Demo with a generated sample PDF (a fictional "Sample Co 001" report, no real data).
Turns a text-based PDF into Markdown or plain text and keeps the layout: headings, bold and italic, paragraph breaks, indentation, simple tables, bullets and page markers. Scanned pages are detected and flagged for OCR instead of guessed, so the tool never invents text it can't see.
Python with PyMuPDF; 16 pytest tests passing. The screenshots compare the original page with the output.
https://github.com/SheldonZhuang/sheldontek-pdf-layout-extract-demo
0 likes · 0 views · Oct 26, 2026
Demo: text PDFs to Markdown with bold, italic, tables, lists and page markers kept; scanned pages flagged for OCR. 16 tests pass.
