audio-feeder is a Flask-based web-app that hosts your audiobooks (or other audio content) as RSS feeds compatible with podcatchers.
The full documentation is at https://audio-feeder.readthedocs.io/.
audio-feeder finds the books in a directory on disk, looks up their metadata from Google Books, and lists them in a web page that also works well on mobile, from which you can subscribe to each book's feed.
- Books can be browsed by the folders they're organized in, and filtered by tag, language, author and the date they were added.
- In addition to a feed of the files as they are on disk, each book can have "rendered" feeds generated with
ffmpeg: the whole book merged into a single file, one file per chapter, or segments of about an hour each. - Each feed has a QR code, so that you can browse the list from a computer and scan the feed you want from your phone.
audio-feeder requires ffmpeg with at least the aac codec (the tests also require the libmp3lame codec). Install it with pip install audio-feeder or uv tool install audio-feeder, and create a server directory with:
$ audio-feeder install --config-dir ~/audiobook_serverThen put your books in ~/audiobook_server/static/media, and from ~/audiobook_server, load them into the database and start the server:
$ audio-feeder update
$ audio-feeder runSee the documentation for how to lay out your books, the configuration settings and how to deploy it. Probably the easiest way to deploy it is via Docker, using the docker-compose configuration in the audio_feeder_docker repository.
To try it out on some example books, run tox -e start_server from a checkout of this repository.
While the state of this is getting much closer to something production-ready, I am emphatically not a front-end developer, and I don't have extremely high confidence in the security of this project. Use at your own risk (and if you do come up with some security and/or usability improvements, please do send a PR).
This is only tested on Linux, but it may also work on other platforms.
All images and documentation contained herein are licensed under CC-0.
The code is released under the Apache 2.0 license.
Pull requests and issues are more than welcome. Please be aware that your contributions will be released under the licenses stated above. If you are not comfortable with that, please do not make a pull request.
User-facing changes should come with a changelog entry, which is managed with scriv. Running tox -e changelog creates a new fragment in changelog.d/; uncomment the section for the kind of change (Features, Bugfixes or Misc; Major for changes that everyone upgrading needs to know about; or Breaking changes for changes that can break an existing installation or customization), describe the change in a sentence or two (in reStructuredText), ending with links to the pull request and any issues it fixes, and commit it with the rest of your changes. When a release is made, the fragments are collected into CHANGELOG.rst with tox -e changelog -- collect --version <version>, under a heading with the version and the release date (e.g. Version 0.6.2 (2022-12-03)). CI checks that every pull request adds a fragment; for changes that don't need one (e.g. refactoring or CI changes), add the skip-changelog label to the pull request.