XML Workshop : Utilizing Relational Data In XML Files

Question

Post reply

XML Workshop : Utilizing Relational Data In XML Files

Matthew Pettit

SSC Enthusiast

Points: 111
More actions
April 24, 2014 at 12:04 am

#286054

Comments posted to this topic are about the item XML Workshop : Utilizing Relational Data In XML Files
[font="Arial"]Matthew[/font]

Viewing 7 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic. Login to reply

Eirikur Eiriksson SSC Guru Points: 182901 More actions · Answer 1

Hi Matthew,

and thank you for this article.

Few comments, firstly the sp_xml_preparedocument and sp_xml_removedocument have some limitations and quite an overhead, I'm very hesitant recommending / implementing them on a production system. XQuery on the XML data type is without those limitations and in terms of importing XML data, only bound by the 2Gb size limit of the XML variable.

Secondly, the SQL Server uses UTF-16 as the encoding for the XML data type, casting from any other encoding will fail with an error. To bypass this, use OPENROWSET(BULK, [FILE_PATH], SINGLE_BLOB) without any typecast, to directly load the data into the destination table as this ignores any encoding specified in the source file.

Import example using XQuery:

USE tempdb;

GO

DECLARE @XMLTABLE TABLE

(

XML_ID INT IDENTITY(1,1) PRIMARY KEY CLUSTERED NOT NULL

,XML_DATA XML NOT NULL

);

INSERT INTO @XMLTABLE (XML_DATA)

SELECT BulkColumn FROM OPENROWSET

(

BULK

'C:\ImportXML\XMLDATA01.xml'

,SINGLE_BLOB

) AS IMPXML

UNION ALL

SELECT BulkColumn FROM OPENROWSET

(

BULK

'C:\ImportXML\XMLDATA02.xml'

,SINGLE_BLOB

) AS IMPXML

SELECT

XT.XML_ID

,PATI.ENT.value('id[1]' ,'INT') AS PATIENT_ID

,PATI.ENT.value('name[1]' ,'NVARCHAR(128)') AS PATIENT_NAME

,PATI.ENT.value('birthdate[1]','DATE') AS PATIENT_BIRTHDATE

FROM @XMLTABLE XT

OUTER APPLY XT.XML_DATA.nodes('root/patient') AS PATI(ENT);

😎

AllenMWhite Ten Centuries Points: 1198 More actions · Answer 2

PowerShell is really a more efficient way to work with XML. Transact-SQL has never really been good at handling string data of any sort, and while XQuery is significantly better than OpenXML, PowerShell just runs circles around both of them.

I recently wrote an article demonstrating this at SQLMag.com.

http://sqlmag.com/powershell/use-powershell-load-xml-data-sql-server

Allen

Eirikur Eiriksson SSC Guru Points: 182901 More actions · Answer 3

It would be interesting to see the performance of PS when working with sets larger than the 2Gb limits of the XML data type. 😎

AllenMWhite Ten Centuries Points: 1198 More actions · Answer 4

Wouldn't you have the same problem in Transact-SQL, though?

Eirikur Eiriksson SSC Guru Points: 182901 More actions · Answer 5

AllenMWhite (4/24/2014)
Wouldn't you have the same problem in Transact-SQL, though?

Yes of course, that is why I am interested in seeing how PS would handle this. Often had to implement all kinds of workarounds when the sets are tens or hundreds of Gbs.

😎

JunkIt SSCommitted Points: 1554 More actions · Answer 6

Nice article

I would add that OPENXML is not currently supported in an Azure SQL database, so if you're planning on migrating any time soon you'll want to stick with the XML data type. Given the overhead I've read about using OPENXML, I'm not sure this is something Microsoft ever plans on supporting.