通过剖析源码单步调试详解MapReduce分组group遍历

通过剖析源码单步调试详解MapReduce分组group遍历:
马克-to-win @ 马克java社区:mapreduce的group知识点是最难理解的,本小节将通过仔细剖析源码,单步调试,来详解之。
另外注意:数据文件写时一定注意:结尾不能有回车和空格,通过在map里面加断点,F8(resume),一轮一轮,调试一行一行的数据,才发现最后一行数据出毛病了,只有是多了一个换行符的毛病。

package com;
import org.apache.hadoop.io.WritableComparable;
import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

public class OrderBean implements WritableComparable<OrderBean>{
    private String orderId;
    private Double amount;

    public OrderBean() {
    }

    public OrderBean(String orderId, Double amount) {
        this.orderId = orderId;
        this.amount = amount;
    }
    public void set(String orderId, Double amount) {
        this.orderId = orderId;
        this.amount = amount;
    }
    public String getOrderId() {
        return orderId;
    }

    public void setOrderId(String orderId) {
        this.orderId = orderId;
    }

    public Double getAmount() {
        return amount;
    }

    public void setAmount(Double amount) {
        this.amount = amount;
    }

    @Override
    public int compareTo(OrderBean o) {
        /*马克-to-win: 如orderId相等,则比较amount,否则比较id即可,这样id相同的连在一起了。因为id先比,amount后比*/
        int cmp = this.getOrderId().compareTo(o.getOrderId());
 
更多内容请见原文,文章转载自:https://blog.csdn.net/qq_44594249/article/details/97616735

上一篇:sql语句去掉前面的0(前导零,零前缀)


下一篇:SQLServer函数 left()、charindex()、stuff()