pdf2epub

Overview

Convert a text-layer PDF to EPUB with real structure: cover image, one chapter per section, and a working 目录 (nav + ncx). Fully deterministic — no LLM in the conversion path. Engine: scripts/pdf2epub.py (pymupdf + ebooklib + Pillow).

When to Use

• A text-layer PDF book should become an EPUB with cover + sections + TOC
• A printed 目录/Contents page needs parsing into real EPUB nav
• Footnote-heavy books (devotional/theological) need notes as linked endnotes, not body-text pollution

Don't use for: image-only scans (refused — route to document-toolkit OCR first) or PDF→Markdown extraction (that is opendataloader-pdf).

Run