Question

我正在使用SQL Server查询构造JSON对象。我在斜杠方面遇到问题，而且我还需要特定格式的数据，如下所述。实际输出中带有斜线。我想从输出查询中消除斜线。

    SELECT (SELECT 
         ('{'+'"value":'+ cast(ISNULL(SP.ACCOUNT_TRAFFIC_DEP_LICENSE_NUMBER,'') as varchar) )                       AS TRAFFIC_DEP_LICENSE_NUMBER
        ,(  '{' +'"value":'+cast(ISNULL(SP.NUMBER_OF_VEHICLES   ,'')    as varchar)     )                                   AS NUMBER_OF_VEHICLES
         ,( '{' +'"value":'+cast(ISNULL(SP.ACCOUNT_TRAFFIC_DEP_LICENSE_START_H_DATE,'') as varchar) )                                       AS TRAFFIC_DEP_LICENSE_START_H_DATE
         ,( '{' +'"value":'+cast(ISNULL(SP.ACCOUNT_TRAFFIC_DEP_LICENSE_END_H_DATE   ,'')    as varchar)     )                           AS TRAFFIC_DEP_LICENSE_END_H_DATE
        ,(  '{' +'"value":'+cast(ISNULL(SP.SUSPENSION_BY_MORROR_STATUS,'')  as varchar)         )                                           AS SUSPENSION_BY_MORROR_STATUS
         ,( '{' +'"value":'+cast(ISNULL(SP.SUSPENSION_BY_VIOLATION  ,'')    as varchar)     )                                           AS SUSPENSION_BY_VIOLATION
         ,( '{' +'"value":'+cast(ISNULL(SP.SUSPENSION_BY_PENALTY        ,'')    as varchar)     )                                           AS SUSPENSION_BY_PENALTY
                from [dbo].[TAMM_CAC_ACCOUNTS_STG]  SP 
                where SP.Account_Reference_Number = CAC.Account_Reference_Number
                AND (
                ACCOUNT_TRAFFIC_DEP_LICENSE_NUMBER IS NOT NULL OR 
                NUMBER_OF_VEHICLES IS NOT NULL OR
                SP.ACCOUNT_TRAFFIC_DEP_LICENSE_END_H_DATE   IS NOT NULL OR
                SP.ACCOUNT_TRAFFIC_DEP_LICENSE_END_H_DATE   IS NOT NULL OR
                SP.SUSPENSION_BY_MORROR_STATUS              IS NOT NULL OR
                SP.SUSPENSION_BY_VIOLATION          IS NOT NULL OR      
                SP.SUSPENSION_BY_PENALTY        IS NOT NULL     )   FOR JSON PATH,INCLUDE_NULL_VALUES


 ) AS specifics
            from dbo.Customer_Account CAC

上面的查询给了我想要的输出，但是有反斜杠。

[
  {
    "TRAFFIC_DEP_LICENSE_NUMBER": "{\"value\":0",
    "NUMBER_OF_VEHICLES": "{\"value\":1009",
    "TRAFFIC_DEP_LICENSE_START_H_DATE": "{\"value\":14351229",
    "TRAFFIC_DEP_LICENSE_END_H_DATE": "{\"value\":14351229",
    "SUSPENSION_BY_MORROR_STATUS": "{\"value\":0",
    "SUSPENSION_BY_VIOLATION": "{\"value\":1",
    "SUSPENSION_BY_PENALTY": "{\"value\":1"
  }
]

实际预期输出低于

“specific”:{
"TRAFFIC_DEP_LICENSE_NUMBER":{“value”:0},
            "NUMBER_OF_VEHICLES":{“value”:22},
            "TRAFFIC_DEP_LICENSE_START_H_DATE":{“value”: 14480608},
            "TRAFFIC_DEP_LICENSE_END_H_DATE":{“value”: 14480608},
            "SUSPENSION_BY_MORROR_STATUS":{“value”:0},
            "SUSPENSION_BY_VIOLATION":{“value”:1},
            "SUSPENSION_BY_PENALTY":{“value”:1},
}

Answer 1

在任何使用字符串值的地方，都可以使用STRING_ESCAPE：例如：

SELECT STRING_ESCAPE('\   /  
\\    "     ', 'json') AS escapedText;

返回：

escapedText  
-------------------------------------------------------------  
\\\t\/\n\\\\\t\"\t

Answer 2

如果我正确理解了您的问题，并且希望删除特殊字符的转义，那么下一个解决方案可能会有所帮助。

FOR JSON子句escapes与JSON输出的\中的特殊字符。

...如果源数据包含特殊字符，则为FOR JSON 子句使用\在JSON输出中对其进行转义，如下表。转义在属性名称中都发生和他们的价值观。 ...

如果确定可以生成有效的JSON，则可以尝试使用JSON_QUERY with FOR JSON。

... JSON_QUERY返回有效的JSON片段。结果，FOR JSON 不会在JSON_QUERY返回值中转义特殊字符。 ...

一个简单的示例，演示您的问题：

-- Statement:
SELECT (
   SELECT '{"value": 1}' AS TRAFFIC_DEP_LICENSE_NUMBER
   FOR JSON PATH, INCLUDE_NULL_VALUES
) AS specifics

-- Output:
-------------------------------------------------
specifics
-------------------------------------------------
[{"TRAFFIC_DEP_LICENSE_NUMBER":"{\"value\": 1}"}]

解决方案：

-- Statement:
SELECT (
   SELECT JSON_QUERY('{"value": 1}') AS TRAFFIC_DEP_LICENSE_NUMBER
   FOR JSON PATH, INCLUDE_NULL_VALUES
) AS specifics

-- Output:
-------------------------------------------------
specifics
-------------------------------------------------
[{"TRAFFIC_DEP_LICENSE_NUMBER":{"value": 1}}]

注意：

您可以使用ISJSON检查JSON：

SELECT (
   SELECT CASE
      WHEN ISJSON('{"value": 1}') = 1 THEN JSON_QUERY('{"value": 1}') 
      ELSE 'Imvalid JSON'
   END AS TRAFFIC_DEP_LICENSE_NUMBER
   FOR JSON PATH, INCLUDE_NULL_VALUES
) AS specifics

从JSON SQL查询中删除反斜杠

2 个答案: