A Dockerized Python solution that extracts structured outlines (titles and headings) from textual PDF files using rule-based heuristics.
-
Updated
Aug 6, 2025 - Python
A Dockerized Python solution that extracts structured outlines (titles and headings) from textual PDF files using rule-based heuristics.
Find the real sections of a PDF (headings, levels, page spans, text) for RAG: PDF outline + rules, optional LLM, hard cost caps. Benchmarked on READoc.
To associate your repository with the heading-detection topic, visit your repo's landing page and select "manage topics."