我正在尝试解析我的 SQL 表之一中的嵌套 JSON 列。嵌套列在多个地方都有一个键值“text”。我希望这个键“文本”的所有值在一个单元格中彼此相邻。
以下是我的示例数据。
{
"version": 1,
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Please assign ticket to RCA team."
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "RCA Team"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tClient code \u2013 5374 \u12345 Apple Financial"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tDividend Task ID \u2013 12345"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPrevious MI IR Number - 12345"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Against the above task please can I have the following data:"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tHURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPartyURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tEligible Share Balance"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCash Currency Dividend Due"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tTax Deducted"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tholder name including any joint holders"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCurrent full address"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tShare balance against task"
}
]
}
]
}
尝试了以下代码
SELECT ID, [Key], Texts
FROM TEST_Table
CROSS APPLY OPENJSON(Description)
WITH
(
Texts varchar(max) '$.content.content.text'
) as TextJson;
但什么也没得到。需要帮助请。
我不确定你为什么想要这个，但这是一个解决方案：
DECLARE @json nvarchar(max) = N'{
"version": 1,
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Please assign ticket to RCA team."
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "RCA Team"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tClient code \u2013 5374 \u12345 Apple Financial"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tDividend Task ID \u2013 12345"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPrevious MI IR Number - 12345"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Against the above task please can I have the following data:"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tHURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPartyURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tEligible Share Balance"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCash Currency Dividend Due"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tTax Deducted"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tholder name including any joint holders"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCurrent full address"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tShare balance against task"
}
]
}
]
}';
WITH cte AS (
SELECT cast(-1 AS bigint) AS parentKey, cast([key] AS bigint) AS [currentKey], value, 1 AS level
FROM OPENJSON(@json, '$.content') o
UNION ALL
SELECT c.currentKey, cast(o.[key] AS bigint), o.value, c.level + 1
FROM cte c
CROSS apply openjson(c.value, '$.content') o
)
SELECT string_agg(json_value(value, '$.text'), ' ') within GROUP (ORDER BY [level], parentKey, currentkey)
FROM cte
WHERE JSON_VALUE(value, '$.type') = 'text'
我使用递归 CTE 调用每个 subjson 部分，然后将其聚合成一个长字符串，按父键和当前键排序。
我使用“ ”作为段落分隔符，但如果您有一些特定需求，您可以使用任何您想要的东西
您可以使用递归 CTE 来提取并解析每个节点中的
contentwith cte as(
select ID,type,text,content
from @table t cross apply openjson(t.Description) with (
type nvarchar(max),
text nvarchar(max),
content nvarchar(max) as json)
as x
union all
select c.ID,u.type,u.text,u.content
from cte as c cross apply Openjson(c.content) with (
type nvarchar(max),
text nvarchar(max),
content nvarchar(max) as json) as u
)
select id,STRING_AGG(text,'\n')
from cte
where type='text'
group by ID
使用此测试 JSON
declare @json nvarchar(max)=N'{
"version": 1,
"type": "doc",
"content": [
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Please assign ticket to RCA team."
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "RCA Team"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tClient code \u2013 5374 \u12345 Apple Financial"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tDividend Task ID \u2013 12345"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPrevious MI IR Number - 12345"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "Against the above task please can I have the following data:"
}
]
},
{
"type": "paragraph",
"content": [
{
"type": "text",
"text": "-\tHURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tPartyURN"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tEligible Share Balance"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCash Currency Dividend Due"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tTax Deducted"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tholder name including any joint holders"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tCurrent full address"
},
{
"type": "hardBreak"
},
{
"type": "text",
"text": "-\tShare balance against task"
}
]
}
]
}';
还有桌子
declare @table table(ID int primary key,Description nvarchar(max))
insert into @table (ID,Description)
VALUES (1,@json),
(2,@json)
这会产生：
id Texts
1 - HURN\n- PartyURN\n- Eligible Share Balance\n- Cash Currency Dividend Due\n- Tax Deducted\n- holder name including any joint holders\n- Current full address\n- Share balance against task\nAgainst the above task please can I have the following data:\n- Client code – 5374 ሴ5 Apple Financial\n- Dividend Task ID – 12345\n- Previous MI IR Number - 12345\nRCA Team\nPlease assign ticket to RCA team.
2 - HURN\n- PartyURN\n- Eligible Share Balance\n- Cash Currency Dividend Due\n- Tax Deducted\n- holder name including any joint holders\n- Current full address\n- Share balance against task\nAgainst the above task please can I have the following data:\n- Client code – 5374 ሴ5 Apple Financial\n- Dividend Task ID – 12345\n- Previous MI IR Number - 12345\nRCA Team\nPlease assign ticket to RCA team.