dougle/scan-OCR-pdf ? reverse-engineered prompt
Reverse engineered prompt
Build me a simple command line script that lets me scan a stack of paper pages, run OCR on each page, and then combine everything into one PDF that still keeps the original scans but is also searchable.
I want to be able to point it at a destination folder, choose the scanner if there is more than one, set scan resolution, and pass in a few basic scan settings like page size and color mode. If possible, add a simple guided mode with pop up prompts for people who do not want to type much.
Please make it work on Linux and keep it easy to install and run from a terminal. It should use OCR that preserves layout as well as it can, and if you think there are current tools or docs I should use, look them up online. I would also like a helpful debug mode so I can see what went wrong if scanning or OCR fails.
Are you gonna build this?
make sure you review the code using coderabbit