JDBC学习4：preparedStatement可以插入blob数据和批量处理

1.blob数据的插入

@Test
	//插入一条blog数据customers
	public void testInsert() throws Exception{
		//1.获取一连接
		Connection conn = JDBCUtils.getConnection();
		
		//2.获取prepared
		String sql = "insert into customers(name,photo) values(?,?)";
		PreparedStatement ps = conn.prepareStatement(sql);
		//3.填充sql
		ps.setString(1,"风景");		
		//单元测试模块默认路径在工程下
		FileInputStream fs = new FileInputStream(new File("girl.jpg"));
		ps.setBlob(2, fs);
		//4.执行
		ps.execute();
		//5.关闭资源
		JDBCUtils.closeResource(conn, ps);
		fs.close();
	}
	@Test
	//查询一条blog数据
	public void testSelect() throws Exception{
		//1.获取一连接
		Connection conn = JDBCUtils.getConnection();
		
		//2.获取prepared
		String sql = "select name,photo from customers where name =?";
		PreparedStatement ps = conn.prepareStatement(sql);
		//3.填充sql
		ps.setString(1,"风景");
		//4.执行
		ResultSet rs = ps.executeQuery();
		//5.保存数据
		if(rs.next()){
			String name = rs.getString(1);
			Blob blob = rs.getBlob(2);
			InputStream is = blob.getBinaryStream();
			FileOutputStream fs = new FileOutputStream("fengjing.jpg");
			byte[] buffer = new byte[1024];
			int len;
			while((len=is.read(buffer))!=-1){
				fs.write(buffer, 0, len);
			}
			fs.close();
			is.close();
			
		}
		
		//6.关闭资源
		JDBCUtils.closeResource(conn, ps);
		rs.close();
	}
	//默认pack大小为1M 如果大于这个，尽管blob支持16m，但是仍然不能增加
	//需要在init文件中 增加 max_allowed_packet=16M 此时可以支持16m
}
	

2.批量处理

//预编译含义：即preparedstatement是预编译的，当其将sql语句送入DBMS时，
//sql语句不需要经过编译校验，直接执行

//addBatch()即正常来说java送入一条语句进入DBMS,但是因为送的时间慢，但是DBMS处理速度很快
//故我们可以将打包一批语句送入DBMS一起执行

//commit()提交操作，正常来说(DML)sql语句进入DBMS时每执行一次自动commit一次，但是可以我们可以执行多次
//然后进行一次commit

public class manyInsert {
	
	
	//1.批量插入方式：通过statement来执行insert语句的拼接操作
	//2.使用preparedStatement批量操作
	@Test
	public void testInsert1() throws Exception{
		Connection conn = JDBCUtils.getConnection();
		
		String sql = "insert into goods(name) values(?)";
		PreparedStatement ps = conn.prepareStatement(sql);
		long cus = System.currentTimeMillis();
		for(int i = 0 ;i<20;i++){
			ps.setObject(1, "lihua"+i);
			ps.execute();
		}
		long cusm = System.currentTimeMillis();
		System.out.println(cus-cusm);
		JDBCUtils.closeResource(conn, ps);
	}
	//3.积攒sql的占位符一起（减少java和数据库的交互次数）
	// 默认情况下 sql不支持batch(需要更改init以及更新驱动)
	@Test
	public void testInsert2() throws Exception{
		Connection conn = JDBCUtils.getConnection();
		System.out.println("连接获取成功");
		String sql = "insert into goods(name) values(?)";
		PreparedStatement ps = conn.prepareStatement(sql);
		long cus = System.currentTimeMillis();
		for(int i = 0 ;i<20000;i++){
			ps.setObject(1, "lihua"+i);
			ps.addBatch();
			if(i%500==0)
			{
				ps.execute();
				ps.clearBatch();
			}
		}
		long cusm = System.currentTimeMillis();
		System.out.println(cus-cusm);
		JDBCUtils.closeResource(conn, ps);
	}
	//4.DML执行一次会commit一次，我们可以延迟到最后再commit
	@Test
	public void testInsert3() throws Exception{
		Connection conn = JDBCUtils.getConnection();
		
		String sql = "insert into goods(name) values(?)";
		PreparedStatement ps = conn.prepareStatement(sql);
		long cus = System.currentTimeMillis();
		conn.setAutoCommit(false);
		
		for(int i = 0 ;i<20000;i++){
			ps.setObject(1, "lihua"+i);
			ps.addBatch();
			if(i%500==0)
			{
				ps.execute();
				ps.clearBatch();
			}
		}
		conn.commit();
		long cusm = System.currentTimeMillis();
		System.out.println(cusm-cus);
		JDBCUtils.closeResource(conn, ps);
	}

}