Catalogging Table - Handling Email Attachments

[aputhuraya]

Hi Fabric Community,

My requirement is to index the attachments from incoming emails.

For this purpose, the current design in Azure is the use of a SQL table, where a logic app adds a row for every attachment along with metadata like subject, sender and filename. Then, on a schedule, a stored procedure looks at every row and it's metadata to index it based on fixed rules on subject/sender/filename.

Now I want to do the same in Microsoft Fabric, but am unable to overcome challenges with having this table set up either as a Lakehouse Delta Table or a Warehouse table.

When I try a Lakehouse Delta Table:

• An Azure Logic App is unable to "insert" rows into it (as Data Manipulation Language isn't supported in a Lakehouse I guess)

When I try a Warehouse Table:

• I cannot easily "update" a row (through "update" statements) using a Spark notebook. 
• I can only use Stored Procedures, which are very restrictive with no support for Temp tables, variables etc.

Any help would be highly appreciated.

Thanks.

[Anonymous · Accepted solution]

Hi @aputhuraya ,

Due to a lack of Data Manipulation Language (DML) support, it is not possible to insert rows into a delta table in Lakehouse.

I saw a case with a similar problem to the one you are experiencing:

Solved: store data from email attachment - Microsoft Fabric Community

His solution was to set up a query that looked at his Outlook account and copied attachments to lakehouse.

He used the “Data Destination” option in the “Query Settings” section of the data flow and selected “Append” from the update method options, which allowed him to add the new data to the lakehouse All-in-One.

You can try his method that will load the attachments in your email he table into lakehouse.

Best Regards,
Yang
Community Support Team

If there is any post helps, then please consider Accept it as the solution  to help the other members find it more quickly.
If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!