MIME type of OOXML file formats (.docx, .xlsx,.pptx)

Hello everyone,

I have a concern regarding the MIME type of uploaded attachments. Our case only supports DOC or DOCX files, and we’ve implemented a validator to ensure that users can only upload these file types based on the MIME type that DS returns. However, in A12 2023.06, DS now uses the Tika library to detect the MIME type, treating all .docx, .xlsx, and .pptx file types as ‘application/x-tika-ooxml.’

I’m aware that Tika can detect the exact MIME type for Microsoft Office file types based on file content. Still, DS doesn’t allow us to access the file content if it hasn’t been linked to an A12 document. Therefore, validating the file type before saving the A12 document is not possible.

Additionally, there’s another issue. Previous DS versions returned an exact MIME type, not the Tika one. As a result, our A12 documents have been stored with that exact MIME type, requiring us to write our own data migration.
Does anyone have any thoughts or suggestions?
Thank you.

For more information regarding the data migration:
This is our old data in the attachment_header table


And we also store the MINE type in our A12 document

The data after upgrading to A12 2023.06


Hi @quynh-still-birch,

if there is a bug in Data Services regarding mime-type detection, please create a Jira ticket, so we can fix it. If you have direct suggestion how to setup Tika to differentiate ooxml files, please attach it to the ticket.

Thank you for your reply, @petr-high-peak
I have created the ticket A12-15170

Update: The issue is fixed in 2023.06-ext3