Question

我的表格XML中的data数据类型列records中存储了一个XML文件。

表格如下：

create table records 
(
     id int,
     type nvarchar(28),
     data xml,
     posted datetime
)

XML数据：

<Properties>
    <data>
        <Name>novel</Name>
        <Gender>Female</Gender>
        <Age>32</Age>
        <Salary>55k</Salary>
        <Phone>123-123</Phone>
    </data>
</Properties>

我目前正在使用以下查询从该XML列中提取数据，该数据在20K记录中花费的时间超过了几分钟。

select
    id,
    posteddate,
    CONVERT( NVARCHAR(500), data.query('data(Properties/data/Name)') ) AS Name,
    CONVERT( NVARCHAR(500), data.query('data(Properties/data/Gender)') ) AS Gender,
    CONVERT( NVARCHAR(500), data.query('data(Properties/data/Age)') ) AS Age,
    CONVERT( NVARCHAR(500), data.query('data(Properties/data/Salary)') ) AS Salary,
    CONVERT( NVARCHAR(500), data.query('data(Properties/data/Phone)') ) AS Phone
from 
    records
where 
    type = 'personnel_xml'

我是SQL Server中XML提取的新手。我想我没有遵循提取XML元素的标准优化方法。

所以，任何人都可以帮助我如何优化这种情况，因为我需要从存储为列的xml中提取100个这样的元素。

Answer 1

要从SQL Server中获取XML值，您应该使用value() Method (xml Data Type)。对于无类型XML，您应指定text()节点以获得更好的性能。

select R.id,
       R.posted,
       R.data.value('(/Properties/data/Name/text())[1]', 'nvarchar(500)') as Name,
       R.data.value('(/Properties/data/Gender/text())[1]', 'nvarchar(10)') as Gender,
       R.data.value('(/Properties/data/Age/text())[1]', 'int') as Age,
       R.data.value('(/Properties/data/Salary/text())[1]', 'nvarchar(10)') as Salary,
       R.data.value('(/Properties/data/Phone/text())[1]', 'nvarchar(30)') as Phone
from dbo.records as R
where type = N'personnel_xml';

Answer 2

假设xml中有多个<data>。注意我添加了一个扩展的xml文件，该文件有两组。

Declare @table table (id int,data xml)
Insert Into @table values (1,'<Properties><data><Name>novel</Name><Gender>Female</Gender><Age>32</Age><Salary>55k</Salary><Phone>123-123</Phone></data>
<data><Name>Another Name</Name><Gender>Male</Gender><Age>45</Age><Salary>75k</Salary><Phone>555-1212</Phone></data>
</Properties>')

;with cte as (
      Select ID
            ,RN   = Row_Number() over (Partition By ID Order By (Select Null))
            ,Data = m.query('.') 
      From   @table AS t
      Cross Apply t.Data.nodes('/Properties/data') AS A(m)
 )
Select ID
      ,RN
      ,Name   = Data.value('(data/Name)[1]'  ,'nvarchar(500)')
      ,Gender = Data.value('(data/Gender)[1]','nvarchar(500)')
      ,Age    = Data.value('(data/Age)[1]'   ,'nvarchar(500)')
      ,Salary = Data.value('(data/Salary)[1]','nvarchar(500)')
      ,Phone  = Data.value('(data/Phone)[1]' ,'nvarchar(500)')
 From  cte

返回

ID  RN  Name            Gender  Age     Salary  Phone
1   1   novel           Female  32      55k     123-123
1   2   Another Name    Male    45      75k     555-1212

SQL Server查询XML数据类型性能问题

2 个答案: