[Investigate] scope PDF parsing functionality

[sidmohan0]

This is still an active topic within RAG extraction space

TODO:

• Review PDF extraction methods in well reviewed RAG tutorials (over past week)
• Break down into what's scope-able for future milestones

[bazooka720]

This will be good one to have.

[sidmohan0]

@bazooka720 we have a beta release (datafog==2.4.0b3) if you're interested in giving it a spin. Planning on releasing 2.4.0 tonight.

Main update:

• added unstructured.io support via an 'upload_file' function in the DataFog class.
• added unstructured[pdf] and unstructured[pptx] packages
• unstructured has out-of-the-box support for a variety of formats so I'm writing some tests to confirm support across those but it did pass unit tests for pdf and pptx.

sample usage:

TODO: expose OCR/Image capabilities (which using Unstructured involves passing a 'high_res' param in for setting) - will try to fit that in for tonight's release.

Let me know if you have any questions/feedback and thanks for all the input!