More Precise Exception Reports for lossless JPEG transcoding

[MemQu]

Brief Description

When it comes to transcode failures, various failures of JPEG encoder are all grouped under the same category: lossless_jpeg_5 in the list of exceptions report we get at the end.

Detailed Description

The exceptions can perhaps not share names. The encoder definitely sends back different details in case of lossless_jpeg_5, like "Can't read image data" or "EncodeImageJXL() failed"

It strikes me that these should each have their own exception ID grouped under lossless_jpeg and we certainly have the detail required to do this change

Additional Information (Optional)

Version 1.2.3

[MemQu]

Enabling normalize does separate these exceptions with different labels. It would be nice if this could happen without requiring that option.

[JacobDev1]

The error ID identifies the line of code where an error occurred. Various errors are grouped under the same ID because they occur on the same line.

The encoding process relies on a loosely coupled abstraction layer which does not share internal encoder states with the Python layer. Encoder errors are already forwarded to the user under the Exception column.

A similar system to what you are trying to achieve exists. However, it focuses on inserting notes for the user. For example:

if not success and "Some error string returned by encoding tool" in stderr:
    raise SomeException("0", f"Encoding failed because...\n\n{stderr}")

Where Encoding failed because... is the developer's note preceding stderr. This is how errors messages are made clearer to users.

Here is what I can do instead:

• Rename ID to something else.
• Insert more notes for the user based on stderr content.