I have over 8 million PDFs that aren't currently full-text searchable. I was generating pages and OCRing them until I realized that was exponentially consuming more space than leaving them as PDFs. However, I still need them Full Text Search-able and my research says I can just Generate Searchable Text. However, I only have access to do that in the UI one or a few at a time. I cannot find a way to do this with Workflow. Schedule OCR only generates text if it has pages. Schedule PDF Page Generation generates pages. I do NOT want pages as they are larger to store than the PDF itself. So at this point I am just wondering...surely I am not the first one to try and solve this problem. What am I missing?
Question
Question
Replies
I think the way to get this would be to use the Distributed Computing Cluster (DCC) and you can initiate it from workflow
Sorry to say, but as far as I know you must have pages in order to have searchable text. The text is generated off the OCR'd pages, not off the image itself. You can save on space by generating pages in black and white instead of color, but you'll have to generate them to get text.
As Craig said, you can generate them in bulk by setting up DCC and running a workflow to queue all your pdf's up for generation, but with 8 million documents it's going to take a while depending on your infrastructure.
I do know that isn't entirely true. A PDF with a text layer can have text generated without generating pages. I think it technically generates pages that are from the text layer only. I can "Generate Searchable Text" in the UI on these PDFs and there are not true pages generated and they become full text searchable. They will have a page count but it isn't a true page. I just don't know how to do that in bulk.
Oh, if your PDF already has a good quality text layer then there's probably no strong need to generate pages and searchable text for them. Unless you need to perform annotations on them, stamp them, or print them nicely from Weblink/Public Portal.