Flow is getting the wrong content-type with pdf email attachment

[RichardUchytil]

I'm using the Document Automation Kit with a trigger on a shared mailbox.  It grabs a PDF on the email, uses AI Builder to parse it and when successful, puts the parsed data in an on-prem SQL Server and uploads the PDF to SharePoint.

Today we finally got some real emails from our vendor.  The flow that grabs the PDF attachment failed because the content-type for the PDF is application/octet-stream, not application/pdf.  Yet if I create a new email, attach that same PDF, email it to myself then drop it in the folder being watched, the flow sees it as application/pdf.  I also modified the flow to say octet-stream is valid.  It processed the PDF fine, uploaded the data to SQL Server and uploaded the PDF to SharePoint.  However, in the Document Automation Kit Power App, the PDF Viewer doesn't show the PDF.  I'm guessing that's because it still has the content-type octet-stream attached to it and the PDF Viewer piece can't view it.

Any ideas how to fix this?  Can I change the content-type after I grab the PDF attachment?  Asking the vendor to change their process isn't an option.  Maybe upload to SharePoint first, then download and process it?

Thanks much!

[abm]

Hi @RichardUchytil 

Think your suggestion of upload to SharePoint is a good option here.

Have a look at this thread.

http - Do I need Content-Type: application/octet-stream for file download? - Stack Overflow

Thanks 

[RichardUchytil]

I wish there was a way to set the content-type in the flow, that would make things much easier.  I really don't want to upload to SharePoint then download again just to fix the content-type, but if that's the only way to do it, that's what I'll do.

Thanks for the help!

[RichardUchytil]

I think I found a way to do it.  In the Document Automation Kit, the first flow imports the attachment.  It checks to see if the content-type is valid (pdf, png or jpg, in our case it will always be a pdf).  I added octet-stream to that so it will continue.  Next it creates a dataverse table record and uploads the attachment.  This table has a field "File Mime Type" that stores what it is.

Adding octet-stream to the list of valid types allowed the flow to continue, the pdf was processed correctly, data was uploaded to SQL Server, and the pdf was uploaded to SharePoint where you could view it.  The only problem was in the Document Automation Kit Power App, the PDF wasn't displaying.

I added a step right after the upload setting the Mime File Type field to application/pdf.  That worked!  Now the Power App can display the PDF!  I need to do some more testing, but I think I have my solution.