File CDK unstructured parser: Improve file type detection

[flash1293]

The current version of the unstructured parsers detects the file type based on the file name - if the file name doesn't have a proper extension (e.g. .pdf for pdf files), it won't process a file.

The logic used for this is here: https://github.com/Unstructured-IO/unstructured/blob/main/unstructured/file_utils/filetype.py#L225

This PR extends the logic the following way:

• Extend RemoteFile with mime type (as some sources like Google Drive can provide it)
• Try to detect the file type using the mime type is available
• If this fails, try to detect the file type using the file name
• If this fails, pass in the raw file to detect the file type by reading the first few bytes

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

1 Ignored Deployment

Name	Status	Preview	Comments	Updated (UTC)
airbyte-docs	⬜️ Ignored (Inspect)	Visit Preview		Oct 31, 2023 0:04am

[flash1293]

@clnoll @aaronsteers As discussed on #31458

[clnoll]

LGTM!