Automated Academic Article-to-EPUB3 Converter
Description
Budget: $30 - $250
The project is to build an automated pipeline that ingests batches of academic articles delivered as DOCX and outputs publication-ready EPUB3 files. The workflow must handle advanced formatting requirements—complex layouts, live internal navigation, clean styling, and support for equations, figures, tables, and a fully generated table of contents. Core workflow • Parse incoming DOCX normalise tags, and map semantic elements to EPUB3 structures. • Transform mathematical expressions into MathML (or high-resolution SVG fallback) so equations render crisply on all major readers. • Place figures, captions, and call-outs accurately with responsive behaviour. • Auto-build a hierarchical TOC that syncs with heading levels and in-text references. • Embed interactive footnotes and hyperlinks; keep architecture open so additional widgets (audio, quizzes, etc.) can be slotted in later without refactoring. • Package, compress, and validate every file with EPUBCheck before delivery. Implementation guidance Python, Node.js, or another scripting language is fine as long as the code is clean, documented, and callable from the command line so it can be wired into our CI environment. Leveraging tools such as Pandoc, Calibre, or custom XSLT is welcome, provided the final output meets EPUB3 spec and passes automated tests. Acceptance criteria 1. Given a sample batch of 50 articles, the tool produces 50 compliant EPUB3 files within one run. 2. All equations render via MathML (or agreed fallback) and survive EPUBCheck without warnings. 3. Figures, tables, and hyperlinks maintain correct placement and functionality on Apple Books, Kindle Previewer, and Thorium Reader. 4. Navigation (TOC, landmarks, page-list) is fully functional and reflects document hierarchy. 5. Script, config files, and brief setup guide are supplied so the pipeline can be reproduced on a fresh machine. Ship the finished scripts, a README, and the test EPUBs; once they pass the criteria above we are good to move to larger volumes.
Skills
Want AI to find more roles like this?
Upload your CV once. Get matched to relevant assignments automatically.