microsoft / microsoft/markitdown
[EPUB] Noisy Markdown output (XML/CSS) and missing chapters due to path resolution issues
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 186k
- Forks
- 13.7k
- Avg merge
- 1d 4h
- Merged PRs (30d)
- 49
Description
Description
Currently, the _epub_converter.py in MarkItDown has two significant limitations when handling EPUB files:
- Noise in output: XHTML files often include XML declarations and
<style>blocks. If theBeautifulSoupsearch for<body>fails (common with namespaces), the entire raw file content is included in the Markdown output. - Missing Content: The simplistic path joining logic
f"{base_path}/{manifest[item_id]}"fails to correctly resolve relative paths (e.g.,../Text/...) used in many commercial EPUB manifests.
Potential Solution
A more robust manifest parser that handles relative paths and a more aggressive XHTML cleaner would resolve these issues. I have implemented a workaround for a private project and would be happy to contribute a PR if interested.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading _epub_converter.py, especially the BeautifulSoup body search and manifest path construction. Reproduce the issue with EPUB files containing namespaced XHTML, XML or style noise, and relative paths such as ../Text/.... Done means Markdown excludes that noise and includes chapters whose manifest paths require relative resolution.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- backend
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100