08/09/26 - Retipedia v3

• Reader adjustable text layouts
• Gutenberg .zim bookshelf support and parser fixes
• Initial StackExchange, iFixIt and MedlinePlus .zim archive support
• Initial image support
• Bug fixes

You need NomadNet 1.4.0 or newer for image rendering

07/06/26 - Retipedia V2

• Formatting overhaul to add anchors, tables, inline citations and better page layouts
• Support for multiple .zim archives
• Add full Gutenberg parser
• Parser improvements
• Server-side caching
• Option to read pages in sections for low-bandwidth users
• And more

Retipedia - A searchable .zim wiki for NomadNet

This lets you host a NomadNet node that serves searchable versions of multiple sites from .zim archives provided by the Kiwix project

Installation

1. Install the python-libzim package for reading .zim archives

pip install libzim

2. Install the BeautifulSoup HTML parser package for converting the HTML from each .zim entry to Micron

pip install beautifulsoup4

3. Move the contents of the git repo to your NomadNet storage/pages/ directory

4. The page folder is detected automatically when the files live under storage/pages. You only need to adjust root_folder in settings.py for other layouts

5. Ensure all micron files for the project are executable

6. Point zims_dir in settings.py at a folder containing one or more .zim archives. A single archive_path is still supported as a fallback

7. Run python3 generate_meta.py once to scan the archives and write the per-archive metadata sidecars into the zims/ folder. Re-run it whenever you add new archives. Only archives with a sidecar are listed, so delete zims/<name>.zim.meta to hide an archive

Using a virtual environment (venv/conda)

The .mu pages start with #!/bin/python3, so NomadNet runs them with the system Python by default. If you installed libzim and beautifulsoup4 into a virtual environment instead, pages will crash with FileNotFoundError. To point every page at the environment's interpreter, run this from inside that environment

python3 generate_meta.py --fix-shebangs

To choose an interpreter explicitly, pass it after = and use quotes if it contains spaces

python3 generate_meta.py --fix-shebangs=/home/user/miniconda3/envs/nomad/bin/python3

python3 generate_meta.py "--fix-shebangs=/usr/bin/env python3"

To go back to the default just run it with --fix-shebangs=/bin/python3

Getting .zim archives

You can download .zim archives provided by the Kiwix project here https://browse.library.kiwix.org

The recommended .zim is wikipedia_en_all_nopic Wikipedia archives in every language are supported

For image support, download the _maxi Wikipedia instead of the nopic Wikipedia .zim archive

Wikipedia https://browse.library.kiwix.org/#lang=&q=&category=wikipedia

Gutenberg https://ebookfoundation.org/openzim.html

iFixIt https://browse.library.kiwix.org/#lang=&q=&category=iFixit

StackExchange (all languages) https://browse.library.kiwix.org/#lang=eng&category=stack_exchange&q=

MedlinePlus, health information from the National Institutes of Health (NIH) https://browse.library.kiwix.org/#lang=&q=medline

Retipedia currently supports Wikipedia in all languages, Gutenberg, StackExchange in all languages, iFixIt and MedlinePlus archives, and also implements a generic .zim parser for anything else. PRs are welcome

Image support

iFixIt and Wikipedia archives with pictures (maxi) can show their images in NomadNet 1.4.0 or newer on terminals with Kitty graphics support such as Konsole, iTerm2, Ghostty and Kitty.

Set images = True in settings.py to turn it on. Readers only download images when their own NomadNet image_loading setting allows it, and is safe for low-bandwidth links.

Images are extracted once into the images/ folder and served from there. You can clear that cache with

python3 clean_images.py        removes every extracted image
python3 clean_images.py 30     removes only images older than 30 days