Ejemplo n.º 1
0
  /** Sets up configuration based on params */
  private static boolean setup(Hashtable<String, String> curConf, Configuration argConf) {

    if (argConf.get("file") == null) {
      logger.fatal("Missing file parameter");
      System.exit(1);
    }

    if (argConf.get("hdfs_base_path") == null) {
      logger.fatal("Missing HDFS base path, check gestore-conf.xml");
      System.exit(1);
    }

    if (argConf.get("hdfs_temp_path") == null) {
      logger.fatal("Missing HDFS temp path, check gestore-conf.xml");
      System.exit(1);
    }

    if (argConf.get("local_temp_path") == null) {
      logger.fatal("Missing local temp path, check gestore-conf.xml");
      System.exit(1);
    }

    // Input paramaters
    curConf.put("run_id", argConf.get("run", ""));
    curConf.put("task_id", argConf.get("task", ""));
    curConf.put("file_id", argConf.get("file"));
    curConf.put("local_path", argConf.get("path", ""));
    curConf.put("type", argConf.get("type", "l2r"));
    curConf.put("timestamp_start", argConf.get("timestamp_start", "1"));
    curConf.put(
        "timestamp_stop", argConf.get("timestamp_stop", Integer.toString(Integer.MAX_VALUE)));
    curConf.put("delimiter", argConf.get("regex", "ID=.*"));
    curConf.put("taxon", argConf.get("taxon", "all"));
    curConf.put("intermediate", argConf.get("full_run", "false"));
    curConf.put("quick_add", argConf.get("quick_add", "false"));
    Boolean full_run = curConf.get("intermediate").matches("(?i).*true.*");
    curConf.put("format", argConf.get("format", "unknown"));
    curConf.put("split", argConf.get("split", "1"));
    curConf.put("copy", argConf.get("copy", "true"));

    // Constants
    curConf.put("base_path", argConf.get("hdfs_base_path"));
    curConf.put("temp_path", argConf.get("hdfs_temp_path"));
    curConf.put("local_temp_path", argConf.get("local_temp_path"));
    curConf.put("db_name_files", argConf.get("hbase_file_table"));
    curConf.put("db_name_runs", argConf.get("hbase_run_table"));
    curConf.put("db_name_updates", argConf.get("hbase_db_update_table"));

    // Timestamps
    Date currentTime = new Date();
    Date endDate = new Date(new Long(curConf.get("timestamp_stop")));
    curConf.put("timestamp_real", Long.toString(currentTime.getTime()));

    return true;
  }
  @Override
  public QueryResult<Variant> getAllVariantsByRegionAndStudy(
      Region region, String sourceId, QueryOptions options) {
    Long start, end, dbstart, dbend;
    start = System.currentTimeMillis();
    QueryResult<Variant> queryResult =
        new QueryResult<>(
            String.format("%s:%d-%d", region.getChromosome(), region.getStart(), region.getEnd()));
    List<Variant> results = new LinkedList<>();

    boolean includeSamples;
    boolean includeStats;
    boolean includeEffects;
    if (!options.containsKey("samples")
        && !options.containsKey("stats")
        && !options.containsKey("effects")) {
      includeSamples = true;
      includeStats = true;
      includeEffects = true;
    } else {
      includeSamples = options.containsKey("samples") && options.getBoolean("samples");
      includeStats = options.containsKey("stats") && options.getBoolean("stats");
      includeEffects = options.containsKey("effects") && options.getBoolean("effects");
    }

    try {
      String startRow = buildRowkey(region.getChromosome(), Long.toString(region.getStart()));
      String stopRow = buildRowkey(region.getChromosome(), Long.toString(region.getEnd()));
      HTable table = new HTable(admin.getConfiguration(), tableName);
      dbstart = System.currentTimeMillis();
      Scan regionScan = new Scan(startRow.getBytes(), stopRow.getBytes());
      ResultScanner scanres = table.getScanner(regionScan);
      dbend = System.currentTimeMillis();
      queryResult.setDbTime(dbend - dbstart);

      // Iterate over results and, optionally, their samples and statistics
      for (Result result : scanres) {
        String[] rowkeyParts = new String(result.getRow(), CHARSET_UTF_8).split("_");
        String chromosome = rowkeyParts[0].replaceFirst("^0+(?!$)", "");
        int position = Integer.parseInt(rowkeyParts[1]);

        // Get basic result fields from Protocol Buffers message
        NavigableMap<byte[], byte[]> infoMap = result.getFamilyMap("i".getBytes());
        byte[] byteInfo = infoMap.get((sourceId + "_data").getBytes());
        VariantFieldsProtos.VariantInfo protoInfo =
            VariantFieldsProtos.VariantInfo.parseFrom(byteInfo);
        String reference = protoInfo.getReference();
        String alternate = StringUtils.join(protoInfo.getAlternateList(), ",");
        String format = StringUtils.join(protoInfo.getFormatList(), ":");
        Variant variant = new Variant(chromosome, position, position, reference, alternate);

        // Set samples if requested
        if (includeSamples) {
          NavigableMap<byte[], byte[]> sampleMap = result.getFamilyMap("d".getBytes());
          Map<String, Map<String, String>> resultSampleMap = new HashMap<>();

          // Set samples
          for (byte[] s : sampleMap.keySet()) {
            String sampleName = (new String(s, CHARSET_UTF_8)).replaceAll(sourceId + "_", "");
            VariantFieldsProtos.VariantSample sample =
                VariantFieldsProtos.VariantSample.parseFrom(sampleMap.get(s));
            String sample1 = sample.getSample();
            String[] values = sample1.split(":");
            String[] fields = format.split(":");
            Map<String, String> singleSampleMap = new HashMap<>();
            for (int i = 0; i < fields.length; i++) {
              singleSampleMap.put(fields[i], values[i]);
            }
            // TODO
            //                        variant.addSampleData(sampleName, singleSampleMap);
          }
        }

        // Set stats if requested
        if (includeStats) {
          byte[] byteStats = infoMap.get((sourceId + "_stats").getBytes());
          VariantFieldsProtos.VariantStats protoStats =
              VariantFieldsProtos.VariantStats.parseFrom(byteStats);
          VariantStats variantStats =
              new VariantStats(
                  chromosome,
                  position,
                  reference,
                  alternate,
                  protoStats.getMaf(),
                  protoStats.getMgf(),
                  protoStats.getMafAllele(),
                  protoStats.getMgfGenotype(),
                  protoStats.getMissingAlleles(),
                  protoStats.getMissingGenotypes(),
                  protoStats.getMendelianErrors(),
                  protoStats.getIsIndel(),
                  protoStats.getCasesPercentDominant(),
                  protoStats.getControlsPercentDominant(),
                  protoStats.getCasesPercentRecessive(),
                  protoStats.getControlsPercentRecessive());
          variant.setStats(variantStats);
        }

        if (includeEffects) {
          QueryResult<VariantEffect> queryEffects = getEffectsByVariant(variant, options);
          variant.setEffect(queryEffects.getResult());
        }

        results.add(variant);
      }
    } catch (IOException e) {
      System.err.println(e.getClass().getName() + ": " + e.getMessage());
    }
    queryResult.setResult(results);
    queryResult.setNumResults(results.size());
    end = System.currentTimeMillis();
    queryResult.setTime(end - start);
    return queryResult;
  }
  public QueryResult getSimpleVariantsByRegion(
      Region region, String sourceId, QueryOptions options) {
    Long start, end, dbstart, dbend;
    start = System.currentTimeMillis();
    boolean includeStats;
    boolean includeEffects;
    if (!options.containsKey("stats") && !options.containsKey("effects")) {
      includeStats = true;
      includeEffects = true;
    } else {
      includeStats = options.containsKey("stats") && options.getBoolean("stats");
      includeEffects = options.containsKey("effects") && options.getBoolean("effects");
    }

    QueryResult<Variant> queryResult =
        new QueryResult<>(
            String.format("%s:%d-%d", region.getChromosome(), region.getStart(), region.getEnd()));
    List<Variant> results = new ArrayList<>();
    String startRow = buildRowkey(region.getChromosome(), Long.toString(region.getStart()));
    String stopRow = buildRowkey(region.getChromosome(), Long.toString(region.getEnd()));
    BasicDBObject query =
        new BasicDBObject("position", new BasicDBObject("$gte", startRow).append("$lte", stopRow))
            .append("sources.sourceId", sourceId);
    DBCollection collection = db.getCollection("variants");
    dbstart = System.currentTimeMillis();
    DBCursor variantInStudies = collection.find(query);
    dbend = System.currentTimeMillis();
    queryResult.setDbTime(dbend - dbstart);

    for (DBObject result : variantInStudies) {
      String[] rowkeyParts = result.get("position").toString().split("_");
      String chromosome = rowkeyParts[0].replaceFirst("^0+(?!$)", "");
      int position = Integer.parseInt(rowkeyParts[1]);
      BasicDBList studies = (BasicDBList) result.get("sources");
      BasicDBObject st = (BasicDBObject) studies.get(0);
      String ref = (String) st.get("ref");
      String alt = StringUtils.join((ArrayList<String>) st.get("alt"), ",");

      // TODO Needs rework
      Variant variant = new Variant(chromosome, position, position, ref, alt);

      // Set stats informations
      if (includeStats) {
        VariantStats stats = new VariantStats();
        BasicDBObject mongoStats = (BasicDBObject) st.get("stats");
        stats.setMaf((float) (double) mongoStats.get("maf"));
        stats.setMafAllele((String) mongoStats.get("alleleMaf"));
        stats.setMissingGenotypes((int) mongoStats.get("missing"));
        List<Genotype> genotypeCount = new ArrayList<>();
        for (BasicDBObject s : (List<BasicDBObject>) mongoStats.get("genotypeCount")) {
          for (Map.Entry<String, Object> entry : s.entrySet()) {
            Genotype genotype = new Genotype(entry.getKey());
            genotype.setCount((Integer) entry.getValue());
            genotypeCount.add(genotype);
          }
        }
        stats.setGenotypes(genotypeCount);
        variant.setStats(stats);
      }

      // TODO Set consequence type names
      if (includeEffects) {
        BasicDBList mongoEffects = (BasicDBList) st.get("effects");
        if (mongoEffects != null) {
          for (Object e : mongoEffects) {
            String effectObo = e.toString();
            VariantEffect effect = new VariantEffect();
            effect.setConsequenceTypeObo(effectObo);
            variant.addEffect(effect);
          }
        }
      }

      results.add(variant);
    }

    queryResult.setResult(results);
    queryResult.setNumResults(results.size());
    end = System.currentTimeMillis();
    queryResult.setTime(end - start);
    return queryResult;
  }
  @Override
  public QueryResult getVariantsHistogramByRegion(
      Region region, String sourceId, boolean histogramLogarithm, int histogramMax) {
    QueryResult<ObjectMap> queryResult =
        new QueryResult<>(
            String.format("%s:%d-%d", region.getChromosome(), region.getStart(), region.getEnd()));
    List<ObjectMap> data = new ArrayList<>();
    String startRow = buildRowkey(region.getChromosome(), Long.toString(region.getStart()));
    String stopRow = buildRowkey(region.getChromosome(), Long.toString(region.getEnd()));

    long startTime = System.currentTimeMillis();

    long startDbTime = System.currentTimeMillis();

    BasicDBObject query =
        new BasicDBObject("position", new BasicDBObject("$gte", startRow).append("$lte", stopRow))
            .append("studies.studyId", sourceId);
    DBCollection collection = db.getCollection("variants");
    DBCursor queryResults = collection.find(query);
    queryResult.setDbTime(System.currentTimeMillis() - startDbTime);

    int resultSize = queryResults.size();

    if (resultSize > histogramMax) { // Need to group results to fit maximum size of the histogram
      int sumChunkSize = resultSize / histogramMax;
      int i = 0, j = 0;
      int featuresCount = 0;
      ObjectMap item = null;

      for (DBObject result : queryResults) {
        //                featuresCount += result.getInt("features_count");
        //                if (i == 0) {
        //                    item = new ObjectMap("chromosome", result.getString("chromosome"));
        //                    item.put("chunkId", result.getInt("chunk_id"));
        //                    item.put("start", result.getInt("start"));
        //                } else if (i == sumChunkSize - 1 || j == resultSize - 1) {
        //                    if (histogramLogarithm) {
        //                        item.put("featuresCount", (featuresCount > 0) ?
        // Math.log(featuresCount) : 0);
        //                    } else {
        //                        item.put("featuresCount", featuresCount);
        //                    }
        //                    item.put("end", result.getInt("end"));
        //                    data.add(item);
        //                    i = -1;
        //                    featuresCount = 0;
        //                }
        //                j++;
        //                i++;
      }
    } else {
      for (DBObject result : queryResults) {
        //                ObjectMap item = new ObjectMap("chromosome",
        // result.getString("chromosome"));
        //                item.put("chunkId", result.getInt("chunk_id"));
        //                item.put("start", result.getInt("start"));
        //                if (histogramLogarithm) {
        //                    int features_count = result.getInt("features_count");
        //                    result.put("featuresCount", (features_count > 0) ?
        // Math.log(features_count) : 0);
        //                } else {
        //                    item.put("featuresCount", result.getInt("features_count"));
        //                }
        //                item.put("end", result.getInt("end"));
        //                data.add(item);
      }
    }

    queryResult.setResult(data);
    queryResult.setNumResults(data.size());
    queryResult.setTime(System.currentTimeMillis() - startTime);

    return queryResult;
  }
Ejemplo n.º 5
0
  public int run(String[] args) throws Exception {
    // printUsage();
    /*
     * SETUP
     */
    Configuration argConf = getConf();
    Hashtable<String, String> confArg = new Hashtable<String, String>();
    setup(confArg, argConf);
    Date currentTime = new Date();
    Date endDate = new Date(new Long(confArg.get("timestamp_stop")));
    Boolean full_run = confArg.get("intermediate").matches("(?i).*true.*");
    Boolean quick_add = confArg.get("quick_add").matches("(?i).*true.*");
    logger.info("Running GeStore");

    // ZooKeeper setup
    Configuration config = HBaseConfiguration.create();
    zkWatcher = new ZooKeeperWatcher(config, "Testing", new HBaseAdmin(config));
    zkInstance =
        new ZooKeeper(
            ZKConfig.getZKQuorumServersString(config),
            config.getInt("zookeeper.session.timeout", -1),
            zkWatcher);

    if (!confArg.get("task_id").isEmpty()) {
      confArg.put("temp_path", confArg.get("temp_path") + confArg.get("task_id"));
    }

    String lockRequest = confArg.get("file_id");
    if (!confArg.get("run_id").isEmpty())
      lockRequest = lockRequest + "_" + confArg.get("run_id") + "_";
    if (!confArg.get("task_id").isEmpty())
      lockRequest = lockRequest + "_" + confArg.get("task_id") + "_";

    // Get type of movement
    toFrom type_move = checkArgs(confArg);
    if (type_move == toFrom.LOCAL2REMOTE && !confArg.get("format").equals("unknown")) {
      List<String> arguments = new ArrayList<String>();
      arguments.add("-Dinput=" + confArg.get("local_path"));
      arguments.add("-Dtable=" + confArg.get("file_id"));
      arguments.add("-Dtimestamp=" + confArg.get("timestamp_stop"));
      arguments.add("-Dtype=" + confArg.get("format"));
      arguments.add("-Dtarget_dir=" + confArg.get("base_path") + "_" + confArg.get("file_id"));
      arguments.add("-Dtemp_hdfs_path=" + confArg.get("temp_path"));
      arguments.add("-Drun_id=" + confArg.get("run_id"));
      if (!confArg.get("run_id").isEmpty()) arguments.add("-Drun_id=" + confArg.get("run_id"));
      if (!confArg.get("task_id").isEmpty()) arguments.add("-Dtask_id=" + confArg.get("task_id"));
      if (quick_add) arguments.add("-Dquick_add=" + confArg.get("quick_add"));
      String lockName = lock(lockRequest);
      String[] argumentString = arguments.toArray(new String[arguments.size()]);
      adddb.main(argumentString);
      unlock(lockName);
      System.exit(0);
    }

    // Database registration

    dbutil db_util = new dbutil(config);
    db_util.register_database(confArg.get("db_name_files"), true);
    db_util.register_database(confArg.get("db_name_runs"), true);
    db_util.register_database(confArg.get("db_name_updates"), true);
    FileSystem hdfs = FileSystem.get(config);
    FileSystem localFS = FileSystem.getLocal(config);

    // Get source type
    confArg.put("source", getSource(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
    confArg.put(
        "database", isDatabase(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
    if (!confArg.get("source").equals("local")
        && type_move == toFrom.REMOTE2LOCAL
        && !confArg.get("timestamp_stop").equals(Integer.toString(Integer.MAX_VALUE))) {
      confArg.put("timestamp_stop", Long.toString(latestVersion(confArg, db_util)));
    }

    /*
     * Get previous timestamp
     */
    Get run_id_get = new Get(confArg.get("run_id").getBytes());
    Result run_get = db_util.doGet(confArg.get("db_name_runs"), run_id_get);
    KeyValue run_file_prev =
        run_get.getColumnLatest(
            "d".getBytes(), (confArg.get("file_id") + "_db_timestamp").getBytes());
    String last_timestamp = new String("0");
    if (null != run_file_prev && !confArg.get("source").equals("local")) {
      long last_timestamp_real = run_file_prev.getTimestamp();
      Long current_timestamp = new Long(confArg.get("timestamp_real"));
      if ((current_timestamp - last_timestamp_real) > 36000) {
        last_timestamp = new String(run_file_prev.getValue());
        Integer lastTimestamp = new Integer(last_timestamp);
        lastTimestamp += 1;
        last_timestamp = lastTimestamp.toString();
        logger.info("Last timestamp: " + last_timestamp + " End data: " + endDate);
        Date last_run = new Date(run_file_prev.getTimestamp());
        if (last_run.before(endDate) && !full_run) {
          confArg.put("timestamp_start", last_timestamp);
        }
      }
    }

    Integer tse = new Integer(confArg.get("timestamp_stop"));
    Integer tss = new Integer(confArg.get("timestamp_start"));
    if (tss > tse) {
      logger.info("No new version of requested file.");
      return 0;
    }

    /*
     * Generate file
     */

    String lockName = lock(lockRequest);

    Get file_id_get = new Get(confArg.get("file_id").getBytes());
    Result file_get = db_util.doGet(confArg.get("db_name_files"), file_id_get);
    if (!file_get.isEmpty()) {
      boolean found =
          hasFile(
              db_util,
              hdfs,
              confArg.get("db_name_files"),
              confArg.get("file_id"),
              getFullPath(confArg));
      if (confArg.get("source").equals("fullfile")) {
        found = false;
      }
      String filenames_put =
          getFileNames(
              db_util, confArg.get("db_name_files"), confArg.get("file_id"), getFullPath(confArg));
      // Filename not found in file database
      if (!found && type_move == toFrom.REMOTE2LOCAL) {
        if (!confArg.get("source").equals("local")) {
          // Generate intermediate file
          if (getFile(hdfs, confArg, db_util) == null) {
            unlock(lockName);
            return 1;
          }
          // Put generated file into file database
          if (!confArg.get("format").equals("fullfile")) {
            putFileEntry(
                db_util,
                hdfs,
                confArg.get("db_name_files"),
                confArg.get("file_id"),
                confArg.get("full_file_name"),
                confArg.get("source"));
          }
        } else {
          logger.warn("Remote file not found, and cannot be generated! File: " + confArg);
          unlock(lockName);
          return 1;
        }
      }
    } else {
      if (type_move == toFrom.REMOTE2LOCAL) {
        logger.warn("Remote file not found, and cannot be generated.");
        unlock(lockName);
        return 1;
      }
    }

    /*
     * Copy file
     * Update tables
     */

    if (type_move == toFrom.LOCAL2REMOTE) {
      if (!confArg.get("format").equals("fullfile")) {
        putFileEntry(
            db_util,
            hdfs,
            confArg.get("db_name_files"),
            confArg.get("file_id"),
            getFullPath(confArg),
            confArg.get("source"));
      }
      putRunEntry(
          db_util,
          confArg.get("db_name_runs"),
          confArg.get("run_id"),
          confArg.get("file_id"),
          confArg.get("type"),
          confArg.get("timestamp_real"),
          confArg.get("timestamp_stop"),
          getFullPath(confArg),
          confArg.get("delimiter"));
      hdfs.copyFromLocalFile(new Path(confArg.get("local_path")), new Path(getFullPath(confArg)));
    } else if (type_move == toFrom.REMOTE2LOCAL) {
      FileStatus[] files = hdfs.globStatus(new Path(getFullPath(confArg) + "*"));
      putRunEntry(
          db_util,
          confArg.get("db_name_runs"),
          confArg.get("run_id"),
          confArg.get("file_id"),
          confArg.get("type"),
          confArg.get("timestamp_real"),
          confArg.get("timestamp_stop"),
          getFullPath(confArg),
          confArg.get("delimiter"));
      unlock(lockName);
      for (FileStatus file : files) {
        Path cur_file = file.getPath();
        Path cur_local_path =
            new Path(new String(confArg.get("local_path") + confArg.get("file_id")));
        String suffix = getSuffix(getFileName(confArg), cur_file.getName());
        if (suffix.length() > 0) {
          cur_local_path = cur_local_path.suffix(new String("." + suffix));
        }
        if (confArg.get("copy").equals("true")) {
          String crc = hdfs.getFileChecksum(cur_file).toString();
          if (checksumLocalTest(cur_local_path, crc)) {
            continue;
          } else {
            hdfs.copyToLocalFile(cur_file, cur_local_path);
            writeChecksum(cur_local_path, crc);
          }
        } else {
          System.out.println(cur_local_path + "\t" + cur_file);
        }
      }
    }
    unlock(lockName);
    return 0;
  }
Ejemplo n.º 6
0
  // Information needed to get a single file:
  // BASE_PATH, FILE_ID, TIMESTAMP_START, TIMESTAMP_STOP, SOURCE, FILESYSTEM
  private static Vector<Path> getFile(
      FileSystem fs, Hashtable<String, String> config, dbutil db_util) throws Exception {
    Long latestVersion = latestVersion(config, db_util);

    try {
      config.put("timestamp_start", config.get("timestamp_start"));
      config.put("timestamp_real", latestVersion.toString());
      config.put("timestamp_stop", latestVersion.toString());
    } catch (Exception E) {
      logger.error("Tryign to get file that is impossible to generate: " + getFullPath(config));
      return null;
    }
    if (Integer.parseInt(config.get("timestamp_start"))
        > Integer.parseInt(config.get("timestamp_stop"))) {
      return null;
    }
    logger.debug(
        "Getting DB for timestamp "
            + config.get("timestamp_start")
            + " to "
            + config.get("timestamp_stop"));

    String final_result = getFullPath(config);

    String temp_path_base =
        config.get("local_temp_path")
            + "_"
            + config.get("task_id")
            + "_"
            + config.get("run_id")
            + "/";
    Path newPath = new Path(final_result + "*");
    Vector<Path> ret_path = new Vector<Path>();
    String lockName = lock(final_result.replaceAll("/", "_"));
    if (fs.globStatus(newPath).length != 0) {
      ret_path.add(newPath);
      unlock(lockName);
      config.put("full_file_name", final_result);
      return ret_path;
    } else {
      if (!config.get("source").equals("local")) {
        config.put("temp_path_base", temp_path_base);

        config.put("timestamp_start", config.get("timestamp_start"));
        config.put("timestamp_real", latestVersion.toString());
        config.put("timestamp_stop", latestVersion.toString());

        Class<?> sourceClass =
            Class.forName("org.gestore.plugin.source." + config.get("source") + "Source");
        Method process_data = sourceClass.getMethod("process", Hashtable.class, FileSystem.class);
        Object processor = sourceClass.newInstance();
        Object retVal;
        try {
          retVal = process_data.invoke(processor, config, fs);
        } catch (InvocationTargetException E) {
          Throwable exception = E.getTargetException();
          logger.error("Unable to call method in child class: " + exception.toString());
          exception.printStackTrace(System.out);
          unlock(lockName);
          return null;
        }
        FileStatus[] files = (FileStatus[]) retVal;
        if (files == null) {
          logger.error("Error getting files, no files returned");
          return null;
        }

        for (FileStatus file : files) {
          Path cur_file = file.getPath();
          Path cur_local_path = new Path(temp_path_base + config.get("file_id"));
          String suffix = getSuffix(config.get("file_id"), cur_file.getName());
          cur_local_path = cur_local_path.suffix(suffix);
          Path res_path = new Path(new String(final_result + suffix));
          logger.debug("Moving file" + cur_file.toString() + " to " + res_path.toString());
          if (config.get("copy").equals("true")) {
            fs.moveFromLocalFile(cur_file, res_path);
          } else {
            fs.rename(cur_file, res_path);
          }
        }

        config.put("full_file_name", final_result);
      }
    }
    unlock(lockName);
    return ret_path;
  }
Ejemplo n.º 7
0
  /**
   * Given the OSM XML file, parse out all of the nodes and ways and then run the import methods to
   * put the values into HBase.
   *
   * @param filename path to the OSM XML file
   * @throws XMLStreamException XML file is invalid
   */
  public static void import_from_file(String filename) throws XMLStreamException {
    System.out.println("Starting...");
    System.out.println(configuration.toString());
    XMLInputFactory xmlInputFactory = XMLInputFactory.newFactory();
    XMLStreamReader xmlStreamReader;
    try {
      xmlStreamReader = xmlInputFactory.createXMLStreamReader(new FileInputStream(filename));
    } catch (FileNotFoundException e) {
      throw new XMLStreamException("OSM XML file not found!");
    }

    // Begin reading the XML
    Node node = null;
    Way way = null;
    while (xmlStreamReader.hasNext()) {
      if (xmlStreamReader.isStartElement()) {
        // Start of Element, consider only the things we care about
        String s1 = xmlStreamReader.getLocalName();
        switch (s1) {
          case "node": // Start new node
            node = new Node();
            // Get the node attributes we care about
            for (int i = 0; i < xmlStreamReader.getAttributeCount(); i++) {
              String value = xmlStreamReader.getAttributeValue(i);
              String s = xmlStreamReader.getAttributeLocalName(i);
              switch (s) {
                case "id":
                  node.setOsmId(value);
                  break;
                case "lat":
                  node.setLat(value);
                  break;
                case "lon":
                  node.setLon(value);
                  break;
              }
              if (node.isComplete()) break;
            }
            break;
          case "way": // Start new way
            way = new Way();
            // Get the way attributes we care about
            for (int i = 0; i < xmlStreamReader.getAttributeCount(); i++) {
              String value = xmlStreamReader.getAttributeValue(i);
              String s = xmlStreamReader.getAttributeLocalName(i);
              if (s.equals("id")) {
                way.setId(value);
              }
            }
            break;
          case "nd": // Start the way's node reference
            if (way == null) {
              System.err.println("Parser Error, Turns out we needed our way object.");
              return;
            }
            // Get the id of the node
            long node_id = Long.MIN_VALUE;
            for (int i = 0; i < xmlStreamReader.getAttributeCount(); i++) {
              String s = xmlStreamReader.getAttributeLocalName(i);
              if (s.equals("ref")) {
                String value = xmlStreamReader.getAttributeValue(i);
                node_id = Long.parseLong(value);
              }
            }
            // Add node to way
            if (node_id != Long.MIN_VALUE) {
              way.addNode(nodes.get(node_id));
            }
            break;
          case "tag": // Start the tag parsing
            String key = "";
            String value = "";
            for (int i = 0; i < xmlStreamReader.getAttributeCount(); i++) {
              String s = xmlStreamReader.getAttributeLocalName(i);
              if (s.equals("k")) {
                key = xmlStreamReader.getAttributeValue(i);
              }
              if (s.equals("v")) {
                value = xmlStreamReader.getAttributeValue(i);
              }
            }
            if (!key.equals("") && !value.equals("")) {
              if (node != null) {
                // This is a node tag
                node.addTag(key, value);
                System.out.println(
                    "Added Node Tag " + key + ": " + value + "for " + node.getOsmId());
              } else if (way != null) {
                // This is a way tag
                way.addTag(key, value);
                System.out.println("Added Way Tag " + key + ": " + value + "for " + way.getId());
              }
              /*
              else {
                  // This will barf because we don't care about 'relation' data types
                  // System.err.println(key + ": " + value + " not associated to node or way!");
              }
              */
            }
            break;
        }

      } else if (xmlStreamReader.isEndElement()) {
        String s = xmlStreamReader.getLocalName();
        if (s.equals("node") && node != null) {
          nodes.put(node.getOsmId(), node);
          System.out.print("\rAdded Node " + String.valueOf(node.getOsmId()));
          node = null;
        } else if (s.equals("way") && way != null) {
          ways.add(way);
          System.out.print("\rAdded Way " + String.valueOf(way.getId()) + "     ");
          way = null;
        }
      }
      xmlStreamReader.next();
    }

    System.out.println("Number of nodes read: " + nodes.size());
    System.out.println("Number of ways read: " + ways.size());

    Main.print_heap_usage();
    import_nodes();
    Main.print_heap_usage();
    import_ways();

    System.out.println("Finished!");
  }