HBase java API 使用实例（二）——查询、分页
上一篇HBase文章中已经介绍了如何使用java API操作HBase（http://blog.csdn.net/wild46cat/article/details/53306621），现在介绍一下如何使用java API对HBase进行查询。
本文主要内容：
1、使用java API对HBase进行常用的查询（行、列过滤）。
2、介绍如何进行分页，这里的分页不是最终的解决办法，只是目前自己测试使用时的一种方案（由于数据被split之后会分布在各个Region Server上，如果没有合理的一个RowKey策略的话，分页是不适用的）。
3、最后会附上源码。
好啦，下面上货：
HBaseUtils.java(代码是全的，主要增加的是query查询的一些方法)
package com.xueyoucto.hbasett;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.*;
import org.apache.hadoop.hbase.client.*;
import org.apache.hadoop.hbase.filter.*;
import org.apache.hadoop.hbase.protobuf.generated.AccessControlProtos;
import org.apache.hadoop.hbase.util.Bytes;
import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
/**
* Created by Administrator on 2016-11-23.
*/
public class HBaseUtils {
static Configuration cfg;
static Connection connection;
static {
cfg = HBaseConfiguration.create();
try {
connection = ConnectionFactory.createConnection(cfg);
} catch (IOException e) {
e.printStackTrace();
}
System.out.println(cfg.get("hbase.master"));
}
//新建表
public static boolean create(String tableName, String columnFamily)
throws Exception {
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
if (admin.tableExists(tableName)) {
System.out.println(tableName + " exists!");
return false;
} else {
String[] columnFamilyArray = columnFamily.split(",");
HColumnDescriptor[] hColumnDescriptor = new HColumnDescriptor[columnFamilyArray.length];
for (int i = 0; i < hColumnDescriptor.length; i++) {
hColumnDescriptor[i] = new HColumnDescriptor(columnFamilyArray[i]);
}
HTableDescriptor familyDesc = new HTableDescriptor(TableName.valueOf(tableName));
for (HColumnDescriptor columnDescriptor : hColumnDescriptor) {
familyDesc.addFamily(columnDescriptor);
}
HTableDescriptor tableDesc = new HTableDescriptor(TableName.valueOf(tableName), familyDesc);
admin.createTable(tableDesc);
System.out.println(tableName + " create successfully!");
return true;
}
}
//插入数据
public static boolean put(String tablename, String row, String columnFamily,
String qualifier, String data) throws Exception {
Table table = connection.getTable(TableName.valueOf(tablename));
Put put = new Put(Bytes.toBytes(row));
put.addColumn(Bytes.toBytes(columnFamily), Bytes.toBytes(qualifier),
Bytes.toBytes(data));
table.put(put);
System.out.println("put '" + row + "', '" + columnFamily + ":" + qualifier
+ "', '" + data + "'");
return true;
}
//把result转换成map，方便返回json数据
public static Map<String, Object> resultToMap(Result result) {
Map<String, Object> resMap = new HashMap<String, Object>();
List<Cell> listCell = result.listCells();
Map<String, Object> tempMap = new HashMap<String, Object>();
String rowname = "";
List<String> familynamelist = new ArrayList<String>();
for (Cell cell : listCell) {
byte[] rowArray = cell.getRowArray();
byte[] familyArray = cell.getFamilyArray();
byte[] qualifierArray = cell.getQualifierArray();
byte[] valueArray = cell.getValueArray();
int rowoffset = cell.getRowOffset();
int familyoffset = cell.getFamilyOffset();
int qualifieroffset = cell.getQualifierOffset();
int valueoffset = cell.getValueOffset();
int rowlength = cell.getRowLength();
int familylength = cell.getFamilyLength();
int qualifierlength = cell.getQualifierLength();
int valuelength = cell.getValueLength();
byte[] temprowarray = new byte[rowlength];
System.arraycopy(rowArray, rowoffset, temprowarray, 0, rowlength);
String temprow = Bytes.toString(temprowarray);
// System.out.println(Bytes.toString(temprowarray));
byte[] tempqulifierarray = new byte[qualifierlength];
System.arraycopy(qualifierArray, qualifieroffset, tempqulifierarray, 0, qualifierlength);
String tempqulifier = Bytes.toString(tempqulifierarray);
// System.out.println(Bytes.toString(tempqulifierarray));
byte[] tempfamilyarray = new byte[familylength];
System.arraycopy(familyArray, familyoffset, tempfamilyarray, 0, familylength);
String tempfamily = Bytes.toString(tempfamilyarray);
// System.out.println(Bytes.toString(tempfamilyarray));
byte[] tempvaluearray = new byte[valuelength];
System.arraycopy(valueArray, valueoffset, tempvaluearray, 0, valuelength);
String tempvalue = Bytes.toString(tempvaluearray);
// System.out.println(Bytes.toString(tempvaluearray));
tempMap.put(tempfamily + ":" + tempqulifier, tempvalue);
// long t= cell.getTimestamp();
// tempMap.put("timestamp",t);
rowname = temprow;
String familyname = tempfamily;
if (familynamelist.indexOf(familyname) < 0) {
familynamelist.add(familyname);
}
}
resMap.put("rowname", rowname);
for (String familyname : familynamelist) {
HashMap<String, Object> tempFilterMap = new HashMap<String, Object>();
for (String key : tempMap.keySet()) {
String[] keyArray = key.split(":");
if (keyArray[0].equals(familyname)) {
tempFilterMap.put(keyArray[1], tempMap.get(key));
}
}
resMap.put(familyname, tempFilterMap);
}
return resMap;
}
//查看某行
public static String get(String tablename, String row) throws Exception {
Table table = connection.getTable(TableName.valueOf(tablename));
Get get = new Get(Bytes.toBytes(row));
Result result = table.get(get);
System.out.println("Get: " + result);
return resultToMap(result).toString();
}
//查看全表
public static String scan(String tablename) throws Exception {
Table table = connection.getTable(TableName.valueOf(tablename));
Scan s = new Scan();
ResultScanner rs = table.getScanner(s);
List<Map<String, Object>> resList = new ArrayList<Map<String, Object>>();
for (Result r : rs) {
Map<String, Object> tempmap = resultToMap(r);
resList.add(tempmap);
}
return resList.toString();
}
//删除表
public static boolean delete(String tableName) throws IOException {
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
if (admin.tableExists(tableName)) {
try {
admin.disableTable(tableName);
admin.deleteTable(tableName);
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
return true;
}
//删除ColumnFamily
public static boolean deleteColumnFamily(String tableName, String columnFamilyName) throws IOException {
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
if (admin.tableExists(tableName)) {
try {
admin.deleteColumn(tableName, columnFamilyName);
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
return true;
}
//删除row
public static boolean deleteRow(String tableName, String rowName) throws IOException {
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
Table table = connection.getTable(TableName.valueOf(tableName));
if (admin.tableExists(tableName)) {
try {
Delete delete = new Delete(rowName.getBytes());
table.delete(delete);
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
return true;
}
//删除qualifier
public static boolean deleteQualifier(String tableName, String rowName, String columnFamilyName, String qualifierName) throws IOException {
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
Table table = connection.getTable(TableName.valueOf(tableName));
if (admin.tableExists(tableName)) {
try {
Delete delete = new Delete(rowName.getBytes());
delete.addColumns(columnFamilyName.getBytes(), qualifierName.getBytes());
table.delete(delete);
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
return true;
}
//单个qualifier的值等于
public static String queryequal(String tablename, String columnFamily, String qualifier, String data) throws Exception {
//某列等于data的
Filter filter = new SingleColumnValueFilter(Bytes.toBytes(columnFamily), Bytes.toBytes(qualifier),
CompareFilter.CompareOp.EQUAL, Bytes.toBytes(data));
FilterList filterList = new FilterList();
filterList.addFilter(filter);
return query(tablename, filterList);
}
//查询年龄在mindata和maxdata之间的数据
public static String queryagebetween(String tablename, String columnFamily, String qualifier, String mindata, String maxdata) throws Exception {
Filter filter = new SingleColumnValueFilter(Bytes.toBytes(columnFamily), Bytes.toBytes(qualifier),
CompareFilter.CompareOp.LESS_OR_EQUAL, Bytes.toBytes(maxdata));
Filter filter1 = new SingleColumnValueFilter(Bytes.toBytes(columnFamily), Bytes.toBytes(qualifier),
CompareFilter.CompareOp.GREATER_OR_EQUAL, Bytes.toBytes(mindata));
FilterList filterList = new FilterList();
filterList.addFilter(filter);
filterList.addFilter(filter1);
return query(tablename, filterList);
}
//过滤列
public static String querycolumn(String tablename, String prefix) throws Exception {
Filter filter = new ColumnPrefixFilter(Bytes.toBytes(prefix));
FilterList filterList = new FilterList();
filterList.addFilter(filter);
return query(tablename, filterList);
}
//得到分页
public static String queryRowCount(String tablename, String startrowname, String count) throws Exception {
Filter filter = new PageFilter(Integer.parseInt(count));
Filter filter1 = new SingleColumnValueFilter(Bytes.toBytes("author"), Bytes.toBytes("age"),
CompareFilter.CompareOp.EQUAL, Bytes.toBytes("30"));
FilterList filterList = new FilterList();
// filterList.addFilter(filter1);
filterList.addFilter(filter);
Table table = connection.getTable(TableName.valueOf(tablename));
Scan s = new Scan(Bytes.toBytes(startrowname));
s.setFilter(filterList);
ResultScanner rs = table.getScanner(s);
List<Map<String, Object>> resList = new ArrayList<Map<String, Object>>();
for (Result r : rs) {
Map<String, Object> tempmap = resultToMap(r);
resList.add(tempmap);
}
return resList.toString();
}
//所有查询都需要的公共query方法
public static String query(String tablename, FilterList filterList) throws Exception {
Table table = connection.getTable(TableName.valueOf(tablename));
Scan s = new Scan();
s.setFilter(filterList);
ResultScanner rs = table.getScanner(s);
List<Map<String, Object>> resList = new ArrayList<Map<String, Object>>();
for (Result r : rs) {
Map<String, Object> tempmap = resultToMap(r);
resList.add(tempmap);
}
return resList.toString();
}
//测试用，添加数据
public static boolean testInsertdata() throws Exception {
Table table = connection.getTable(TableName.valueOf("test"));
for (int i = 20; i < 12000; i++) {
Put put = new Put(Bytes.toBytes("r" + String.valueOf(i)));
put.addColumn(Bytes.toBytes("author"), Bytes.toBytes("age"),
Bytes.toBytes(String.valueOf(i)));
table.put(put);
System.out.println("put '" + "r" + String.valueOf(i)+ "', '" + "author" + ":" + "age"
+ "', '" + String.valueOf(i) + "'");
}
return true;
}
//测试用删除数据
public static boolean deleteInsertdata() throws Exception{
HBaseAdmin admin = (HBaseAdmin) connection.getAdmin();
Table table = connection.getTable(TableName.valueOf("test"));
if (admin.tableExists("test")) {
try {
for (int i = 20; i < 12000; i++) {
Delete delete = new Delete(("r" + String.valueOf(i)).getBytes());
table.delete(delete);
}
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
return true;
}
}
对于新增加的方法进行测试：
1、测试查询某个qualifier为特定值的列。例如：博客作者中年龄是30的内容。
package com.xueyoucto.hbasett;
import javax.servlet.ServletException;
import javax.servlet.annotation.WebServlet;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import java.io.IOException;
import java.io.PrintWriter;
/**
* Created by Administrator on 2016-11-24.
*/
@WebServlet(name = "ServletQuery", urlPatterns = "/HBaseServletQueryEqual")
public class ServletQuery extends HttpServlet {
protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
doGet(request, response);
}
protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
String tname = request.getParameter("tablename");
String columnFamily = request.getParameter("columnfamily");
String qualifier = request.getParameter("qualifier");
String data = request.getParameter("data");
String msg = "-";
try {
msg = HBaseUtils.queryequal(tname, columnFamily,qualifier,data);
} catch (Exception e) {
e.printStackTrace();
}
response.setCharacterEncoding("utf-8");
response.setContentType("text/html;charset=utf-8");
PrintWriter out = response.getWriter();
out.println(msg);
out.flush();
out.close();
}
}
首先看一下所有数据：
查询博客作者年龄是30岁的数据（有脏数据，为了测试后面的功能）：
注意：所有的数据必须都有author:age这个quailifer，不然，默认情况下如果没有这个列，那么依然会被查出来。
2、测试博客作者年龄在40到50岁之间的内容：
package com.xueyoucto.hbasett;
import javax.servlet.ServletException;
import javax.servlet.annotation.WebServlet;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import java.io.IOException;
import java.io.PrintWriter;
/**
* Created by Administrator on 2016-11-24.
*/
@WebServlet(name = "ServletQueryAgeBetween",urlPatterns = "/HBaseServletQueryAgeBetween")
public class ServletQueryAgeBetween extends HttpServlet {
protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
doGet(request,response);
}
protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
String tname = request.getParameter("tablename");
String columnFamily = request.getParameter("columnfamily");
String qualifier = request.getParameter("qualifier");
String mindata = request.getParameter("mindata");
String maxdata = request.getParameter("maxdata");
String msg = "-";
try {
msg = HBaseUtils.queryagebetween(tname, columnFamily, qualifier, mindata, maxdata);
} catch (Exception e) {
e.printStackTrace();
}
response.setCharacterEncoding("utf-8");
response.setContentType("text/html;charset=utf-8");
PrintWriter out = response.getWriter();
out.println(msg);
out.flush();
out.close();
}
}
查询：
注意：这里的40和50是字符串查询，也就是说，如果有个age=400，那么依然会被包含在结果中。比较大小的时候是通过ASCII码。
3、测试查询后的结果只取指定列（查询blog作者的年龄）：
package com.xueyoucto.hbasett;
import javax.servlet.ServletException;
import javax.servlet.annotation.WebServlet;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import java.io.IOException;
import java.io.PrintWriter;
/**
* Created by Administrator on 2016-11-24.
*/
@WebServlet(name = "ServletQueryColumn", urlPatterns = "/HBaseServletQueryColumn")
public class ServletQueryColumn extends HttpServlet {
protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
doGet(request, response);
}
protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
String tname = request.getParameter("tablename");
String prefix = request.getParameter("prefix");
String msg = "-";
try {
msg = HBaseUtils.querycolumn(tname, prefix);
} catch (Exception e) {
e.printStackTrace();
}
response.setCharacterEncoding("utf-8");
response.setContentType("text/html;charset=utf-8");
PrintWriter out = response.getWriter();
out.println(msg);
out.flush();
out.close();
}
}
执行查询：
4、测试分页（得到从startrow开始的5条数据）：
package com.xueyoucto.hbasett;
import javax.servlet.ServletException;
import javax.servlet.annotation.WebServlet;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import java.io.IOException;
import java.io.PrintWriter;
/**
* Created by Administrator on 2016-11-24.
*/
@WebServlet(name = "ServletQueryCount", urlPatterns = "/HBaseServletQueryCount")
public class ServletQueryCount extends HttpServlet {
protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
doGet(request, response);
}
protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
String tname = request.getParameter("tablename");
String startrowname = request.getParameter("startrowname");
String count = request.getParameter("count");
String msg = "-";
try {
msg = HBaseUtils.queryRowCount(tname, startrowname, count);
} catch (Exception e) {
e.printStackTrace();
}
response.setCharacterEncoding("utf-8");
response.setContentType("text/html;charset=utf-8");
PrintWriter out = response.getWriter();
out.println(msg);
out.flush();
out.close();
}
}
执行查询：
注意：这里的查询出来的数据有些不对，应该是只有5条记录。
原因是这样的，先看下图：
注意看，对于表test，我把它按照rowkey进行split了，分成了三个Region。
再看下图，PageFilter类的说明：
大概的意思就是说PageFilter是FilterBase的一个实现类，在查询的时候，如果大于filter中设置的number，那么就会停止。需要注意的是，这个filter不能保证返回给client的数目就一定小于等于设置的pagesize。原因是这样的，filter是分别过滤不同region server上的数据，而这些regionserver大部分情况下是分开的。但是，如果能够保证某个表是在单独的region上，那么返回的返回的client的数据就会小于等于设置的pagesize。
好，那么现在有两种方法保证分页的正确性：
1、保证数据不会被split，也就是数据始终在一个region上。
2、被查询出的数据都在一个region上。
首先这两种方式都可以。各有各的好处。首先第一种的优点是方便，缺点明显，就是可能导致单点负载较高。第二种的优点是负载均衡，缺点是需要根据业务设置rowkey，从而保证同一业务的数据根据设置好的规则split后能够被分配到相同的region中。然后根据startrow和PagerFilter设置的pagesize进行分页操作（略了，有兴趣可以自己实现以下）。
源码下载地址：
http://download.csdn.net/detail/wild46cat/9692647
扫一扫
2045
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
