comment : COMMENT) {
      if (comment.contains("\n")) {
        throw new UserException("Comments can not contain a new line");
      }
      samFileHeader.addComment(comment);
    }

    BamFileIoUtils.reheaderBamFile(samFileHeader, INPUT, OUTPUT, CREATE_MD5_FILE, CREATE_INDEX);

    return null;
  }
Example #2
0
  @Override
  public int doWork(String[] args) {
    boolean compressed = false;
    int maxRecordsInRAM = 100000;
    long count = -1L;
    File fileout = null;
    com.github.lindenb.jvarkit.util.cli.GetOpt opt =
        new com.github.lindenb.jvarkit.util.cli.GetOpt();
    int c;
    while ((c = opt.getopt(args, getGetOptDefault() + "o:n:N:T:b")) != -1) {
      switch (c) {
        case 'b':
          compressed = true;
          break;
        case 'N':
          maxRecordsInRAM = Integer.parseInt(opt.getOptArg());
          break;
        case 'n':
          count = Long.parseLong(opt.getOptArg());
          break;
        case 'o':
          fileout = new File(opt.getOptArg());
          break;
        case 'T':
          this.addTmpDirectory(new File(opt.getOptArg()));
          break;
        default:
          {
            switch (handleOtherOptions(c, opt, null)) {
              case EXIT_FAILURE:
                return -1;
              case EXIT_SUCCESS:
                return 0;
              default:
                break;
            }
          }
      }
    }
    if (count < -1L) // -1 == infinite
    {
      error("Bad count:" + count);
      return -1;
    }
    SamReader samReader = null;
    SAMRecordIterator iter = null;
    SAMFileWriter samWriter = null;
    Random random = new Random();
    CloseableIterator<RandSamRecord> iter2 = null;
    try {
      SamFileReaderFactory.setDefaultValidationStringency(ValidationStringency.SILENT);
      if (opt.getOptInd() == args.length) {
        info("Reading from stdin");
        samReader = SamFileReaderFactory.mewInstance().openStdin();
      } else if (opt.getOptInd() + 1 == args.length) {
        File filename = new File(args[opt.getOptInd()]);
        info("Reading from " + filename);
        samReader = SamFileReaderFactory.mewInstance().open(filename);
      } else {
        error("Illegal number of arguments.");
        return -1;
      }
      SAMFileHeader header = samReader.getFileHeader();

      header = header.clone();
      header.setSortOrder(SortOrder.unsorted);
      header.addComment("Processed with " + getProgramName() + " : " + getProgramCommandLine());
      SAMFileWriterFactory sfw = new SAMFileWriterFactory();
      sfw.setCreateIndex(false);
      sfw.setCreateMd5File(false);
      if (fileout == null) {
        if (compressed) {
          samWriter = sfw.makeBAMWriter(header, true, System.out);
        } else {
          samWriter = sfw.makeSAMWriter(header, true, System.out);
        }
      } else {
        samWriter = sfw.makeSAMOrBAMWriter(header, true, fileout);
        this.addTmpDirectory(fileout);
      }
      iter = samReader.iterator();
      SAMSequenceDictionaryProgress progress =
          new SAMSequenceDictionaryProgress(samReader.getFileHeader().getSequenceDictionary());

      SortingCollection<RandSamRecord> sorter =
          SortingCollection.newInstance(
              RandSamRecord.class,
              new RandSamRecordCodec(header),
              new RandSamRecordComparator(),
              maxRecordsInRAM,
              getTmpDirectories());
      sorter.setDestructiveIteration(true);
      while (iter.hasNext()) {
        RandSamRecord r = new RandSamRecord();
        r.rand_index = random.nextInt();
        r.samRecord = progress.watch(iter.next());

        sorter.add(r);
      }
      iter.close();
      iter = null;

      sorter.doneAdding();
      iter2 = sorter.iterator();
      if (count == -1) {
        while (iter2.hasNext()) {
          samWriter.addAlignment(iter2.next().samRecord);
        }
      } else {
        while (iter2.hasNext() && count > 0) {
          samWriter.addAlignment(iter2.next().samRecord);
          count--;
        }
      }
      iter2.close();
      iter2 = null;
      sorter.cleanup();
      progress.finish();
    } catch (Exception e) {
      error(e);
      return -1;
    } finally {
      CloserUtil.close(iter);
      CloserUtil.close(iter2);
      CloserUtil.close(samReader);
      CloserUtil.close(samWriter);
    }
    return 0;
  }
Example #3
0
  /** Combines multiple SAM/BAM files into one. */
  @Override
  protected int doWork() {
    boolean matchedSortOrders = true;

    // read interval list if it is defined
    final List<Interval> intervalList =
        (INTERVALS == null ? null : IntervalList.fromFile(INTERVALS).uniqued().getIntervals());
    // map reader->iterator used if INTERVALS is defined
    final Map<SamReader, CloseableIterator<SAMRecord>> samReaderToIterator =
        new HashMap<SamReader, CloseableIterator<SAMRecord>>(INPUT.size());

    // Open the files for reading and writing
    final List<SamReader> readers = new ArrayList<SamReader>();
    final List<SAMFileHeader> headers = new ArrayList<SAMFileHeader>();
    {
      SAMSequenceDictionary dict = null; // Used to try and reduce redundant SDs in memory

      for (final File inFile : INPUT) {
        IOUtil.assertFileIsReadable(inFile);
        final SamReader in =
            SamReaderFactory.makeDefault().referenceSequence(REFERENCE_SEQUENCE).open(inFile);
        if (INTERVALS != null) {
          if (!in.hasIndex())
            throw new PicardException(
                "Merging with interval but Bam file is not indexed " + inFile);
          final CloseableIterator<SAMRecord> samIterator =
              new SamRecordIntervalIteratorFactory()
                  .makeSamRecordIntervalIterator(in, intervalList, true);
          samReaderToIterator.put(in, samIterator);
        }

        readers.add(in);
        headers.add(in.getFileHeader());

        // A slightly hackish attempt to keep memory consumption down when merging multiple files
        // with
        // large sequence dictionaries (10,000s of sequences). If the dictionaries are identical,
        // then
        // replace the duplicate copies with a single dictionary to reduce the memory footprint.
        if (dict == null) {
          dict = in.getFileHeader().getSequenceDictionary();
        } else if (dict.equals(in.getFileHeader().getSequenceDictionary())) {
          in.getFileHeader().setSequenceDictionary(dict);
        }

        matchedSortOrders = matchedSortOrders && in.getFileHeader().getSortOrder() == SORT_ORDER;
      }
    }

    // If all the input sort orders match the output sort order then just merge them and
    // write on the fly, otherwise setup to merge and sort before writing out the final file
    IOUtil.assertFileIsWritable(OUTPUT);
    final boolean presorted;
    final SAMFileHeader.SortOrder headerMergerSortOrder;
    final boolean mergingSamRecordIteratorAssumeSorted;

    if (matchedSortOrders
        || SORT_ORDER == SAMFileHeader.SortOrder.unsorted
        || ASSUME_SORTED
        || INTERVALS != null) {
      log.info(
          "Input files are in same order as output so sorting to temp directory is not needed.");
      headerMergerSortOrder = SORT_ORDER;
      mergingSamRecordIteratorAssumeSorted = ASSUME_SORTED;
      presorted = true;
    } else {
      log.info("Sorting input files using temp directory " + TMP_DIR);
      headerMergerSortOrder = SAMFileHeader.SortOrder.unsorted;
      mergingSamRecordIteratorAssumeSorted = false;
      presorted = false;
    }
    final SamFileHeaderMerger headerMerger =
        new SamFileHeaderMerger(headerMergerSortOrder, headers, MERGE_SEQUENCE_DICTIONARIES);
    final MergingSamRecordIterator iterator;
    // no interval defined, get an iterator for the whole bam
    if (intervalList == null) {
      iterator =
          new MergingSamRecordIterator(headerMerger, readers, mergingSamRecordIteratorAssumeSorted);
    } else {
      // show warning related to https://github.com/broadinstitute/picard/pull/314/files
      log.info(
          "Warning: merged bams from different interval lists may contain the same read in both files");
      iterator = new MergingSamRecordIterator(headerMerger, samReaderToIterator, true);
    }
    final SAMFileHeader header = headerMerger.getMergedHeader();
    for (final String comment : COMMENT) {
      header.addComment(comment);
    }
    header.setSortOrder(SORT_ORDER);
    final SAMFileWriterFactory samFileWriterFactory = new SAMFileWriterFactory();
    if (USE_THREADING) {
      samFileWriterFactory.setUseAsyncIo(true);
    }
    final SAMFileWriter out = samFileWriterFactory.makeSAMOrBAMWriter(header, presorted, OUTPUT);

    // Lastly loop through and write out the records
    final ProgressLogger progress = new ProgressLogger(log, PROGRESS_INTERVAL);
    while (iterator.hasNext()) {
      final SAMRecord record = iterator.next();
      out.addAlignment(record);
      progress.record(record);
    }

    log.info("Finished reading inputs.");
    for (final CloseableIterator<SAMRecord> iter : samReaderToIterator.values())
      CloserUtil.close(iter);
    CloserUtil.close(readers);
    out.close();
    return 0;
  }