Question: I have created a sample xml document containing data conforming to the atom 1.0 schema. When I import the contents of this file (for testing purposes) in PowerPivot, it creates columns for each atom element in each entry, instead of creating a column per content element. Why is this?
Background: A customer wants to import data from a web service which provides a feed that uses a custom XML schema that is not supported by PowerPivot. The service provides the ability for the caller to supply an XSLT template that will be applied to the feed. I am hoping to be able to transform this feed into a valid atom feed thereby allowing the customer to import data into PowerPivot.
Sample atom xml:
<?xml version="1.0" encoding="UTF-8"?><feed xmlns:d="http://schemas.microsoft.com/ado/2007/08/dataservices" xmlns="http://www.w3.org/2005/Atom" xmlns:m="http://schemas.microsoft.com/ado/2007/08/dataservices/metadata"><title type="text">My Data Feed</title><id>http://temp/feed</id><updated>2012-12-13T00:00:00Z</updated><entry><id>http://temp/feed/1</id><title type="text">Title</title><author><name>Author</name></author><updated>2012-12-13T00:00:00Z</updated><content type="application/xml"><d:Name>John Smith</d:Name><d:Address>Address</d:Address><d:Zip>1234</d:Zip></content></entry></feed>When imported into PowerPivot (selecting "From Data Feeds", clicking "Browse" and pointing out the xml file), it looks like this:
I was excpecting three columns: Name, Address and Zip. If I change "Include Atom Elements" from Auto to False in the connection configuration, no columns are imported.