You Don't Need OCR for a PDF That Already Has a Text Layer

There's a reflex baked into a lot of document pipelines: a PDF shows up, so the first step is OCR, then extraction. Half the time that first step is unnecessary, and it's worth understanding why befor