Query on differing XML-BLOB values

[Ylreeb]

Dear Power-BI-Forum

I'm trying to excerpt descriptive metadata from an Enterprise Media Asset Management System stored as binary (blob) within an Oracle-DB. The data itself is an XML.

So far so easy, I proceeded this way:

1. Selected the affected Oracle-table as PowerBI-Import
2. Removed all unneeded columns
3. XML-Parsed the blob-column and removed errors (former blob's with null-values)
4. Transformed one particular item by expanding and pivoting

In order to execute this for all items I just edited the responsible "row filtering" M-Code to "each true". That gives me the following error:  

Expression.Error: We cannot convert the value "xxx" to type Table.
Details:
    Value=xxx
    Type=Type

The reason for that behavior:

descriptive metadata can differ in their amount of values and so in the amount of columns after the BI-query/transformation. That is because the vendor on facilitates some kind of flexibility handling descriptive metadata within the relational database of the Media Asset Management system.

My procedure even works fine with items keeping the same amount of values.

So I'm coming up to you - is there any way reducing the values within the blob/XML (eg. to asset.siteIdentity, sports.sport, asset.materialType, asset.title) as a kind of a defined record/value-subset pivoting to?

I also tried "Combine Files". But transforming the receiving CSV (Sample File) the way Parsed-XML does by using delimiters is quite impossible.

THX in advance

[v-chuncz-msft]

@Ylreeb,

You'll need to do research into The Advanced Editor.