pdfminer bug

[Coniferish]

Closes #2212.

Summary

This PR implements logic to fall back to the "inferred_layout + OCR" if pdfminer fails in the hi_res pipeline (discussed in this slack channel.

Testing

PDF: NASA-SNA-8-D-027III-Rev2-CsmLmSpacecraftOperationalDataBook-Volume3-MassProperties-pg856.pdf

elements = partition_pdf(
    filename="NASA-SNA-8-D-027III-Rev2-CsmLmSpacecraftOperationalDataBook-Volume3-MassProperties-pg856.pdf",
    strategy="hi_res",
)

[Coniferish]

@christinestraub, I'm still hitting the TypeError with these changes when partitioning via hi_res

[christinestraub]

@christinestraub, I'm still hitting the TypeError with these changes when partitioning via hi_res

@Coniferish Does it get the TypeError with the result? or only raise the TypeError?

[Coniferish]

@christinestraub Ah, it gets the TypeError and the results

[christinestraub]

@Coniferish For clarity, I reverted the changes (refactoring) in the "fast" strategy workflow. I think it would be better to do this refactoring work in any future PR (a refactor PR).