尝试使用 SQL Query 和 OpenJson 读取 JSON 字符串

如何解决尝试使用 SQL Query 和 OpenJson 读取 JSON 字符串

我有一个非常大的 JSON 字符串，我正在尝试使用 SQL 读取某些报告，但是 JSON 中嵌入了至少几个级别的对象，有人可以建议我如何使用一个读取所有数据使用 OpenJson 的 SQL 查询。

我尝试了以下方法

DECLARE @json NVarChar(max) =
(select top 1 Blob
from Vehicles
where Entity = 'Case' and EntityId = 20)

SELECT * FROM OpenJson(@json);

结果看起来像这样我只粘贴了一行，但存在很多这样的行，任何帮助都会非常有帮助 - 提前致谢。

{
   "InputFieldId":500,"Name":"CA Port of Intermodal Location","Description":null,"InputFieldEntity":"VehicleField","InputFieldEntityId":-1,"Group":"Drayage","Type":"dropdown","Order":710,"IsDeleted":false,"IsRequired":false,"InputDropdowns":[
      {
         "InputDropdownId":144,"InputField":null,"InputFieldId":500,"Name":"Crocket C&H Sugar Port","Order":10
      },{
         "InputDropdownId":145,"Name":"Hueneme Port","Order":20
      },{
         "InputDropdownId":146,"Name":"Long Beach Port","Order":30
      },{
         "InputDropdownId":147,"Name":"Oakland Port","Order":40
      },{
         "InputDropdownId":148,"Name":"Los Angeles Port","Order":50
      },{
         "InputDropdownId":149,"Name":"Redwood City Port","Order":60
      },{
         "InputDropdownId":150,"Name":"San Diego Port","Order":70
      },{
         "InputDropdownId":151,"Name":"San Francisco Port","Order":80
      },{
         "InputDropdownId":152,"Name":"Stockton Port","Order":90
      },{
         "InputDropdownId":153,"Name":"West Sacramento Port","Order":100
      },{
         "InputDropdownId":154,"Name":"Georgia Pacific Antioch Port","Order":110
      },{
         "InputDropdownId":155,"Name":"IMTT Port","Order":120
      },{
         "InputDropdownId":156,"Name":"Kindermorgan LA Port","Order":130
      },{
         "InputDropdownId":157,"Name":"Kindermorgan Richmond Port","Order":140
      },{
         "InputDropdownId":158,"Name":"Koch Carbon Long Beach Port","Order":150
      },{
         "InputDropdownId":159,"Name":"Mitsubishi Cement Long Beach. Etc. Port","Order":160
      },{
         "InputDropdownId":160,"Name":"UP City of Industry","Order":170
      },{
         "InputDropdownId":161,"Name":"UP ICTF","Order":180
      },{
         "InputDropdownId":162,"Name":"UP LATC","Order":190
      },{
         "InputDropdownId":163,"Name":"UP Lathrop","Order":200
      },{
         "InputDropdownId":164,"Name":"UP LA","Order":210
      },{
         "InputDropdownId":165,"Name":"UP Oakland","Order":220
      },{
         "InputDropdownId":166,"Name":"BNSF Oakland","Order":230
      },{
         "InputDropdownId":167,"Name":"BNSF LA","Order":240
      },{
         "InputDropdownId":168,"Name":"BNSF Richmond","Order":250
      },{
         "InputDropdownId":169,"Name":"BNSF San Bernardino","Order":260
      },{
         "InputDropdownId":170,"Name":"BNSF Stockton","Order":270
      }
   ],"InputValue":{
      "InputValueId":0,"InputValueEntity":"Vehicle","InputValueEntityId":0,"Value":null,"InputDropdown":null,"InputDropdownId":null,"DateCreated":"0001-01-01T08:00:00Z","DateUpdated":"0001-01-01T08:00:00Z","CreatedByUserId":null,"UpdatedByUserId":null,"CreatedByUser":null,"UpdatedByUser":null
   },"isHidden":false
}

我正在寻找一个查询，它可以将所有数据显示为一个表格格式，我进入字符串的所有子部分并将结果放入一个表格中。请提供任何帮助，即使行具有重复值，因为某些父对象值会为子对象重复。对任何 json 字符串运行应该是一个通用查询，因为这里我们不知道嵌套了多少对象。

查询必须动态地给我列和行，这取决于 json 中字段的深度和数量。例如，如果 Continent、Country、States 和 Cities，那么对于每个 City，State、Country 和 Continent 字段将重复，但问题是，我们不确切知道深度和列，它可能会因每个 jSon 字符串而改变，查询必须能够毫无故障地为所有 json 字符串运行 - 我的朋友有什么帮助吗？

解决方法

当你使用openjson解析根级别的列时，嵌入的数组应该被解析为json列。然后你可以再次使用 openjson 进一步解析这个 json 列。请查看示例中如何解析 col_2。

对于同级别的嵌入对象，可以通过在openjson中指定字段路径直接解析出它们的列。请查看示例中如何解析 col_3。

这是例子-

declare 
  @json nvarchar(max) = '[
      {"col_1":"c1-1","col_2":[{"col_2_1":"c21-1-1","col_2_2":"c22-1-1"},{"col_2_1":"c21-1-2","col_2_2":"c22-1-2"}],"col_3":{"col_3_1":"c31-1","col_3_2":"c32-1"}
      },{"col_1":"c1-2","col_2":[{"col_2_1":"c21-2-1","col_2_2":"c22-2-1"},{"col_2_1":"c21-2-2","col_2_2":"c22-2-2"}],"col_3":{"col_3_1":"c31-2","col_3_2":"c32-2"}
      }
  ]';

with d1 as (
select *
  from openjson(@json) with (
         col_1 nvarchar(max),col_2 nvarchar(max) as json,col_3_1 nvarchar(max) '$.col_3.col_3_1',col_3_2 nvarchar(max) '$.col_3.col_3_2'
       )
)
select d1.col_1,d1.col_3_1,d1.col_3_2,d2.col_2_1,d2.col_2_2
  from d1 cross apply openjson(d1.col_2) with (col_2_1 nvarchar(max),col_2_2 nvarchar(max)) d2

输出将是 -

c1-1    c31-1   c32-1   c21-1-1   c22-1-1
c1-1    c31-1   c32-1   c21-1-2   c22-1-2
c1-2    c31-2   c32-2   c21-2-1   c22-2-1
c1-2    c31-2   c32-2   c21-2-2   c22-2-2

您需要确保以这种方式获得的数据对您的业务有意义。