import com.google.common.collect.Maps;
import org.apache.hadoop.hbase.Cell;
import org.apache.hadoop.hbase.CellUtil;
import org.apache.hadoop.hbase.TableName;
import org.apache.hadoop.hbase.client.*;
import org.apache.hadoop.hbase.filter.CompareFilter;
import org.apache.hadoop.hbase.filter.RowFilter;
import org.apache.hadoop.hbase.filter.SubstringComparator;
import org.apache.hadoop.hbase.io.compress.Compression;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hbase.thirdparty.com.google.common.collect.Lists;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import org.springframework.util.Assert;
import org.springframework.util.StringUtils;
import java.io.IOException;
import java.math.BigDecimal;
import java.nio.ByteBuffer;
import java.nio.charset.StandardCharsets;
import java.util.*;
import static org.apache.hadoop.hbase.CompareOperator.EQUAL;
@Component
public class HbaseUtil {
@Autowired
private Connection connection;
/**
* 创建表
*
* @param tableName 表名
* @param families 列族
*/
public void createTable(String tableName, List<String> families, Compression.Algorithm type) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(!admin.tableExists(name), "This table already exists");
TableDescriptorBuilder desc = TableDescriptorBuilder.newBuilder(name);
for (String cf : families) {
ColumnFamilyDescriptorBuilder builder = ColumnFamilyDescriptorBuilder.newBuilder(cf.getBytes(StandardCharsets.UTF_8));
if (type != null) {
builder.setCompressionType(type);
}
//指定最大版本1,值会被覆盖
builder.setMaxVersions(1);
desc.setColumnFamily(builder.build());
}
admin.createTable(desc.build());
} catch (IOException e) {
throw e;
}
}
public void addFamily2Table(String tableName, List<String> families, Compression.Algorithm type) throws IOException {
TableName table = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(admin.tableExists(table), "This table not exists");
for (String cf : families) {
ColumnFamilyDescriptorBuilder builder = ColumnFamilyDescriptorBuilder.newBuilder(cf.getBytes(StandardCharsets.UTF_8));
if (type != null) {
builder.setCompressionType(type);
}
builder.setMaxVersions(1);//指定最大版本1,值会被覆盖(当时间戳大于等于已存数据的时间戳时会覆盖原来的值)
admin.addColumnFamily(table, builder.build());
}
} catch (IOException e) {
throw e;
}
}
/**
* 删除表操作
*
* @param tableName
*/
public void deleteTable(String tableName) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(admin.tableExists(name), "This table not exists");
admin.disableTable(name);
admin.deleteTable(name);
} catch (IOException e) {
throw e;
}
}
/**
* 插入记录(单行单列族-单列单值-指定时间戳)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
* @param timestamp 值对应的时间戳
*/
public void insertDouble(String tableName, String rowKey, String family, String column, Double value, Long timestamp) throws IOException {
insertObject(tableName, rowKey, family, column, value, timestamp);
}
/**
* 插入记录(单行单列族-单列单值-不指定时间戳)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
*/
public void insertDouble(String tableName, String rowKey, String family, String column, Double value) throws IOException {
insertObject(tableName, rowKey, family, column, value, null);
}
/**
* 插入记录(单行单列族-单列单值)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
* @param timestamp 值对应的时间戳
*/
public void insertObject(String tableName, String rowKey, String family, String column, Object value, Long timestamp) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Put put = new Put(Bytes.toBytes(rowKey));
if (timestamp != null) {
put.setTimestamp(timestamp);
}
put.addColumn(Bytes.toBytes(family), Bytes.toBytes(column), getBytes(value));
table.put(put);
} catch (IOException e) {
throw e;
}
}
/**
* 插入记录(单行单列族-多列多值)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param columns 列名(数组)
* @param values 值(数组)(且需要和列一一对应)
*/
public void insertDoubles(String tableName, String rowKey, String family, List<String> columns, List<Double> values) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Put put = new Put(Bytes.toBytes(rowKey));
for (int i = 0; i < columns.size(); i++) {
put.addColumn(Bytes.toBytes(family), Bytes.toBytes(columns.get(i)), Bytes.toBytes(values.get(i)));
table.put(put);
}
} catch (IOException e) {
throw e;
}
}
/**
* 查找一行记录
*
* @param tableName 表名
* @param rowKey 行名
*/
public Result selectData(String tableName, String rowKey) throws IOException {
return this.selectData(tableName, rowKey, null);
}
/**
* 查找一行记录
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族
*/
public Result selectData(String tableName, String rowKey, String family) throws IOException {
return this.selectData(tableName, rowKey, family, null);
}
/**
* 查找一行记录
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族
* @param qualifier 列名
*/
public Result selectData(String tableName, String rowKey, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Get g = new Get(rowKey.getBytes(StandardCharsets.UTF_8));
if (!StringUtils.isEmpty(family)) {
g.addFamily(Bytes.toBytes(family));
if (!StringUtils.isEmpty(qualifier)) {
g.addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
}
}
Result rs = table.get(g);
return rs;
} catch (IOException e) {
throw e;
}
}
/**
* @param tableName 表格名称
* @param rowKeyStart 开始rowKey
* @param rowKeyEnd 结束rowKey
* @return
* @throws IOException
*/
public ResultScanner selectDataList(String tableName, String rowKeyStart, String rowKeyEnd) throws IOException {
return this.selectDataList(tableName, rowKeyStart, rowKeyEnd, null);
}
/**
* @param tableName 表格名称
* @param rowKeyStart 开始rowKey
* @param rowKeyEnd 结束rowKey
* @param family 列族
* @return
* @throws IOException
*/
public ResultScanner selectDataList(String tableName, String rowKeyStart, String rowKeyEnd, String family) throws IOException {
return this.selectDataList(tableName, rowKeyStart, rowKeyEnd, family, null);
}
/**
* @param tableName 表格名称
* @param rowKeyStart 开始rowKey
* @param rowKeyEnd 结束rowKey
* @param family 列族
* @param qualifier 列
* @return
* @throws IOException
*/
public ResultScanner selectDataList(String tableName, String rowKeyStart, String rowKeyEnd, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Scan scan = new Scan();
boolean isHasFamily = StringUtils.isEmpty(family);
boolean isHasQualifier = StringUtils.isEmpty(qualifier);
if (!isHasFamily) {
scan.addFamily(Bytes.toBytes(family));
if (!isHasQualifier) {
scan.addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
}
}
scan.withStartRow(Bytes.toBytes(rowKeyStart));
scan.withStopRow(Bytes.toBytes(rowKeyEnd));
ResultScanner scanner = table.getScanner(scan);
return scanner;
} catch (IOException e) {
throw e;
}
}
/**
* @param qualifierMap 列集合
* @param entries 列族集合
* @return 经过转换的返回值
*/
private List<Map<String, Object>> assembleReturnData(Map<String, List<Object>> qualifierMap, Set<Map.Entry<String, Set<String>>> entries) {
List<Map<String, Object>> families = Lists.newArrayList();
for (Iterator<Map.Entry<String, Set<String>>> it = entries.iterator(); it.hasNext(); ) {
Map.Entry<String, Set<String>> next = it.next();
String key = next.getKey();
Set<String> value = next.getValue();
Map<String, Object> familyData = Maps.newHashMap();
List<Map<String, Object>> qualifiers = Lists.newArrayList();
familyData.put("family", key);
familyData.put("qualifiers", qualifiers);
families.add(familyData);
for (String cell : value) {
Map<String, Object> qualifierData = Maps.newHashMap();
qualifierData.put("qualifier", cell);
qualifierData.put("times", qualifierMap.get(cell + "Times"));
qualifierData.put("values", qualifierMap.get(cell + "Values"));
qualifiers.add(qualifierData);
}
}
return families;
}
/**
* 组装数据
*
* @param familyList 列族集合
* @param qualifierList 列集合
* @param rs
* @param rowKeyList 行键集合
*/
private void assembleCellData(List<String> familyList, List<Map<String, String>> qualifierList, Result rs, List<String> rowKeyList) {
if (rs.size() == 0) {
return;
}
for (Cell cell : rs.listCells()) {
String family = Bytes.toString(CellUtil.cloneFamily(cell));
String qualifier = Bytes.toString(CellUtil.cloneQualifier(cell));
String value = Bytes.toString(CellUtil.cloneValue(cell));
String rowKey = Bytes.toString(CellUtil.cloneRow(cell));
familyList.add(family);
if(!rowKeyList.contains(rowKey)){
rowKeyList.add(rowKey);
}
Map<String, String> map = new HashMap<>();
map.put("key", qualifier);
map.put("value", value);
map.put("row",rowKey);
qualifierList.add(map);
}
}
/**
* 删除一行记录
*
* @param tableName 表名
* @param rowKey 行名
*/
public void deleteRow(String tableName, String rowKey) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes(StandardCharsets.UTF_8));
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 删除单行单列族记录
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
*/
public void deleteFamily(String tableName, String rowKey, String family) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes(StandardCharsets.UTF_8)).addFamily(Bytes.toBytes(family));
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 删除单行单列族单列记录
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param qualifier 列名
*/
public void deleteColumn(String tableName, String rowKey, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes(StandardCharsets.UTF_8)).addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 查找单行单列族单列记录
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @return
*/
public Object selectValue(String tableName, String rowKey, String family, String column) throws IOException {
TableName name = TableName.valueOf(tableName);
Table table = connection.getTable(name);
Get g = new Get(rowKey.getBytes(StandardCharsets.UTF_8));
g.addColumn(Bytes.toBytes(family), Bytes.toBytes(column));
Result rs = table.get(g);
return getObject(rs.value(), String.class);
}
/**
* 根据rowKey关键字查询报告记录
*
* @param tableName
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeyword(String tableName, String rowKeyword) throws IOException {
List list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tableName));
Scan scan = new Scan();
//添加行键过滤器,根据关键字匹配
RowFilter rowFilter = new RowFilter(CompareFilter.CompareOp.EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此处根据业务来自定义实现
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 根据rowKey关键字和时间戳范围查询报告记录
*
* @param tableName
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeywordTimestamp(String tableName, String rowKeyword, Long minStamp, Long maxStamp) throws IOException {
ArrayList list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tableName));
Scan scan = new Scan();
//添加scan的时间范围
scan.setTimeRange(minStamp, maxStamp);
RowFilter rowFilter = new RowFilter(EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此处根据业务来自定义实现
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 利用协处理器进行全表count统计
*
* @param tableName
*/
/*public Long countRowsWithCoprocessor(String tableName) throws Throwable {
TableName name = TableName.valueOf(tableName);
Admin admin = connection.getAdmin();
HTableDescriptor descriptor = admin.getTableDescriptor(name);
String coprocessorClass = "org.apache.hadoop.hbase.coprocessor.AggregateImplementation";
if (!descriptor.hasCoprocessor(coprocessorClass)) {
admin.disableTable(name);
descriptor.addCoprocessor(coprocessorClass);
admin.modifyTable(name, descriptor);
admin.enableTable(name);
}
//计时
StopWatch stopWatch = new StopWatch();
stopWatch.start();
Scan scan = new Scan();
//Hbase服务端需要开启该服务
AggregationClient aggregationClient = new AggregationClient(connection.getConfiguration());
Long count = aggregationClient.rowCount(name, new LongColumnInterpreter(), scan);
stopWatch.stop();
log.debug("RowCount:{} ,全表count统计耗时:{}", count, stopWatch.getTotalTimeMillis());
return count;
}*/
/**
* 根据不同的类型获取不同的byte数组
*
* @param object
* @return
*/
private static byte[] getBytes(Object object) {
Class className = object.getClass();
if (className.equals(java.lang.Boolean.class)) {
return Bytes.toBytes((Boolean) object);
}
if (className.equals(java.lang.Byte.class)) {
return Bytes.toBytes((Byte) object);
}
if (className.equals(java.lang.Short.class)) {
return Bytes.toBytes((Short) object);
}
if (className.equals(java.lang.Character.class)) {
return Bytes.toBytes((Character) object);
}
if (className.equals(java.lang.Integer.class)) {
return Bytes.toBytes((Integer) object);
}
if (className.equals(java.lang.Long.class)) {
return Bytes.toBytes((Long) object);
}
if (className.equals(java.lang.Float.class)) {
return Bytes.toBytes((Float) object);
}
if (className.equals(java.lang.Double.class)) {
return Bytes.toBytes((Double) object);
}
if (className.equals(java.lang.String.class)) {
return Bytes.toBytes((String) object);
}
if (className.equals(java.nio.ByteBuffer.class)) {
return Bytes.toBytes((ByteBuffer) object);
}
if (className.equals(java.math.BigDecimal.class)) {
return Bytes.toBytes((BigDecimal) object);
}
return ByteUtil.toBytes(object);
}
/**
* 根据不同的byte数组获取不同的对象
*
* @param bytes
* @return
*/
private static Object getObject(byte[] bytes, Class className) {
if (className.equals(java.lang.Boolean.class)) {
return Bytes.toBoolean(bytes);
}
if (className.equals(java.lang.Byte.class)) {
return Bytes.toShort(bytes);
}
if (className.equals(java.lang.Short.class)) {
return Bytes.toShort(bytes);
}
if (className.equals(java.lang.Character.class)) {
return Bytes.toString(bytes);
}
if (className.equals(java.lang.Integer.class)) {
return Bytes.toInt(bytes);
}
if (className.equals(java.lang.Long.class)) {
return Bytes.toLong(bytes);
}
if (className.equals(java.lang.Float.class)) {
return Bytes.toFloat(bytes);
}
if (className.equals(java.lang.Double.class)) {
return Bytes.toDouble(bytes);
}
if (className.equals(java.lang.String.class)) {
return Bytes.toString(bytes);
}
if (className.equals(java.nio.ByteBuffer.class)) {
return ByteBuffer.wrap(bytes);
}
if (className.equals(java.math.BigDecimal.class)) {
return Bytes.toBigDecimal(bytes);
}
return ByteUtil.toObject(bytes);
}
}
HBase工具类
最新推荐文章于 2022-07-17 17:30:54 发布
这个Java类提供了HBase的基本操作,包括创建和删除表,向表中插入数据(单行单列、多列),查询单行或多行数据,以及删除数据。类中还实现了基于行键的过滤查询和时间戳范围查询。此外,它支持设置列族的压缩算法和最大版本数。
摘要由CSDN通过智能技术生成