Layout-Parser/layout-parser ? reverse-engineered prompt
Reverse engineered prompt
Build me a Python library for document image analysis that helps me find and work with the layout of scanned pages and PDFs.
I want to be able to load a document image or PDF, detect things like text blocks, titles, tables, and other regions on the page, then crop those regions, run OCR on them, and draw the results back on the image so I can see what was found. It should also let me load and save layout data in simple formats like JSON and CSV, and work in a way that feels easy to use in notebooks or small scripts.
Please make it feel like a practical toolkit for real document workflows, with a clean API, a few example notebooks or scripts, and basic docs so someone can get started quickly. If you need to look up current docs online while building it, go ahead and do that.
Are you gonna build this?
make sure you review the code using coderabbit