Import XML to SQL table and format data - sql

Question

Would someone be able to help me I am trying to import an xml file to a sql server table. I can import the required data but having some trouble getting it in the format required.
declare #input XML = '<Sub>
<Results>
<Result>
<ids>
<id>
<type>code</type>
<value>9004a3d2</value>
</id>
<id>
<type>username</type>
<value>jbloggs001</value>
<date>20160725</date>
</id>
<id>
<type>EmployeeID</type>
<value>01234</value>
<date>20160725</date>
</id>
</ids>
</Result>
</Results>
</Sub>'
SELECT
datatype = XCol.value('(type)[1]','varchar(25)'),
datavalue = XCol.value('(value)[1]','varchar(50)')
FROM
#input.nodes('/Sub/Results/Result/ids/id') AS XTbl(XCol)
This gives 3 columns like:
datatype datavalue
--------------------------------
code 9004a3d2
username jbloggs001
employeeID 01234
Would it be possible to get it to import as?
EmployeeID USername Code
---------------------------------
01234 jbloggs 0019004a3d2
Thanks

Answer 1

As was suggested in comments that could be done using PIVOT:
SELECT *
FROM (
SELECT
datatype = XCol.value('(type)[1]','varchar(25)'),
datavalue = XCol.value('(value)[1]','varchar(50)')
FROM
#input.nodes('/Sub/Results/Result/ids/id') AS XTbl(XCol)
) as p
PIVOT (
MAX(datavalue) FOR datatype IN (EmployeeID,username,code)
) as pvt
Output:
EmployeeID username code
01234 jbloggs001 9004a3d2
If input is always with same types:
SELECT XCol.value('(id/value)[3]','varchar(50)') as EmployeeID,
XCol.value('(id/value)[2]','varchar(50)') as username,
XCol.value('(id/value)[1]','varchar(50)') as code
FROM #input.nodes('/Sub/Results/Result/ids') AS XTbl(XCol)

Answer 2

Another option is XQuery contains function
declare #input XML = '<Sub>
<Results>
<Result>
<ids>
<id>
<type>code</type>
<value>9004a3d2</value>
</id>
<id>
<type>username</type>
<value>jbloggs001</value>
<date>20160725</date>
</id>
<id>
<type>EmployeeID</type>
<value>01234</value>
<date>20160725</date>
</id>
</ids>
</Result>
</Results>
</Sub>';
SELECT
code = XCol.value('(id[contains((./type)[1],"code")]/value)[1]','varchar(50)'),
username = XCol.value('(id[contains((./type)[1],"username")]/value)[1]','varchar(50)'),
EmployeeID = XCol.value('(id[contains((./type)[1],"EmployeeID")]/value)[1]','varchar(50)')
FROM
#input.nodes('/Sub/Results/Result/ids') AS XTbl(XCol);

Answer 3

Below query works for your solution
Select [EmployeeID],[Username],[Code]
FROM #TEMP
PIVOT
(
MAX(DATAVALUE) FOR DATATYPE IN (code,username,employeeID)
)A

Answer 4

Another very straight and typesafe approach was this:
SELECT id.value('(id[type="code"]/value)[1]','varchar(max)') AS code
,id.value('(id[type="username"]/value)[1]','varchar(max)') AS username
,id.value('(id[type="EmployeeID"]/value)[1]','varchar(max)') AS EmployeeID
FROM #input.nodes('Sub/Results/Result/ids') AS A(id)