如何总结这些值并将其转换为 varchar 以便我可以在 SQL Server 2005 中以 JSON 格式获得输出？答案

【问题标题】：How to sum up the values and also cast it to varchar so that I can get output in JSON format in SQL server 2005?如何总结这些值并将其转换为 varchar 以便我可以在 SQL Server 2005 中以 JSON 格式获得输出？
【发布时间】：2016-02-01 07:10:50
【问题描述】：

我想总结收入的值并按产品名称对其进行分组。我使用了以下方式。但是我无法汇总这些值并将其同时转换为 varchar，因为只有将其转换为 varchar 时才能获得 JSON 格式的结果。这是我的查询

select '[' + STUFF(( select ',{"id":' + cast(id as varchar(max)) + ',"ProductName":"' + ProductName + '"' + ',"Revenue":'

+ cast(Revenue as varchar(max)) + ',"InYear":' + CONVERT(VARCHAR(10), InYear, 111) +'}' 来自 dummyRevenueData 其中 InYear>='2014/01/01' 和 InYear

以上查询结果：

[{"id":2,"ProductName":"P1","Revenue":700,"InYear":2014/01/01},{"id":3,"ProductName":"p1","Revenue":800,"InYear":2015/01/01},{"id":4,"ProductName":"p1","Revenue":900,"InYear":2016/01/01},{"id":6,"ProductName":"P2","Revenue":600,"InYear":2014/01/01},{"id":7,"ProductName":"P2","Revenue":200,"InYear":2015/01/01},{"id":8,"ProductName":"P2","Revenue":400,"InYear":2016/01/01},{"id":10,"ProductName":"p3","Revenue":100,"InYear":2014/01/01},{"id":11,"ProductName":"p3","Revenue":500,"InYear":2015/01/01},{"id":12,"ProductName":"p3","Revenue":400,"InYear":2016/01/01}]

我想总结收入，我的输出应该在

[{"ProductName":"P1","Revenue":2400},

{"ProductName":"p2","Revenue":1000}, {"ProductName":"p3","Revenue":1200}]

我尝试了以下查询，但没有得到结果：

 select '[' + STUFF(( select ',{"id":' + cast(id as varchar(max)) + ',"ProductName":"' + ProductName + '"' + ',"Revenue":'

+ cast((sum(Revenue) as Rev) as varchar(max)) + ',"InYear":' + CONVERT(VARCHAR(10), InYear, 111) +'}' 来自 dummyRevenueData 其中 InYear>='2014/01/01' 和 InYear

【问题讨论】：

标签： sql-server json casting

【解决方案1】：

为什么不在第一次选择时使用 GROUP BY 来检索数据

然后使用 CAST 获取 JSON 格式的结果

SELECT CONCAT('[', GROUP_CONCAT(jsonEntry SEPARATOR ','), ']') AS jsonEntries
FROM 
  (
      SELECT '{'+ 
               '",ProductName":"' + ProductName + '"' + 
               ',"Revenue":"' + CAST(SumRevenue as varchar(SumRevenue)) + '"' + 
             '}' AS jsonEntry
      FROM
      (
           SELECT id, ProductName, SUM(Revenue) as SumRevenue 
           FROM dummyRevenueData 
           WHERE InYear>='2014/01/01' AND InYear<='2016/01/01'
           GROUP BY ProductName
      ) AS rawData

  ) AS stringData

虚拟收入数据：

id | ProductName | Revenue | InYear
---------------------------------------
2  | p1          | 700     | 2014/01/01
3  | p1          | 800     | 2015/01/01
4  | p1          | 900     | 2016/01/01
6  | P2          | 600     | 2014/01/01
7  | P2          | 200     | 2015/01/01
8  | P2          | 400     | 2016/01/01
10 | p3          | 100     | 2014/01/01
11 | p3          | 500     | 2015/01/01
12 | p3          | 400     | 2016/01/01

原始数据：

ProductName | SumRevenue
------------------------
p1          | 2400
P2          | 1200
p3          | 1000

字符串数据：

jsonEntry
------------------------------------
{"ProductName":"p1","Revenue":"2400"}
{"ProductName":"P2","Revenue":"1200"}
{"ProductName":"p3","Revenue":"1000"}

jsonEntries：

[{"ProductName":"p1","Revenue":"2400"},{"ProductName":"P2","Revenue":"1200"},{"ProductName":"p3","Revenue":"1000"}]

在dummyRevenueData中添加ProductId，这样就可以按ProductId分组了

【讨论】：