親測，只能調用一次，如果想想在一次reduce重復使用叠代器中的數據，得先取出來放在list中然後在從list中取出來！！多次讀取reduce函數中叠代器的數據

  public static class FindFriendReducer extends Reducer<Text, AllInfoBean, AllInfoBean, NullWritable>
  {
    protected void reduce(Text Keyin, Iterable<AllInfoBean> values, Reducer<Text, AllInfoBean, AllInfoBean, NullWritable>.Context context)
      throws IOException, InterruptedException
    {
      AllInfoBean allInfoBean 
 
= new AllInfoBean();
      allInfoBean.setPid(Keyin.toString());
      for (AllInfoBean bean : values) {
        if (bean.getOderid() != 0)
          continue;
        allInfoBean.setPname(bean.getPname());
        allInfoBean.setCategory_id(bean.getCategory_id());
        allInfoBean.setPrice(bean.getPrice());
      }

      
 
for (AllInfoBean bean : values)
      {
        context.write(bean, NullWritable.get());
      }
    }
  }

這樣是取不出來數據的！！

就 象jdbc 的resultset一樣只能取一次

mapreduce中reduce中的叠代器只能調用一次！