Java Code Examples for org.apache.hadoop.hbase.client.Scan#setStartRow()

private Scan createScan(Application application, Range range, byte[] family) {
    range = rangeFactory.createStatisticsRange(range);
    if (logger.isDebugEnabled()) {
        logger.debug("scan time:{} ", range.prettyToString());
    }

    // start key is replaced by end key because timestamp has been reversed
    byte[] startKey = ApplicationMapStatisticsUtils.makeRowKey(application.getName(), application.getServiceTypeCode(), range.getTo());
    byte[] endKey = ApplicationMapStatisticsUtils.makeRowKey(application.getName(), application.getServiceTypeCode(), range.getFrom());

    final Scan scan = new Scan();
    scan.setCaching(this.scanCacheSize);
    scan.setStartRow(startKey);
    scan.setStopRow(endKey);
    scan.addFamily(family);
    scan.setId("ApplicationSelfScan");

    return scan;
}
 

public static Scan buildScan(RawScan rawScan) {
    Scan scan = new Scan();
    scan.setCaching(rawScan.hbaseCaching);
    scan.setMaxResultSize(rawScan.hbaseMaxResultSize);
    scan.setCacheBlocks(true);
    scan.setAttribute(Scan.SCAN_ATTRIBUTES_METRICS_ENABLE, Bytes.toBytes(Boolean.TRUE));

    if (rawScan.startKey != null) {
        scan.setStartRow(rawScan.startKey);
    }
    if (rawScan.endKey != null) {
        scan.setStopRow(rawScan.endKey);
    }
    if (rawScan.fuzzyKeys != null) {
        applyFuzzyFilter(scan, rawScan.fuzzyKeys);
    }
    if (rawScan.hbaseColumns != null) {
        applyHBaseColums(scan, rawScan.hbaseColumns);
    }

    return scan;
}
 

private Scan createScan(String agentId, long currentTime) {
    Scan scan = new Scan();

    byte[] agentIdBytes = Bytes.toBytes(agentId);
    long startTime = TimeUtils.reverseTimeMillis(currentTime);
    byte[] startKeyBytes = RowKeyUtils.concatFixedByteAndLong(agentIdBytes, HbaseTableConstatns.AGENT_NAME_MAX_LEN, startTime);
    byte[] endKeyBytes = RowKeyUtils.concatFixedByteAndLong(agentIdBytes, HbaseTableConstatns.AGENT_NAME_MAX_LEN, Long.MAX_VALUE);

    scan.setStartRow(startKeyBytes);
    scan.setStopRow(endKeyBytes);
    scan.addFamily(descriptor.getColumnFamilyName());

    scan.setMaxVersions(1);
    scan.setCaching(SCANNER_CACHING);

    return scan;
}
 

protected Scan getMultiScanner(final FilterList filterList) {
  // Single scan w/ multiple ranges
  final Scan multiScanner = scanProvider.get();
  final List<ByteArrayRange> ranges = readerParams.getQueryRanges().getCompositeQueryRanges();

  final MultiRowRangeFilter filter = operations.getMultiRowRangeFilter(ranges);
  if (filter != null) {
    filterList.addFilter(filter);

    final List<RowRange> rowRanges = filter.getRowRanges();
    multiScanner.setStartRow(rowRanges.get(0).getStartRow());

    final RowRange stopRowRange = rowRanges.get(rowRanges.size() - 1);
    byte[] stopRowExclusive;
    if (stopRowRange.isStopRowInclusive()) {
      // because the end is always exclusive, to make an inclusive
      // stop row into exlusive all we need to do is add a traling 0
      stopRowExclusive = HBaseUtils.getInclusiveEndKey(stopRowRange.getStopRow());
    } else {
      stopRowExclusive = stopRowRange.getStopRow();
    }
    multiScanner.setStopRow(stopRowExclusive);
  }
  return multiScanner;
}
 

@Test
public void testGetLowerUnboundSplits() throws Exception {
    long ts = nextTimestamp();
    initTableValues(ts);
    String url = getUrl() + ";" + PhoenixRuntime.CURRENT_SCN_ATTRIB + "=" + ts;
    Properties props = new Properties(TEST_PROPERTIES);
    Connection conn = DriverManager.getConnection(url, props);

    Scan scan = new Scan();
    
    ConnectionQueryServices services = driver.getConnectionQueryServices(getUrl(), TEST_PROPERTIES);
    TableRef table = getTableRef(conn,ts);
    services.getStatsManager().updateStats(table);
    scan.setStartRow(HConstants.EMPTY_START_ROW);
    scan.setStopRow(K1);
    List<KeyRange> keyRanges = getSplits(conn, ts, scan);
    assertEquals("Unexpected number of splits: " + keyRanges, 3, keyRanges.size());
    assertEquals(newKeyRange(KeyRange.UNBOUND, new byte[] {'7'}), keyRanges.get(0));
    assertEquals(newKeyRange(new byte[] {'7'}, new byte[] {'M'}), keyRanges.get(1));
    assertEquals(newKeyRange(new byte[] {'M'}, K3), keyRanges.get(2));
}
 

/**
 * Returns the most recent {@link Flow} runs within that time range, up to
 * {@code limit} instances. If the {@code version} parameter is non-null, the
 * returned results will be restricted to those matching this app version.
 *
 * @param cluster the cluster where the jobs were run
 * @param user the user running the jobs
 * @param appId the application identifier for the jobs
 * @param version if non-null, only flows matching this application version
 *          will be returned
 * @param startTime the start time for the flows to be looked at
 * @param endTime the end time for the flows to be looked at
 * @param populateTasks if {@code true}, then TaskDetails will be populated
 *          for each job
 * @param limit the maximum number of flows to return
 * @return
 */
public List<Flow> getFlowSeries(String cluster, String user, String appId,
    String version, boolean populateTasks, long startTime, long endTime,
    int limit) throws IOException {
  // TODO: use RunMatchFilter to limit scan on the server side
  byte[] rowPrefix = Bytes.toBytes(
      cluster + Constants.SEP + user + Constants.SEP + appId + Constants.SEP);
  Scan scan = createFlowScan(rowPrefix, limit, version);

  // set the start and stop rows for scan so that it's time bound
  if (endTime != 0) {
    byte[] scanStartRow;
    // use end time in start row, if present
    long endRunId = FlowKey.encodeRunId(endTime);
    scanStartRow =
        Bytes.add(rowPrefix, Bytes.toBytes(endRunId), Constants.SEP_BYTES);
    scan.setStartRow(scanStartRow);
  }

  if (startTime != 0) {
    byte[] scanStopRow;
    // use start time in stop row, if present
    long stopRunId = FlowKey.encodeRunId(startTime);
    scanStopRow =
        Bytes.add(rowPrefix, Bytes.toBytes(stopRunId), Constants.SEP_BYTES);
    scan.setStopRow(scanStopRow);
  }
  return createFromResults(scan, populateTasks, limit);
}
 

@Override
public RegionScanner preScannerOpen(ObserverContext<RegionCoprocessorEnvironment> e, Scan scan, RegionScanner s)
        throws IOException {
    s = super.preScannerOpen(e, scan, s);
    if (ScanUtil.isAnalyzeTable(scan)) {
        // We are setting the start row and stop row such that it covers the entire region. As part
        // of Phonenix-1263 we are storing the guideposts against the physical table rather than 
        // individual tenant specific tables.
        scan.setStartRow(HConstants.EMPTY_START_ROW);
        scan.setStopRow(HConstants.EMPTY_END_ROW);
        scan.setFilter(null);
    }
    return s;
}
 

/**
 * Creates the scan request.
 * 
 * @param cf
 *          the cf
 * @param cq
 *          the cq
 * @param startKey
 *          the start key
 * @param endKey
 *          the end key
 * @param maxResultSize
 *          the max result size
 * @param startTime
 *          the start time
 * @param endTime
 *          the end time
 * @return the scan
 * @throws IOException
 *           Signals that an I/O exception has occurred.
 */
@VisibleForTesting
Scan createScanRequest(byte[] cf, byte[] cq, String startKey, String endKey,
    long maxResultSize, long startTime, long endTime) throws IOException {
  Scan scan = new Scan();
  scan.addColumn(cf, cq);
  scan.setMaxVersions(ConfigurationUtil.getConfiguration().getInt(
      "hbase.table.column.maxVersions"));
  scan.setStartRow(startKey.getBytes());
  if (endKey != null) {
    scan.setStopRow(endKey.getBytes());
  }
  scan.setMaxResultSize(maxResultSize);
  boolean setTimeRange = true;
  if (startTime < 0 && endTime < 0) {
    setTimeRange = false;
  }
  if (setTimeRange) {
    if (startTime < 0) {
      startTime = 0;
    } else {
      startTime = PcapHelper.convertToDataCreationTimeUnit(startTime);
    }
    if (endTime < 0) {
      endTime = Long.MAX_VALUE;
    } else {
      endTime = PcapHelper.convertToDataCreationTimeUnit(endTime);
    }
    Assert.isTrue(startTime < endTime,
        "startTime value must be less than endTime value");
  }
  // create Scan request;
  if (setTimeRange) {
    scan.setTimeRange(startTime, endTime);
  }
  return scan;
}
 

/**
 * TODO If the required field is null for a row, then this row will not be fetched. That could be a
 * problem for counting Need another version of read to strictly get the number of rows which will return
 * all the columns for a column family
 */
@Override
public void open() throws IOException {
    if (isOpen) {
        return; // silently return
    }
    try {
        tbl = EagleConfigFactory.load().getHTable(schema.getTable());
    } catch (RuntimeException ex) {
        throw new IOException(ex);
    }

    String rowkeyRegex = buildRegex2(searchTags);
    RegexStringComparator regexStringComparator = new RegexStringComparator(rowkeyRegex);
    regexStringComparator.setCharset(Charset.forName("ISO-8859-1"));
    RowFilter filter = new RowFilter(CompareOp.EQUAL, regexStringComparator);
    FilterList filterList = new FilterList();
    filterList.addFilter(filter);
    Scan s1 = new Scan();
    // reverse timestamp, startRow is stopKey, and stopRow is startKey
    s1.setStartRow(stopKey);
    s1.setStopRow(startKey);
    s1.setFilter(filterList);
    // TODO the # of cached rows should be minimum of (pagesize and 100)
    s1.setCaching(100);
    // TODO not optimized for all applications
    s1.setCacheBlocks(true);
    // scan specified columnfamily and qualifiers
    for (byte[] qualifier : qualifiers) {
        s1.addColumn(schema.getColumnFamily().getBytes(), qualifier);
    }
    rs = tbl.getScanner(s1);
    isOpen = true;
}
 

@Test
public void testBuildIndexWhenUserTableAlreadyHasData() throws Exception {
    String tableName = schemaName + "." + generateUniqueName();
    String indexName = "IDX_" + generateUniqueName();
    String indexTableName = schemaName + "." + indexName;
    TableName physicalTableName = SchemaUtil.getPhysicalTableName(tableName.getBytes(), isNamespaceMapped);
    String indexPhysicalTableName = physicalTableName.getNameAsString();

    createBaseTable(tableName, null, "('e','i','o')");
    Connection conn1 = DriverManager.getConnection(getUrl());
    conn1.createStatement().execute("UPSERT INTO "+tableName+" values('b',1,2,4,'z')");
    conn1.createStatement().execute("UPSERT INTO "+tableName+" values('f',1,2,3,'z')");
    conn1.createStatement().execute("UPSERT INTO "+tableName+" values('j',2,4,2,'a')");
    conn1.createStatement().execute("UPSERT INTO "+tableName+" values('q',3,1,1,'c')");
    conn1.commit();
    conn1.createStatement().execute("CREATE LOCAL INDEX " + indexName + " ON " + tableName + "(v1)");
    ResultSet rs = conn1.createStatement().executeQuery("SELECT COUNT(*) FROM " + indexTableName);
    assertTrue(rs.next());
    assertEquals(4, rs.getInt(1));
    Admin admin = driver.getConnectionQueryServices(getUrl(), TestUtil.TEST_PROPERTIES).getAdmin();
    org.apache.hadoop.hbase.client.Connection hbaseConn = admin.getConnection();
    Table indexTable = hbaseConn.getTable(TableName.valueOf(indexPhysicalTableName));
    Pair<byte[][], byte[][]> startEndKeys = hbaseConn.getRegionLocator(TableName.valueOf(indexPhysicalTableName)).getStartEndKeys();
    byte[][] startKeys = startEndKeys.getFirst();
    byte[][] endKeys = startEndKeys.getSecond();
    for (int i = 0; i < startKeys.length; i++) {
        Scan s = new Scan();
        s.addFamily(QueryConstants.DEFAULT_LOCAL_INDEX_COLUMN_FAMILY_BYTES);
        s.setStartRow(startKeys[i]);
        s.setStopRow(endKeys[i]);
        ResultScanner scanner = indexTable.getScanner(s);
        int count = 0;
        for(Result r:scanner){
            count++;
        }
        scanner.close();
        assertEquals(1, count);
    }
    indexTable.close();
}
 

public static void setupReverseScan(Scan scan) {
    if (isReversed(scan)) {
        byte[] startRow = scan.getStartRow();
        byte[] stopRow = scan.getStopRow();
        byte[] newStartRow = startRow;
        byte[] newStopRow = stopRow;
        if (startRow.length != 0) {
            /*
             * Must get previous key because this is going from an inclusive start key to an exclusive stop key, and
             * we need the start key to be included. We get the previous key by decrementing the last byte by one.
             * However, with variable length data types, we need to fill with the max byte value, otherwise, if the
             * start key is 'ab', we lower it to 'aa' which would cause 'aab' to be included (which isn't correct).
             * So we fill with a 0xFF byte to prevent this. A single 0xFF would be enough for our primitive types (as
             * that byte wouldn't occur), but for an arbitrary VARBINARY key we can't know how many bytes to tack
             * on. It's lame of HBase to force us to do this.
             */
            newStartRow = Arrays.copyOf(startRow, startRow.length + MAX_FILL_LENGTH_FOR_PREVIOUS_KEY.length);
            if (ByteUtil.previousKey(newStartRow, startRow.length)) {
                System.arraycopy(MAX_FILL_LENGTH_FOR_PREVIOUS_KEY, 0, newStartRow, startRow.length, MAX_FILL_LENGTH_FOR_PREVIOUS_KEY.length);
            } else {
                newStartRow = HConstants.EMPTY_START_ROW;
            }
        }
        if (stopRow.length != 0) {
            // Must add null byte because we need the start to be exclusive while it was inclusive
            newStopRow = ByteUtil.concat(stopRow, QueryConstants.SEPARATOR_BYTE_ARRAY);
        }
        scan.setStartRow(newStopRow);
        scan.setStopRow(newStartRow);
        scan.setReversed(true);
    }
}
 

List<WorkflowInstanceExecutionData> executionData(WorkflowId workflowId, String offset, int limit)
    throws IOException {
  try (final Table eventsTable = connection.getTable(EVENTS_TABLE_NAME)) {
    final Scan scan = new Scan()
        .setRowPrefixFilter(Bytes.toBytes(workflowId.toKey() + '#'))
        .setFilter(new FirstKeyOnlyFilter());

    if (!Strings.isNullOrEmpty(offset)) {
      final WorkflowInstance offsetInstance = WorkflowInstance.create(workflowId, offset);
      scan.setStartRow(Bytes.toBytes(offsetInstance.toKey() + '#'));
    }

    final Set<WorkflowInstance> workflowInstancesSet = Sets.newHashSet();
    try (ResultScanner scanner = eventsTable.getScanner(scan)) {
      Result result = scanner.next();
      while (result != null) {
        final String key = new String(result.getRow());
        final int lastHash = key.lastIndexOf('#');

        final WorkflowInstance wfi = WorkflowInstance.parseKey(key.substring(0, lastHash));
        workflowInstancesSet.add(wfi);
        if (workflowInstancesSet.size() == limit) {
          break;
        }

        result = scanner.next();
      }
    }

    return executionData(workflowInstancesSet);
  }
}
 

private Scan createScanForInitialAgentInfo(String agentId) {
    Scan scan = new Scan();
    byte[] agentIdBytes = Bytes.toBytes(agentId);
    byte[] reverseStartKey = RowKeyUtils.concatFixedByteAndLong(agentIdBytes, HbaseTableConstatns.AGENT_NAME_MAX_LEN, Long.MAX_VALUE);
    scan.setStartRow(reverseStartKey);
    scan.setReversed(true);
    scan.setMaxVersions(1);
    scan.setCaching(SCANNER_CACHING);
    return scan;
}
 

@SuppressWarnings("unused")
public static void main(String[] args) throws IOException {
  Configuration conf = HBaseConfiguration.create();
  String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();
  if (otherArgs.length < 2) {
    System.err.println("Usage: java -cp <this jar>:<hbase classpath> "
        + ValidateWordCount.class.getName() + " <table-name> <expected count>");
    System.exit(2);
  }

  TableName tableName = TableName.valueOf(otherArgs[0]);
  int expectedCount = Integer.parseInt(otherArgs[1]);

  Scan scan = new Scan();
  scan.addFamily(Bytes.toBytes("cf"));
  scan.setStartRow(Bytes.toBytes(""));
  int count = 0;

  try (Connection conn = ConnectionFactory.createConnection(conf);
      Table table = conn.getTable(tableName);
      ResultScanner rs = table.getScanner(scan)) {
    for (Result result : rs) {
      count++;
    }
  }

  System.out.println("Count: " + count + ".  Expected: " + expectedCount);
  System.exit(count == expectedCount ? 0 : 1);
}
 

/**
 * 初始化scan集合
 * 
 * @param job
 * @return
 */
private List<Scan> initScans(Job job) {
    Configuration conf = job.getConfiguration();
    // 获取运行时间: yyyy-MM-dd
    String date = conf.get(GlobalConstants.RUNNING_DATE_PARAMES);
    long startDate = TimeUtil.parseString2Long(date);
    long endDate = startDate + GlobalConstants.DAY_OF_MILLISECONDS;

    Scan scan = new Scan();
    // 定义hbase扫描的开始rowkey和结束rowkey
    scan.setStartRow(Bytes.toBytes("" + startDate));
    scan.setStopRow(Bytes.toBytes("" + endDate));

    FilterList filterList = new FilterList();
    // 定义mapper中需要获取的列名
    String[] columns = new String[] { 
    		EventLogConstants.LOG_COLUMN_NAME_UUID, // 用户id
            EventLogConstants.LOG_COLUMN_NAME_SERVER_TIME, // 服务器时间
            EventLogConstants.LOG_COLUMN_NAME_PLATFORM, // 平台名称
            EventLogConstants.LOG_COLUMN_NAME_BROWSER_NAME, // 浏览器名称
            EventLogConstants.LOG_COLUMN_NAME_BROWSER_VERSION // 浏览器版本号
    };
    filterList.addFilter(this.getColumnFilter(columns));

    scan.setAttribute(Scan.SCAN_ATTRIBUTES_TABLE_NAME, Bytes.toBytes(EventLogConstants.HBASE_NAME_EVENT_LOGS));
    scan.setFilter(filterList);
    return Lists.newArrayList(scan);
}
 

/**
 * Get parallel scans of the specified scan boundaries. This can be used for getting parallel
 * scans when there is split/merges while scanning a chunk. In this case we need not go by all
 * the regions or guideposts.
 * @param scan
 * @return
 * @throws SQLException
 */
private List<List<Scan>> getParallelScans(Scan scan) throws SQLException {
    List<HRegionLocation> regionLocations = getRegionBoundaries(scanGrouper);
    List<byte[]> regionBoundaries = toBoundaries(regionLocations);
    int regionIndex = 0;
    int stopIndex = regionBoundaries.size();
    if (scan.getStartRow().length > 0) {
        regionIndex = getIndexContainingInclusive(regionBoundaries, scan.getStartRow());
    }
    if (scan.getStopRow().length > 0) {
        stopIndex = Math.min(stopIndex, regionIndex + getIndexContainingExclusive(regionBoundaries.subList(regionIndex, stopIndex), scan.getStopRow()));
    }
    List<List<Scan>> parallelScans = Lists.newArrayListWithExpectedSize(stopIndex - regionIndex + 1);
    List<Scan> scans = Lists.newArrayListWithExpectedSize(2);
    while (regionIndex <= stopIndex) {
        HRegionLocation regionLocation = regionLocations.get(regionIndex);
        RegionInfo regionInfo = regionLocation.getRegion();
        Scan newScan = ScanUtil.newScan(scan);
        byte[] endKey;
        if (regionIndex == stopIndex) {
            endKey = scan.getStopRow();
        } else {
            endKey = regionBoundaries.get(regionIndex);
        }
        if(ScanUtil.isLocalIndex(scan)) {
            ScanUtil.setLocalIndexAttributes(newScan, 0, regionInfo.getStartKey(),
                regionInfo.getEndKey(), newScan.getAttribute(SCAN_START_ROW_SUFFIX),
                newScan.getAttribute(SCAN_STOP_ROW_SUFFIX));
        } else {
            if(Bytes.compareTo(scan.getStartRow(), regionInfo.getStartKey())<=0) {
                newScan.setAttribute(SCAN_ACTUAL_START_ROW, regionInfo.getStartKey());
                newScan.setStartRow(regionInfo.getStartKey());
            }
            if(scan.getStopRow().length == 0 || (regionInfo.getEndKey().length != 0 && Bytes.compareTo(scan.getStopRow(), regionInfo.getEndKey())>0)) {
                newScan.setStopRow(regionInfo.getEndKey());
            }
        }
        scans = addNewScan(parallelScans, scans, newScan, endKey, true, regionLocation);
        regionIndex++;
    }
    if (!scans.isEmpty()) { // Add any remaining scans
        parallelScans.add(scans);
    }
    return parallelScans;
}
 

public SlicedRowFilterGTSDecoderIterator(long now, long timespan, List<Metadata> metadatas, Connection conn, TableName tableName, byte[] colfam, boolean writeTimestamp, KeyStore keystore, boolean useBlockCache) {
    
  this.keystore = keystore;
  this.now = now;
  this.timespan = timespan;
  this.hbaseAESKey = keystore.getKey(KeyStore.AES_HBASE_DATA);
  this.writeTimestamp = writeTimestamp;
  
  //
  // Check that if 'timespan' is < 0 then 'now' is either Long.MAX_VALUE or congruent to 0 modulo DEFAULT_MODULUS
  //
  
  if (timespan < 0) {
    if (Long.MAX_VALUE != now && 0 != (now % Constants.DEFAULT_MODULUS)) {
      throw new RuntimeException("Incompatible 'timespan' (" + timespan + ") and 'now' (" + now + ")");
    }
  }
  
  //
  // Create a SlicedRowFilter for the prefix, class id, labels id and ts
  // We include the prefix so we exit the filter early when the last
  // matching row has been reached
  //
  
  // 128BITS
  
  int[] bounds = { 0, 24 };
  
  //
  // Create singleton for each classId/labelsId combo
  //
  // TODO(hbs): we should really create multiple scanner, one per class Id for example,
  // 
  
  List<Pair<byte[], byte[]>> ranges = new ArrayList<Pair<byte[], byte[]>>();
  
  for (Metadata metadata: metadatas) {
    byte[][] keys = getKeys(metadata, now, timespan);
    byte[] lower = keys[0];
    byte[] upper = keys[1];
    
    this.metadatas.put(new String(Arrays.copyOfRange(lower, prefix.length, prefix.length + 16), StandardCharsets.ISO_8859_1), metadata);
    
    Pair<byte[],byte[]> range = new Pair<byte[],byte[]>(lower, upper);
    
    ranges.add(range);
  }
              
  SlicedRowFilter filter = new SlicedRowFilter(bounds, ranges, timespan < 0 ? -timespan : Long.MAX_VALUE);

  //
  // Create scanner. The start key is the lower bound of the first range
  //
  
  Scan scan = new Scan();
  scan.addFamily(colfam); // (HBaseStore.GTS_COLFAM, Longs.toByteArray(Long.MAX_VALUE - modulus));
  scan.setStartRow(filter.getStartKey());
  byte[] filterStopKey = filter.getStopKey();
  // Add one byte at the end (we can do that because we know the slice is the whole key)
  byte[] stopRow = Arrays.copyOf(filterStopKey, filterStopKey.length + 1);
  scan.setStopRow(stopRow);
  scan.setFilter(filter);
  
  scan.setMaxResultSize(1000000L);
  scan.setBatch(50000);
  scan.setCaching(50000);
  
  scan.setCacheBlocks(useBlockCache);

  Sensision.update(SensisionConstants.SENSISION_CLASS_CONTINUUM_HBASE_CLIENT_FILTERED_SCANNERS, Sensision.EMPTY_LABELS, 1);
  Sensision.update(SensisionConstants.SENSISION_CLASS_CONTINUUM_HBASE_CLIENT_FILTERED_SCANNERS_RANGES, Sensision.EMPTY_LABELS, ranges.size());

  try {
    this.htable = conn.getTable(tableName);
    this.scanner = this.htable.getScanner(scan);
    iter = scanner.iterator();          
  } catch (IOException ioe) {
    LOG.error("",ioe);
    this.iter = null;
  }
}
 

public static void main(String[] args) throws IOException {

        if (args == null || args.length != 3) {
            logger.info(
                    "Usage: hbase org.apache.hadoop.util.RunJar kylin-job-latest.jar org.apache.kylin.job.tools.RowCounterCLI [HTABLE_NAME] [STARTKEY] [ENDKEY]");
            return; // if no enough arguments provided, return with above message
        }

        logger.info(args[0]);
        String htableName = args[0];
        logger.info(args[1]);
        byte[] startKey = BytesUtil.fromReadableText(args[1]);
        logger.info(args[2]);
        byte[] endKey = BytesUtil.fromReadableText(args[2]);

        if (startKey == null) {
            logger.info("startkey is null ");
        } else {
            logger.info("startkey lenght: {}", startKey.length);
        }
        if(logger.isInfoEnabled()){
            logger.info("start key in binary: {}", Bytes.toStringBinary(startKey));
            logger.info("end key in binary: {}", Bytes.toStringBinary(endKey));
        }

        Configuration conf = HBaseConnection.getCurrentHBaseConfiguration();

        Scan scan = new Scan();
        scan.setCaching(512);
        scan.setCacheBlocks(true);
        scan.setStartRow(startKey);
        scan.setStopRow(endKey);

        logger.info("My Scan {}", scan);
        try (Connection conn = ConnectionFactory.createConnection(conf);
                Table tableInterface = conn.getTable(TableName.valueOf(htableName))) {
            Iterator<Result> iterator = tableInterface.getScanner(scan).iterator();
            int counter = 0;
            while (iterator.hasNext()) {
                iterator.next();
                counter++;
                if (counter % 1000 == 1) {
                    logger.info("number of rows: {}", counter);
                }
            }
            logger.info("number of rows: {}", counter);
        }
    }
 

/**
 * Setup scan condition if necessary.
 * 
 * @param conf
 * @param line
 * @throws IOException
 */
public static void setScanIfNecessary(Configuration conf, CommandLine line) throws IOException {
	String startRow = line.getOptionValue("startRow");
	String endRow = line.getOptionValue("endRow");
	String startTime = line.getOptionValue("startTime");
	String endTime = line.getOptionValue("endTime");

	boolean enabledScan = false;
	Scan scan = new Scan();
	// Row
	if (isNotBlank(startRow)) {
		conf.set(TableInputFormat.SCAN_ROW_START, startRow);
		scan.setStartRow(Bytes.toBytes(startRow));
		enabledScan = true;
	}
	if (isNotBlank(endRow)) {
		Assert2.hasText(startRow, "Argument for startRow and endRow are used simultaneously");
		conf.set(TableInputFormat.SCAN_ROW_STOP, endRow);
		scan.setStopRow(Bytes.toBytes(endRow));
		enabledScan = true;
	}

	// Row TimeStamp
	if (isNotBlank(startTime) && isNotBlank(endTime)) {
		conf.set(TableInputFormat.SCAN_TIMERANGE_START, startTime);
		conf.set(TableInputFormat.SCAN_TIMERANGE_END, endTime);
		try {
			Timestamp stime = new Timestamp(Long.parseLong(startTime));
			Timestamp etime = new Timestamp(Long.parseLong(endTime));
			scan.setTimeRange(stime.getTime(), etime.getTime());
			enabledScan = true;
		} catch (Exception e) {
			throw new IllegalArgumentException(String.format("Illegal startTime(%s) and endTime(%s)", startTime, endTime), e);
		}
	}

	if (enabledScan) {
		ClientProtos.Scan proto = ProtobufUtil.toScan(scan);
		log.info("All other SCAN configuration are ignored if\n"
				+ "		 * this is specified.See TableMapReduceUtil.convertScanToString(Scan)\n"
				+ "		 * for more details.");
		conf.set(TableInputFormat.SCAN, Base64.encodeBytes(proto.toByteArray()));
	}
}
 

private static void jumpScan(Connection conn, boolean[] hits, Stats stats) throws IOException {

        final int jumpThreshold = 6; // compensate for Scan() overhead, totally by experience

        Table table = conn.getTable(TableName.valueOf(TEST_TABLE));
        try {

            stats.markStart();

            int i = 0;
            while (i < N_ROWS) {
                // find the first hit
                int start = i;
                while (start + 1 < N_ROWS && !hits[start]) start++;

                // find the last hit within jumpThreshold
                int end = start + 1;
                int jump = end + 1;
                while (jump < N_ROWS && (end + jumpThreshold > jump)) {
                    if (hits[jump]) {
                        end = jump;
                    }
                    jump++;
                }

                if (start < N_ROWS) {
                    Scan scan = new Scan();
                    scan.setStartRow(Bytes.toBytes(start));
                    scan.setStopRow(Bytes.toBytes(end));
                    scan.addFamily(CF);
                    ResultScanner scanner = table.getScanner(scan);
                    i = start;
                    for (Result r : scanner) {
                        stats.consume(r);
                        dot(i, N_ROWS);
                        i++;
                    }
                }
                i = end;
            }

            stats.markEnd();

        } finally {
            IOUtils.closeQuietly(table);
        }
    }