spring DATA (jpa) @query 和JdbcTemplate

这几天做项目的时候，需求是统计订单表的数据，其中技术持久层的技术是jpa ,用实体映射比较困难。后来先想到jdbc的 JdbcTemplate，可是头说不行，之后用@query 原生语句查询。

1 、 JdbcTemplate实现，较为复杂的sql语句。重点是实体对应的实现。

service里

@Autowired

    JdbcTemplate jdbcTemplate;

 //计算出昨日流量渠道统计量

    public List<OrderChannelQualityDTO> queryByHour(String txNo,Date fromTime,Date toTime) {
//String selectSQL = " select t.id,t.channel_id,t.interface_time,t.trade_time,t.process,t.timeout,t.create_time from t_ar_performance t where t.create_time>=? andt.create_time<=? order
by t.create_time desc ,t.channel_id asc ";

String minDate = "round((UNIX_TIM
4000
ESTAMP(t.update_time)-UNIX_TIMESTAMP(t.created_time))/60)";

      String selectSQL = "SELECT SUM(IF(t.status ='success',1,0))as successCount,SUM(IF(t.status ='fail',1,0))as failCount ,"
+ "SUM(CASE WHEN  0<="+minDate+" and "+minDate+"<5 THEN 1 ELSE 0 END) as fiveCount,"
+ "SUM(CASE
WHEN  5<="+minDate+" and "+minDate+"<1 THEN 1 ELSE 0 END) as tenCount,"
+ "SUM(CASE
WHEN  10<="+minDate+" and "+minDate+"<30     THEN 1 ELSE 0 END) as thirtyContent,"
+ "SUM(CASE
WHEN   30<="+minDate+" and "+minDate+"<60     THEN 1 ELSE 0 END) as sixtyContent,"
+ "SUM(CASE
WHEN   60<="+minDate+" and "+minDate+"<120     THEN 1 ELSE 0 END) as twohouContent,"
+ "SUM(CASE
WHEN   120<="+minDate+"   THEN 1 ELSE 0 END) as twohouMoreContent,"
+ "t.resource_id,t.mo,t.province from "
+ "(SELECT a.*,b.resource_id from t_order a LEFT JOIN  t_channel b ON a.channel_id = b.id where a.update_time>=? and  a.update_time<=?  and  a.status != 'init') t "
+ "GROUP BY t.mo,t.province,t.resource_id";

List<OrderChannelQualityDTO> list = jdbcTemplate.query(selectSQL, new Object[]{fromTime,toTime},new ARPerformanceDTORowMapper());
if(log.isInfoEnabled()){

            //log.info(" sql:{} time:{} return:{} ", selectSQL,startTime,list);

        }
return list;
}

 private class ARPerformanceDTORowMapper implements RowMapper<OrderChannelQualityDTO> {

        @Override

        public OrderChannelQualityDTO mapRow(ResultSet rs, int rowNum) throws SQLException {

        OrderChannelQualityDTO tarPerformance = new OrderChannelQualityDTO();

        tarPerformance.setSuccessCount(rs.getInt("successCount"));

        tarPerformance.setFailCount(rs.getInt("failCount"));

        tarPerformance.setFiveCount(rs.getInt("fiveCount"));

        tarPerformance.setTenCount(rs.getInt("tenCount"));

        tarPerformance.setThirtyContent(rs.getInt("thirtyContent"));

        tarPerformance.setSixtyContent(rs.getInt("sixtyContent"));

        tarPerformance.setTwohourContent(rs.getInt("twohouContent"));

        tarPerformance.setTwohourMoreContent(rs.getInt("twohouMoreContent"));

        tarPerformance.setResourceId(rs.getString("resource_id"));

        tarPerformance.setProvince(rs.getString("province"));

        tarPerformance.setMo(rs.getString("mo"));

            return tarPerformance;

        }

    }

2.@query

在持久层内，重点是@query查出是一个List<Object[]> ，没有属性名称对应，只是查出object[]，索引值是对应sql的顺序。

String minDate = "round((UNIX_TIMESTAMP(t.update_time)-UNIX_TIMESTAMP(t.created_time))/60)";
@Query(value = "SELECT SUM(IF(t.status ='success',1,0))as successCount,SUM(IF(t.status ='fail',1,0))as failCount ,"
+ "SUM(CASE WHEN  0<="+minDate+" and "+minDate+"<5 THEN 1 ELSE 0 END) as fiveCount,"
+ "SUM(CASE
WHEN  5<="+minDate+" and "+minDate+"<1 THEN 1 ELSE 0 END) as tenCount,"
+ "SUM(CASE
WHEN  10<="+minDate+" and "+minDate+"<30     THEN 1 ELSE 0 END) as thirtyContent,"
+ "SUM(CASE
WHEN   30<="+minDate+" and "+minDate+"<60     THEN 1 ELSE 0 END) as sixtyContent,"
+ "SUM(CASE
WHEN   60<="+minDate+" and "+minDate+"<120     THEN 1 ELSE 0 END) as twohouContent,"
+ "SUM(CASE
WHEN   120<="+minDate+"   THEN 1 ELSE 0 END) as twohouMoreContent,"
+ "t.resource_id,t.mo,t.province from "
+ "(SELECT a.*,b.resource_id from t_order a LEFT JOIN  t_channel b ON a.channel_id = b.id where a.update_time>=:fromDate and  a.update_time<=:nextDate  and  a.status != 'init') t "
+ "GROUP BY t.mo,t.province,t.resource_id"
,nativeQuery=true)
List<Object[]> objectFlowList(@Param("fromDate") Date fromDate,@Param("nextDate") Date nextDate);

在 service

 //查出昨日流量渠道统计
   List<Object[]> channelQualityObjectArrayDTO = channelQualityRepo.objectFlowList(fromDate, nextDate);
   List<TChannelQuality> channelQualityList = new ArrayList<TChannelQuality>();
   if(channelQualityObjectArrayDTO != null && channelQualityObjectArrayDTO.size() >0){
    for(Object[] channelQualityObjectParamDTO :channelQualityObjectArrayDTO){
    TChannelQuality channelQuality = new TChannelQuality();
    channelQuality.setDay(dateFormat.parse(dateString));
    channelQuality.setBizType("flow");
    channelQuality.setSuccessCount(getInt((BigDecimal)channelQualityObjectParamDTO[0]));
    channelQuality.setFailCount(getInt((BigDecimal)channelQualityObjectParamDTO[1]));
    channelQuality.setFiveCount(getInt((BigDecimal)channelQualityObjectParamDTO[2]));
    channelQuality.setTenCount(getInt((BigDecimal)channelQualityObjectParamDTO[3]));
    channelQuality.setThirtyContent(getInt((BigDecimal)channelQualityObjectParamDTO[4]));
    channelQuality.setSixtyContent(getInt((BigDecimal)channelQualityObjectParamDTO[5]));
    channelQuality.setTwohourContent(getInt((BigDecimal)channelQualityObjectParamDTO[6]));
    channelQuality.setTwohourMoreContent(getInt((BigDecimal)channelQualityObjectParamDTO[7]));
    channelQuality.setResourceId(((String)channelQualityObjectParamDTO[8]));
    channelQuality.setMo(((String)channelQualityObjectParamDTO[9]));
    channelQuality.setProvince(((String)channelQualityObjectParamDTO[10]));
    channelQualityList.add(channelQuality);
    }
   List<TChannelQuality> result =channelQualityRepo.save(channelQualityList);

总结：

JdbcTemplate 是下下策，头让改的时候，竟然问了句为什么JdbcTemplate 不好？感觉好low端。。。。，jpa是一个很强大的持久层框架。