将Java数据对象(JDO)与HBase一起使用

2018-08-13 15:03 更新

将Java数据对象(JDO)与HBase一起使用

Java数据对象(JDO)是一种访问数据库中持久数据的标准方法,使用普通的旧Java对象(PO​​JO)来表示持久数据。

依赖

此代码示例具有以下依赖项:

  1. HBase 0.90.x或更高版本
  2. commons-beanutils.jar(https://commons.apache.org/)
  3. commons-pool-1.5.5.jar(https://commons.apache.org/)
  4. transactional-tableindexed for HBase 0.90(https://github.com/hbase-trx/hbase-transactional-tableindexed)

下载 hbase-jdo

从http://code.google.com/p/hbase-jdo/下载代码。

JDO示例

此示例使用JDO创建一个表和一个索引,在表中插入行,获取行,获取列值,执行查询以及执行一些其他HBase操作。

package com.apache.hadoop.hbase.client.jdo.examples;

import java.io.File;
import java.io.FileInputStream;
import java.io.InputStream;
import java.util.Hashtable;

import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hbase.client.tableindexed.IndexedTable;

import com.apache.hadoop.hbase.client.jdo.AbstractHBaseDBO;
import com.apache.hadoop.hbase.client.jdo.HBaseBigFile;
import com.apache.hadoop.hbase.client.jdo.HBaseDBOImpl;
import com.apache.hadoop.hbase.client.jdo.query.DeleteQuery;
import com.apache.hadoop.hbase.client.jdo.query.HBaseOrder;
import com.apache.hadoop.hbase.client.jdo.query.HBaseParam;
import com.apache.hadoop.hbase.client.jdo.query.InsertQuery;
import com.apache.hadoop.hbase.client.jdo.query.QSearch;
import com.apache.hadoop.hbase.client.jdo.query.SelectQuery;
import com.apache.hadoop.hbase.client.jdo.query.UpdateQuery;

/**
 * Hbase JDO Example.
 *
 * dependency library.
 * - commons-beanutils.jar
 * - commons-pool-1.5.5.jar
 * - hbase0.90.0-transactionl.jar
 *
 * you can expand Delete,Select,Update,Insert Query classes.
 *
 */
public class HBaseExample {
  public static void main(String[] args) throws Exception {
    AbstractHBaseDBO dbo = new HBaseDBOImpl();

    //*drop if table is already exist.*
    if(dbo.isTableExist("user")){
     dbo.deleteTable("user");
    }

    //*create table*
    dbo.createTableIfNotExist("user",HBaseOrder.DESC,"account");
    //dbo.createTableIfNotExist("user",HBaseOrder.ASC,"account");

    //create index.
    String[] cols={"id","name"};
    dbo.addIndexExistingTable("user","account",cols);

    //insert
    InsertQuery insert = dbo.createInsertQuery("user");
    UserBean bean = new UserBean();
    bean.setFamily("account");
    bean.setAge(20);
    bean.setEmail("ncanis@gmail.com");
    bean.setId("ncanis");
    bean.setName("ncanis");
    bean.setPassword("1111");
    insert.insert(bean);

    //select 1 row
    SelectQuery select = dbo.createSelectQuery("user");
    UserBean resultBean = (UserBean)select.select(bean.getRow(),UserBean.class);

    // select column value.
    String value = (String)select.selectColumn(bean.getRow(),"account","id",String.class);

    // search with option (QSearch has EQUAL, NOT_EQUAL, LIKE)
    // select id,password,name,email from account where id='ncanis' limit startRow,20
    HBaseParam param = new HBaseParam();
    param.setPage(bean.getRow(),20);
    param.addColumn("id","password","name","email");
    param.addSearchOption("id","ncanis",QSearch.EQUAL);
    select.search("account", param, UserBean.class);

    // search column value is existing.
    boolean isExist = select.existColumnValue("account","id","ncanis".getBytes());

    // update password.
    UpdateQuery update = dbo.createUpdateQuery("user");
    Hashtable<String, byte[]> colsTable = new Hashtable<String, byte[]>();
    colsTable.put("password","2222".getBytes());
    update.update(bean.getRow(),"account",colsTable);

    //delete
    DeleteQuery delete = dbo.createDeleteQuery("user");
    delete.deleteRow(resultBean.getRow());

    ////////////////////////////////////
    // etc

    // HTable pool with apache commons pool
    // borrow and release. HBasePoolManager(maxActive, minIdle etc..)
    IndexedTable table = dbo.getPool().borrow("user");
    dbo.getPool().release(table);

    // upload bigFile by hadoop directly.
    HBaseBigFile bigFile = new HBaseBigFile();
    File file = new File("doc/movie.avi");
    FileInputStream fis = new FileInputStream(file);
    Path rootPath = new Path("/files/");
    String filename = "movie.avi";
    bigFile.uploadFile(rootPath,filename,fis,true);

    // receive file stream from hadoop.
    Path p = new Path(rootPath,filename);
    InputStream is = bigFile.path2Stream(p,4096);

  }
}
以上内容是否对您有帮助:
在线笔记
App下载
App下载

扫描二维码

下载编程狮App

公众号
微信公众号

编程狮公众号