我需要帮助解析T-SQL中的XML

Can someone help me to parse the following XML in T-SQL (SQL Server 2005)?

有人可以帮我解析T-SQL中的以下XML(SQL Server 2005)吗?

<Tx>
    <T>1</T>
    <C>1</C>
    <T>2</T>
    <C>1</C>
    <T>3</T>
    <C>1</C>
    <T>4</T>
    <C>1</C>
</Tx>

I tried the following:

我尝试了以下方法:

SELECT
        Tx.query('T').value('.', 'varchar(10)') AS [Column 1],
        Tx.query('C').value('.', 'varchar(10)') AS [Column 2]
    FROM @MyXml.nodes('Tx') x(Tx)

but it didn't work, as I get the following result:

但它没有用,因为我得到以下结果:

Column 1    Column 2
--------    --------
1234        1111

while expecting (what I want to achieve):

期待(我想要实现的目标):

Column 1    Column 2
--------    --------
1           1
2           1
3           1
4           1

The string is obviously a valid XML, but is it valid for T-SQL?

该字符串显然是一个有效的XML,但它对T-SQL有效吗?

I'll be grateful too if someone can provide a reference that explains how XML works in T-SQL.

如果有人能提供解释XML如何在T-SQL中工作的参考,我将不胜感激。

Thanks in advance.

提前致谢。

3 个解决方案

#1

Here is a way to pair the values by position in. First T is matched against first C and so on.

这是一种按位置对值进行配对的方法。首先将T与第一个C匹配,依此类推。

declare @XML xml = 
'<Tx>
    <T>1</T>
    <C>4</C>
    <T>2</T>
    <C>3</C>
    <T>3</T>
    <C>2</C>
    <T>4</T>
    <C>1</C>
</Tx>'

select @XML.value('(/Tx/T[position() = sql:column("N.number")])[1]', 'int') as Column1,
       @XML.value('(/Tx/C[position() = sql:column("N.number")])[1]', 'int') as Column2
from master..spt_values as N
where N.type = 'P' and
      N.number between 1 and @XML.value('max((count(/Tx/T), count(/Tx/C)))', 'int')

With a different structure of the XML like suggested by marc_s the query is much simpler.

使用marc_s建议的不同XML结构,查询更简单。

declare @XML xml = 
'<Tx>
  <row>
    <T>1</T>
    <C>4</C>
  </row>
  <row>
    <T>2</T>
    <C>3</C>
  </row>
  <row>
    <T>3</T>
    <C>2</C>
  </row>
  <row>
    <T>4</T>
    <C>1</C>
  </row>
</Tx>'

select T.R.value('T[1]', 'int') as Column1,
       T.R.value('C[1]', 'int') as Column2
from @XML.nodes('/Tx/row') as T(R)

#2

What are you trying to achieve??

你想要实现什么?

Your current select statement will list out all the <T> elements below <Tx> - but only those - not the <C> nodes, too - and you seem to want to get the values from both kinds of subnodes - right??

您当前的select语句将列出下面的所有元素 - 但只有那些 - 而不是节点 - 并且您似乎想从两种子节点中获取值 - 对吗?

This will give you all the elements from the <T> nodes - is that what you're looking for?

这将为您提供节点中的所有元素 - 您正在寻找什么?

SELECT
    Tx.x.value('.', 'varchar(10)') AS [Column 1]
FROM @MyXml.nodes('Tx/T') Tx(x)

This will give you all the elements from inside the <Tx> node, including their "type" (C or T) - is that what you're looking for?

这将为您提供节点内的所有元素,包括它们的“类型”(C或T) - 您正在寻找什么?

SELECT
    ColumType = Tx.x.value('local-name(.)', 'varchar(10)'),
    ColumnValue = Tx.x.value('.', 'varchar(10)')
FROM @MyXml.nodes('Tx/*') Tx(x)

Update: as resources for learning how to use XQuery in SQL Server, I'd recommend

更新:作为学习如何在SQL Server中使用XQuery的资源,我建议

SQL Server 2005 XQuery and XML-DML

SQL Server 2005 XQuery和XML-DML

Introduction to XQuery in SQL Server 2005

SQL Server 2005中的XQuery简介

Your XML is not well suited to do what you're trying to do - there's nothing that "keeps together" the T and C element that obviously belong together - so there's really nothing XQuery can do to parse this the way you want it to.

你的XML并不适合做你想要做的事情 - 没有任何东西可以“整合”明显属于一起的T和C元素 - 所以XQuery根本无法以你想要的方式解析它。

If you had the XML something like this:

如果你有这样的XML:

<Tx>
   <Pair>
      <T>....</T>
      <C>....</C>
   </Pair>
   <Pair>
      <T>....</T>
      <C>....</C>
   </Pair>
   ....
</Tx>

then you could get the list of Tx/Pair nodes and get the T and C elements from that XML fragment. But right now - all you can do is parse out all 8 subnodes of <Tx> and display their values - that's really all you can do

然后你可以得到Tx / Pair节点列表,并从该XML片段中获取T和C元素。但是现在 - 你所能做的只是解析的所有8个子节点并显示它们的值 - 这就是你所能做的一切

#3

Use this:

select a.t, b.c
from
(
    select t.c.value('.[1]', 'int') [t]
        , ROW_NUMBER() OVER(ORDER BY t.c) [rn]
    from @MyXml.nodes('Tx/T') t(c)
)a
join
(
    select t.c.value('.[1]', 'int') [c]
        , ROW_NUMBER() OVER(ORDER BY t.c) [rn]
    from @MyXml.nodes('Tx/C') t(c)
)b on b.rn = a.rn

or another way:

或另一种方式:

select t.t [T]
    , @MyXml.value('(Tx/C[position() = sql:column("rn")])[1]', 'int') [C]
from
(
    select t.c.value('.[1]', 'int') [t]
        , ROW_NUMBER() OVER(ORDER BY t.c) [rn]
    from @MyXml.nodes('Tx/T') t(c)
)t

#1