The Script fails on huge pdf files

[tomgoeck]

This might be because of the token-limit of 10000 tokens per Request to the LLM

[cdeschenes]

I'm looking for a way to get the actual title and author or PDF books that do not have any metadata to pull from. Is it possible to limit how many pages to OCR? This would also fix the above issue of maxing token requests. Maybe a way to specify a page range would help.

[tomgoeck]

Thats a very good idea to limit the number of pages. But I need the entire pdf as an ocr-pdf to be able to search within the PDF files.
For your case I have created a different file: ai_scan_rename_only.sh
It only scans the first 5 pages of your book and checks for the author, title and year. The file will then be renamed and moved to the Renamed folder. I hope thats what your looking for. To test it in Terminal just navigate where the script is located and type ./ai_scan_rename_only.sh ./ to scan the current folder.

[cdeschenes]

Thanks! I will definitely give this a go this weekend and report back.