iBatis中RowHandler处理大数据量的导出

背景：一次性导出几十万条数据到excel，此时的导出效率会严重损失。

解决方法： 用ibatis中的RowHandler，一次只处理一条数据，内存中只保持一条数据，导出时每5万条创建一个sheet

具体实现方法：

MyRowHandler处理类：

import java.util.Map;

import org.apache.poi.hssf.usermodel.HSSFCell;
import org.apache.poi.hssf.usermodel.HSSFRichTextString;
import org.apache.poi.hssf.usermodel.HSSFRow;
import org.apache.poi.hssf.usermodel.HSSFSheet;
import org.apache.poi.hssf.usermodel.HSSFWorkbook;

import com.ibatis.sqlmap.client.event.RowHandler;

public class MyRowHandler implements RowHandler {
	HSSFWorkbook wb;
	HSSFSheet sheet ; 
	HSSFRow row;
	HSSFCell cell;
	int dataNum=0;
	int rowStart = 1;
	@Override
	public void handleRow(Object arg0){
		dataNum++;
		if(dataNum%50000==0){
			//每5万条记录分页
			setSheet();
		}
		
		Map map = (Map)arg0;
		//在这里处理逻辑，插入数据到excel
		System.out.println(map.get("key"));
		System.out.println(map.get("value"));
		
	}
	public HSSFWorkbook getWb() {
		return wb;
	}
	public void setWb(HSSFWorkbook wb) {
		this.wb = wb;
		setSheet();
	}
	public void setSheet(){
		this.sheet= wb.createSheet();
		rowStart = 1;
		//写表头
		writeHeader();
	}
	public void writeHeader(){
		String[] header = {"ID","类型"};
		row = sheet.createRow(0);
		int n=0;
		for (String  head : header) {
			cell = row.createCell((short) n);
			HSSFRichTextString str = new HSSFRichTextString(head);
			cell.setCellValue(str);
			n++; 
		}
	}
	
}

 action中执行：

public String execute(){
		
		String fileName="F:\\05-JAVA\\测试.xls";
		HSSFWorkbook wb = new HSSFWorkbook();
		MyRowHandler handler = new MyRowHandler();
		handler.setWb(wb);
		DaoFactory.getIbatisDao().getSqlMapClientTemplate().queryWithRowHandler("term.base", handler);
		wb = handler.getWb();
		
		OutputStream os = null;
		try {
			os = new FileOutputStream(fileName);
			wb.write(os);
		}catch(Exception e){
			
		}finally{
			try {
				os.close();
			} catch (IOException e) {
				e.printStackTrace();
			} 
		}
             return "success";
	}