2

My source data in a SQL Server table has XML structured data. For each of the source rows I need to query the following sample XML structure in a way that I get a calculated value of all A sub elements. For example the SUM of bb and product of cc * dd over all A elements. The number of A elements in the structure is unknown.

<root>
 <A>
 <bb>bb1</bb>
 <cc>cc1</cc>
 <dd>dd1</dd>
 </A>
 <A>
 <bb>bb2</bb>
 <cc>cc2</cc>
 <dd>dd2</dd>
 </A>
 <A>
 <bb>bb3</bb>
 <cc>cc3</cc>
 <dd>dd3</dd>
 </A>
</root>

I experimented with something like this, but it looks like not being the right way:

SELECT Convert(xml,myXmlCol).value('(/root/A[1]/bb)[1]', 'int') as MyResult FROM SourceTable
Tom V
15.8k7 gold badges66 silver badges87 bronze badges
asked Feb 9, 2016 at 14:58
0

1 Answer 1

1

The below query uses this sample data:

Declare @xml1 xml = CAST('<root>
 <A><bb>1</bb><cc>6</cc><dd>3</dd></A>
 <A><bb>2</bb><cc>5</cc><dd>6</dd></A>
 <A><bb>3</bb><cc>4</cc><dd>9</dd></A>
</root>' as xml);
Declare @xml2 xml = CAST('<root>
 <A><bb>4</bb><cc>3</cc><dd>7</dd></A>
 <A><bb>5</bb><cc>2</cc><dd>4</dd></A>
 <A><bb>6</bb><cc>1</cc><dd>1</dd></A>
</root>' as xml);
Declare @data TABLE(id int identity(0, 1), x xml)
Insert Into @data values(@xml1), (@xml2)

It declares 2 different variables and insert in in a dummy table. This is only a table with 2 rows:

0 <root><A><bb>1</bb><cc>6</cc><dd>3</dd></A>...</A></root>
1 <root><A><bb>4</bb><cc>3</cc><dd>7</dd></A>...</A></root>

Query:

SELECT d.id
 , xml.node.value('./bb[1]', 'varchar(50)') as colBB
 , xml.node.value('./cc[1]', 'varchar(50)') as colCC
 , xml.node.value('./dd[1]', 'varchar(50)') as colDD
FROM @data d
CROSS APPLY x.nodes('/root/A') as xml(node)

It read each row from the table and extract XML data.

Output:

id colBB colCC colDD
0 1 6 3
0 2 5 6
0 3 4 9
1 4 3 7
1 5 2 4
1 6 1 1

If can then be used with a GROUP BY:

SELECT id
 , SUM(colBB) as BB
 , SUM(colCC * colDD) as CC*DD
FROM (
 SELECT d.id
 , xml.node.value('./bb[1]', 'int') as colBB
 , xml.node.value('./cc[1]', 'int') as colCC
 , xml.node.value('./dd[1]', 'int') as colDD
 FROM @data d
 CROSS APPLY x.nodes('/root/A') as xml(node)
) x
GROUP BY id;

Output

id BB CC*DD
0 6 84
1 15 30
answered Feb 9, 2016 at 15:33
0

Your Answer

Draft saved
Draft discarded

Sign up or log in

Sign up using Google
Sign up using Email and Password

Post as a guest

Required, but never shown

Post as a guest

Required, but never shown

By clicking "Post Your Answer", you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.