HBASE-476 RegexpRowFilter behaves incorectly when there are multiple store files (Clint Morgan via Jim Kellerman)

HBASE-527   RegexpRowFilter does not work when there are columns from multiple families (Clint Morgan via Jim Kellerman)


git-svn-id: https://svn.apache.org/repos/asf/hadoop/hbase/trunk@640106 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Jim Kellerman 2008-03-22 22:35:36 +00:00
parent e7b705a808
commit aabe9f09a9
12 changed files with 389 additions and 31 deletions

View File

@ -50,6 +50,10 @@ Hbase Change Log
HBASE-531 Merge tool won't merge two overlapping regions (port HBASE-483 to HBASE-531 Merge tool won't merge two overlapping regions (port HBASE-483 to
trunk) trunk)
HBASE-537 Wait for hdfs to exit safe mode HBASE-537 Wait for hdfs to exit safe mode
HBASE-476 RegexpRowFilter behaves incorectly when there are multiple store
files (Clint Morgan via Jim Kellerman)
HBASE-527 RegexpRowFilter does not work when there are columns from
multiple families (Clint Morgan via Jim Kellerman)
IMPROVEMENTS IMPROVEMENTS
HBASE-415 Rewrite leases to use DelayedBlockingQueue instead of polling HBASE-415 Rewrite leases to use DelayedBlockingQueue instead of polling

View File

@ -19,17 +19,9 @@
*/ */
package org.apache.hadoop.hbase.filter; package org.apache.hadoop.hbase.filter;
import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import java.util.TreeMap;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.io.Text; import org.apache.hadoop.io.Text;
/**
/*
* Subclass of StopRowFilter that filters rows > the stop row, * Subclass of StopRowFilter that filters rows > the stop row,
* making it include up to the last row but no further. * making it include up to the last row but no further.
*/ */
@ -49,6 +41,8 @@ public class InclusiveStopRowFilter extends StopRowFilter{
super(stopRowKey); super(stopRowKey);
} }
/** {@inheritDoc} */
@Override
public boolean filter(final Text rowKey) { public boolean filter(final Text rowKey) {
if (rowKey == null) { if (rowKey == null) {
if (this.stopRowKey == null) { if (this.stopRowKey == null) {

View File

@ -22,7 +22,7 @@ package org.apache.hadoop.hbase.filter;
import java.io.DataInput; import java.io.DataInput;
import java.io.DataOutput; import java.io.DataOutput;
import java.io.IOException; import java.io.IOException;
import java.util.TreeMap; import java.util.SortedMap;
import org.apache.commons.logging.Log; import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.LogFactory;
@ -137,7 +137,7 @@ public class PageRowFilter implements RowFilterInterface {
* {@inheritDoc} * {@inheritDoc}
*/ */
public boolean filterNotNull(@SuppressWarnings("unused") public boolean filterNotNull(@SuppressWarnings("unused")
final TreeMap<Text, byte[]> columns) { final SortedMap<Text, byte[]> columns) {
return filterAllRemaining(); return filterAllRemaining();
} }

View File

@ -27,7 +27,7 @@ import java.util.HashMap;
import java.util.HashSet; import java.util.HashSet;
import java.util.Map; import java.util.Map;
import java.util.Set; import java.util.Set;
import java.util.TreeMap; import java.util.SortedMap;
import java.util.Map.Entry; import java.util.Map.Entry;
import java.util.regex.Pattern; import java.util.regex.Pattern;
@ -196,7 +196,7 @@ public class RegExpRowFilter implements RowFilterInterface {
* *
* {@inheritDoc} * {@inheritDoc}
*/ */
public boolean filterNotNull(final TreeMap<Text, byte[]> columns) { public boolean filterNotNull(final SortedMap<Text, byte[]> columns) {
for (Entry<Text, byte[]> col : columns.entrySet()) { for (Entry<Text, byte[]> col : columns.entrySet()) {
if (nullColumns.contains(col.getKey()) if (nullColumns.contains(col.getKey())
&& !HLogEdit.isDeleted(col.getValue())) { && !HLogEdit.isDeleted(col.getValue())) {
@ -211,7 +211,7 @@ public class RegExpRowFilter implements RowFilterInterface {
if (!columns.containsKey(col)) { if (!columns.containsKey(col)) {
if (LOG.isDebugEnabled()) { if (LOG.isDebugEnabled()) {
LOG.debug("filterNotNull returning true for colKey: " + col + LOG.debug("filterNotNull returning true for colKey: " + col +
", column not found in given TreeMap<Text, byte[]>."); ", column not found in given SortedMap<Text, byte[]>.");
} }
return true; return true;
} }

View File

@ -19,7 +19,7 @@
*/ */
package org.apache.hadoop.hbase.filter; package org.apache.hadoop.hbase.filter;
import java.util.TreeMap; import java.util.SortedMap;
import org.apache.hadoop.io.Text; import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.Writable; import org.apache.hadoop.io.Writable;
@ -108,7 +108,7 @@ public interface RowFilterInterface extends Writable {
* @param columns * @param columns
* @return true if null/non-null criteria not met. * @return true if null/non-null criteria not met.
*/ */
boolean filterNotNull(final TreeMap<Text, byte[]> columns); boolean filterNotNull(final SortedMap<Text, byte[]> columns);
/** /**
* Validates that this filter applies only to a subset of the given columns. * Validates that this filter applies only to a subset of the given columns.

View File

@ -24,7 +24,7 @@ import java.io.DataOutput;
import java.io.IOException; import java.io.IOException;
import java.util.HashSet; import java.util.HashSet;
import java.util.Set; import java.util.Set;
import java.util.TreeMap; import java.util.SortedMap;
import org.apache.commons.logging.Log; import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.LogFactory;
@ -232,7 +232,7 @@ public class RowFilterSet implements RowFilterInterface {
} }
/** {@inheritDoc} */ /** {@inheritDoc} */
public boolean filterNotNull(final TreeMap<Text, byte[]> columns) { public boolean filterNotNull(final SortedMap<Text, byte[]> columns) {
boolean resultFound = false; boolean resultFound = false;
boolean result = operator == Operator.MUST_PASS_ONE; boolean result = operator == Operator.MUST_PASS_ONE;
for (RowFilterInterface filter : filters) { for (RowFilterInterface filter : filters) {

View File

@ -22,7 +22,7 @@ package org.apache.hadoop.hbase.filter;
import java.io.DataInput; import java.io.DataInput;
import java.io.DataOutput; import java.io.DataOutput;
import java.io.IOException; import java.io.IOException;
import java.util.TreeMap; import java.util.SortedMap;
import org.apache.commons.logging.Log; import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.LogFactory;
@ -132,7 +132,7 @@ public class StopRowFilter implements RowFilterInterface {
* @param columns * @param columns
*/ */
public boolean filterNotNull(@SuppressWarnings("unused") public boolean filterNotNull(@SuppressWarnings("unused")
final TreeMap<Text, byte[]> columns) { final SortedMap<Text, byte[]> columns) {
return filterAllRemaining(); return filterAllRemaining();
} }

View File

@ -22,7 +22,7 @@ package org.apache.hadoop.hbase.filter;
import java.io.DataInput; import java.io.DataInput;
import java.io.DataOutput; import java.io.DataOutput;
import java.io.IOException; import java.io.IOException;
import java.util.TreeMap; import java.util.SortedMap;
import org.apache.commons.logging.Log; import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.LogFactory;
@ -31,7 +31,7 @@ import org.apache.hadoop.io.Text;
/** /**
* WhileMatchRowFilter is a wrapper filter that filters everything after the * WhileMatchRowFilter is a wrapper filter that filters everything after the
* first filtered row. Once the nested filter returns true for either of it's * first filtered row. Once the nested filter returns true for either of it's
* filter(..) methods or filterNotNull(TreeMap<Text, byte[]>), this wrapper's * filter(..) methods or filterNotNull(SortedMap<Text, byte[]>), this wrapper's
* filterAllRemaining() will return true. All filtering methods will * filterAllRemaining() will return true. All filtering methods will
* thereafter defer to the result of filterAllRemaining(). * thereafter defer to the result of filterAllRemaining().
*/ */
@ -115,7 +115,7 @@ public class WhileMatchRowFilter implements RowFilterInterface {
} }
/** {@inheritDoc} */ /** {@inheritDoc} */
public boolean filterNotNull(final TreeMap<Text, byte[]> columns) { public boolean filterNotNull(final SortedMap<Text, byte[]> columns) {
changeFAR(this.filter.filterNotNull(columns)); changeFAR(this.filter.filterNotNull(columns));
boolean result = filterAllRemaining(); boolean result = filterAllRemaining();
if (LOG.isDebugEnabled()) { if (LOG.isDebugEnabled()) {

View File

@ -1721,12 +1721,14 @@ public class HRegion implements HConstants {
private HInternalScannerInterface[] scanners; private HInternalScannerInterface[] scanners;
private TreeMap<Text, byte []>[] resultSets; private TreeMap<Text, byte []>[] resultSets;
private HStoreKey[] keys; private HStoreKey[] keys;
private RowFilterInterface filter;
/** Create an HScanner with a handle on many HStores. */ /** Create an HScanner with a handle on many HStores. */
@SuppressWarnings("unchecked") @SuppressWarnings("unchecked")
HScanner(Text[] cols, Text firstRow, long timestamp, HStore[] stores, HScanner(Text[] cols, Text firstRow, long timestamp, HStore[] stores,
RowFilterInterface filter) RowFilterInterface filter)
throws IOException { throws IOException {
this.filter = filter;
this.scanners = new HInternalScannerInterface[stores.length]; this.scanners = new HInternalScannerInterface[stores.length];
try { try {
for (int i = 0; i < stores.length; i++) { for (int i = 0; i < stores.length; i++) {
@ -1737,8 +1739,8 @@ public class HRegion implements HConstants {
// At least WhileMatchRowFilter will mess up the scan if only // At least WhileMatchRowFilter will mess up the scan if only
// one shared across many rows. See HADOOP-2467. // one shared across many rows. See HADOOP-2467.
scanners[i] = stores[i].getScanner(timestamp, cols, firstRow, scanners[i] = stores[i].getScanner(timestamp, cols, firstRow,
(i > 0 && filter != null)? filter != null ?
(RowFilterInterface)WritableUtils.clone(filter, conf): filter); (RowFilterInterface)WritableUtils.clone(filter, conf) : filter);
} }
} catch(IOException e) { } catch(IOException e) {
for (int i = 0; i < this.scanners.length; i++) { for (int i = 0; i < this.scanners.length; i++) {
@ -1839,6 +1841,12 @@ public class HRegion implements HConstants {
} }
} }
} }
if (filter != null && filter.filterNotNull(results)) {
LOG.warn("Filter return true on assembled Results in hstore");
return moreToFollow == true && this.next(key, results);
}
return moreToFollow; return moreToFollow;
} }

View File

@ -32,7 +32,6 @@ import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory; import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.io.Text; import org.apache.hadoop.io.Text;
import org.apache.hadoop.hbase.HStoreKey; import org.apache.hadoop.hbase.HStoreKey;
import org.apache.hadoop.hbase.io.ImmutableBytesWritable;
import org.apache.hadoop.hbase.filter.RowFilterInterface; import org.apache.hadoop.hbase.filter.RowFilterInterface;
import org.apache.hadoop.hbase.HConstants; import org.apache.hadoop.hbase.HConstants;
@ -167,11 +166,6 @@ class HStoreScanner implements HInternalScannerInterface {
break; break;
} }
// Filter out null criteria columns that are not null
if (dataFilter != null) {
filtered = dataFilter.filterNotNull(resultSets[i]);
}
// NOTE: We used to do results.putAll(resultSets[i]); // NOTE: We used to do results.putAll(resultSets[i]);
// but this had the effect of overwriting newer // but this had the effect of overwriting newer
// values with older ones. So now we only insert // values with older ones. So now we only insert

View File

@ -0,0 +1,225 @@
/**
* Copyright 2008 The Apache Software Foundation
*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.hbase.filter;
import java.io.IOException;
import java.util.Arrays;
import java.util.HashMap;
import java.util.Map;
import java.util.TreeMap;
import junit.framework.Assert;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.hbase.HBaseClusterTestCase;
import org.apache.hadoop.hbase.HColumnDescriptor;
import org.apache.hadoop.hbase.HConstants;
import org.apache.hadoop.hbase.HScannerInterface;
import org.apache.hadoop.hbase.HStoreKey;
import org.apache.hadoop.hbase.HTableDescriptor;
import org.apache.hadoop.hbase.client.HBaseAdmin;
import org.apache.hadoop.hbase.client.HTable;
import org.apache.hadoop.hbase.io.BatchUpdate;
import org.apache.hadoop.io.Text;
/** Test regexp filters HBASE-476 */
public class TestRowFilterAfterWrite extends HBaseClusterTestCase {
@SuppressWarnings("hiding")
private static final Log LOG = LogFactory.getLog(TestRowFilterAfterWrite.class.getName());
static final String TABLE_NAME = "TestTable";
static final String FAMILY = "C:";
static final String COLUMN1 = FAMILY + "col1";
static final Text TEXT_COLUMN1 = new Text(COLUMN1);
static final String COLUMN2 = FAMILY + "col2";
static final Text TEXT_COLUMN2 = new Text(COLUMN2);
private static final Text[] columns = {
TEXT_COLUMN1, TEXT_COLUMN2
};
private static final int NUM_ROWS = 10;
private static final int VALUE_SIZE = 1000;
private static final byte[] VALUE = new byte[VALUE_SIZE];
private static final int COL_2_SIZE = 5;
private static final int KEY_SIZE = 9;
private static final int NUM_REWRITES = 10;
static {
Arrays.fill(VALUE, (byte) 'a');
}
/** constructor */
public TestRowFilterAfterWrite() {
super();
// Make sure the cache gets flushed so we get multiple stores
conf.setInt("hbase.hregion.memcache.flush.size", (NUM_ROWS * (VALUE_SIZE + COL_2_SIZE + KEY_SIZE)));
LOG.info("memcach flush : " + conf.get("hbase.hregion.memcache.flush.size"));
conf.setInt("hbase.regionserver.optionalcacheflushinterval", 100000000);
// Avoid compaction to keep multiple stores
conf.setInt("hbase.hstore.compactionThreshold", 10000);
// Make lease timeout longer, lease checks less frequent
conf.setInt("hbase.master.lease.period", 10 * 1000);
conf.setInt("hbase.master.lease.thread.wakefrequency", 5 * 1000);
// For debugging
conf.setInt("hbase.regionserver.lease.period", 20 * 60 * 1000);
conf.setInt("ipc.client.timeout", 20 * 60 * 1000);
}
/**
* {@inheritDoc}
*/
@Override
public void setUp() throws Exception {
// this.conf.set(HConstants.HBASE_DIR, "file:///opt/benchmark/hadoop/hbase");
this.conf.set(HConstants.MASTER_ADDRESS, "0.0.0.0:60100");
// Must call super.setup() after starting mini dfs cluster. Otherwise
// we get a local file system instead of hdfs
super.setUp();
}
/**
* {@inheritDoc}
*/
@Override
public void tearDown() throws Exception {
super.tearDown();
}
/**
* Test hbase mapreduce jobs against single region and multi-region tables.
*
* @throws IOException
* @throws InterruptedException
*/
public void testAfterWrite() throws IOException, InterruptedException {
singleTableTest();
}
/*
* Test against a single region. @throws IOException
*/
private void singleTableTest() throws IOException, InterruptedException {
HTableDescriptor desc = new HTableDescriptor(TABLE_NAME);
desc.addFamily(new HColumnDescriptor(FAMILY));
// Create a table.
HBaseAdmin admin = new HBaseAdmin(this.conf);
admin.createTable(desc);
// insert some data into the test table
HTable table = new HTable(conf, new Text(TABLE_NAME));
for (int i = 0; i < NUM_ROWS; i++) {
BatchUpdate b =
new BatchUpdate(new Text("row_" + String.format("%1$05d", i)));
b.put(TEXT_COLUMN1, VALUE);
b.put(TEXT_COLUMN2, String.format("%1$05d", i).getBytes());
table.commit(b);
}
// LOG.info("Print table contents using scanner before map/reduce for " + TABLE_NAME);
// scanTable(TABLE_NAME, false);
// LOG.info("Print table contents using scanner+filter before map/reduce for " + TABLE_NAME);
// scanTableWithRowFilter(TABLE_NAME, false);
// Do some identity write operations on one column of the data.
for (int n = 0; n < NUM_REWRITES; n++) {
for (int i = 0; i < NUM_ROWS; i++) {
BatchUpdate b =
new BatchUpdate(new Text("row_" + String.format("%1$05d", i)));
b.put(TEXT_COLUMN2, String.format("%1$05d", i).getBytes());
table.commit(b);
}
}
// Wait for the flush to happen
LOG.info("Waiting, for flushes to complete");
Thread.sleep(5 * 1000);
// Wait for the flush to happen
LOG.info("Done. No flush should happen after this");
// Do another round so to populate the mem cache
for (int i = 0; i < NUM_ROWS; i++) {
BatchUpdate b =
new BatchUpdate(new Text("row_" + String.format("%1$05d", i)));
b.put(TEXT_COLUMN2, String.format("%1$05d", i).getBytes());
table.commit(b);
}
LOG.info("Print table contents using scanner after map/reduce for " + TABLE_NAME);
scanTable(TABLE_NAME, true);
LOG.info("Print table contents using scanner+filter after map/reduce for " + TABLE_NAME);
scanTableWithRowFilter(TABLE_NAME, true);
}
private void scanTable(final String tableName, final boolean printValues) throws IOException {
HTable table = new HTable(conf, new Text(tableName));
HScannerInterface scanner = table.obtainScanner(columns, HConstants.EMPTY_START_ROW);
int numFound = doScan(scanner, printValues);
Assert.assertEquals(NUM_ROWS, numFound);
}
private void scanTableWithRowFilter(final String tableName, final boolean printValues) throws IOException {
HTable table = new HTable(conf, new Text(tableName));
Map<Text, byte[]> columnMap = new HashMap<Text, byte[]>();
columnMap.put(TEXT_COLUMN1, VALUE);
RegExpRowFilter filter = new RegExpRowFilter(null, columnMap);
HScannerInterface scanner = table.obtainScanner(columns, HConstants.EMPTY_START_ROW, filter);
int numFound = doScan(scanner, printValues);
Assert.assertEquals(NUM_ROWS, numFound);
}
private int doScan(final HScannerInterface scanner, final boolean printValues) throws IOException {
{
int count = 0;
try {
HStoreKey key = new HStoreKey();
TreeMap<Text, byte[]> results = new TreeMap<Text, byte[]>();
while (scanner.next(key, results)) {
if (printValues) {
LOG.info("row: " + key.getRow());
for (Map.Entry<Text, byte[]> e : results.entrySet()) {
LOG.info(" column: " + e.getKey() + " value: "
+ new String(e.getValue(), HConstants.UTF8_ENCODING));
}
}
count++;
}
} finally {
scanner.close();
}
return count;
}
}
}

View File

@ -0,0 +1,133 @@
/**
* Copyright 2008 The Apache Software Foundation
*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.hbase.filter;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import java.util.TreeMap;
import junit.framework.Assert;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.hbase.HBaseClusterTestCase;
import org.apache.hadoop.hbase.HColumnDescriptor;
import org.apache.hadoop.hbase.HConstants;
import org.apache.hadoop.hbase.HScannerInterface;
import org.apache.hadoop.hbase.HStoreKey;
import org.apache.hadoop.hbase.HTableDescriptor;
import org.apache.hadoop.hbase.client.HBaseAdmin;
import org.apache.hadoop.hbase.client.HTable;
import org.apache.hadoop.hbase.io.BatchUpdate;
import org.apache.hadoop.io.Text;
/**
* Test for regexp filters (HBASE-527)
*/
public class TestRowFilterOnMultipleFamilies extends HBaseClusterTestCase {
private static final Log LOG = LogFactory.getLog(TestRowFilterOnMultipleFamilies.class.getName());
static final String TABLE_NAME = "TestTable";
static final String COLUMN1 = "A:col1";
static final Text TEXT_COLUMN1 = new Text(COLUMN1);
static final String COLUMN2 = "B:col2";
static final Text TEXT_COLUMN2 = new Text(COLUMN2);
private static final Text[] columns = {
TEXT_COLUMN1, TEXT_COLUMN2
};
private static final int NUM_ROWS = 10;
private static final byte[] VALUE = "HELLO".getBytes();
/** @throws IOException */
public void testMultipleFamilies() throws IOException {
HTableDescriptor desc = new HTableDescriptor(TABLE_NAME);
desc.addFamily(new HColumnDescriptor("A:"));
desc.addFamily(new HColumnDescriptor("B:"));
// Create a table.
HBaseAdmin admin = new HBaseAdmin(this.conf);
admin.createTable(desc);
// insert some data into the test table
HTable table = new HTable(conf, new Text(TABLE_NAME));
for (int i = 0; i < NUM_ROWS; i++) {
BatchUpdate b =
new BatchUpdate(new Text("row_" + String.format("%1$05d", i)));
b.put(TEXT_COLUMN1, VALUE);
b.put(TEXT_COLUMN2, String.format("%1$05d", i).getBytes());
table.commit(b);
}
LOG.info("Print table contents using scanner before map/reduce for " + TABLE_NAME);
scanTable(TABLE_NAME, true);
LOG.info("Print table contents using scanner+filter before map/reduce for " + TABLE_NAME);
scanTableWithRowFilter(TABLE_NAME, true);
}
private void scanTable(final String tableName, final boolean printValues) throws IOException {
HTable table = new HTable(conf, new Text(tableName));
HScannerInterface scanner = table.obtainScanner(columns, HConstants.EMPTY_START_ROW);
int numFound = doScan(scanner, printValues);
Assert.assertEquals(NUM_ROWS, numFound);
}
private void scanTableWithRowFilter(final String tableName, final boolean printValues) throws IOException {
HTable table = new HTable(conf, new Text(tableName));
Map<Text, byte[]> columnMap = new HashMap<Text, byte[]>();
columnMap.put(TEXT_COLUMN1, VALUE);
RegExpRowFilter filter = new RegExpRowFilter(null, columnMap);
HScannerInterface scanner = table.obtainScanner(columns, HConstants.EMPTY_START_ROW, filter);
int numFound = doScan(scanner, printValues);
Assert.assertEquals(NUM_ROWS, numFound);
}
private int doScan(final HScannerInterface scanner, final boolean printValues) throws IOException {
{
int count = 0;
try {
HStoreKey key = new HStoreKey();
TreeMap<Text, byte[]> results = new TreeMap<Text, byte[]>();
while (scanner.next(key, results)) {
if (printValues) {
LOG.info("row: " + key.getRow());
for (Map.Entry<Text, byte[]> e : results.entrySet()) {
LOG.info(" column: " + e.getKey() + " value: "
+ new String(e.getValue(), HConstants.UTF8_ENCODING));
}
}
Assert.assertEquals(2, results.size());
count++;
}
} finally {
scanner.close();
}
return count;
}
}
}