Question

表trans_data的列tran_type，其值为＆＃39; V＆＃39; M＆＃39; M＆＃39;＆＃39; MC＆＃39;＆＃39; P＆＃39;

目前使用多个联接：

(          
    select count(tran_type) Visa, to_char(request_datetime,'hh24') src_hour from trans_data where tran_type = 'V' group by to_char(request_datetime,'hh24')
  ) b on a.src_hour = b.src_hour left join

  (
    select count(tran_type) Maestro, to_char(request_datetime,'hh24') src_hour from trans_data where switched_type = 'M' group by to_char(request_datetime,'hh24')                   
  ) c on  a.src_hour = c.src_hour left join

上述块需要重复进行其他输入，例如＆＃39; MC＆＃39; - 万事达卡和＆＃39; P＆＃39;贝宝。将来，可能会有更多的交易类型。有没有办法，上面的代码可以根据tran_type的数量动态创建？

对于前。如果明天，我们包括＆＃39; P＆＃39; PayPal，动态查询应如下所示：

(          
    select count(tran_type) Visa, to_char(request_datetime,'hh24') src_hour from trans_data where tran_type = 'V' group by to_char(request_datetime,'hh24')
  ) b on a.src_hour = b.src_hour left join

  (
    select count(tran_type) Maestro, to_char(request_datetime,'hh24') src_hour from trans_data where tran_type = 'M' group by to_char(request_datetime,'hh24')                   
  ) c on  a.src_hour = c.src_hour left join

(
    select count(tran_type) PayPal, to_char(request_datetime,'hh24') src_hour from trans_data where tran_type = 'P' group by to_char(request_datetime,'hh24')                   
  ) d on  a.src_hour = d.src_hour left join

Answer 1

对于此类查询类型，您应该在select语句中使用PIVOT子句。请求的tran_type值列表可以在“for tran_type in”列表中动态更改。

select * from
(select to_char(request_datetime,'hh24') src_hour, tran_type from trans_data)
pivot
(
   count(tran_type)
   for tran_type in ('V' as "Visa",'M' as "Maestro",'P' as "PayPal")
)
order by src_hour

PL / SQL-基于列值创建动态查询块

1 个答案: