MediaBridge is a project being developed at the Noisebridge hackerspace in San Francisco, CA, USA. See also the Noisebridge hompage and the wiki entry for this project.
MediaBridge is in a very early stage of the development. It's intended functionality is to provide recommendations that bridge media types. So for example, you might say you're interested in the film Saw and MediaBrige might recommend the video game Silent Hill or a Stephen King book. For now, we are working on simply returning recommendations for movies, based on the Netflix Prize dataset.
Currently, we are only accepting contributions from members of the project who meet in person at Noisebridge.
This code uses Python 3. It is tested on Python 3.12, but will probably work on versions back to 3.9.
To install the project dependencies, first install pipenv globally with pip install pipenv. Then create a virtual env/install dependencies with pipenv install --dev.
To run code in the project, prefix your command with pipenv run, a la pipenv run python -m mediabridge.main.
To run unit tests,
- Ensure
pipenvis installed - Run
pipenv run pytest
There is a GitHub actions "check" for passing tests, which must pass for you to be able to merge your PR.
We use ruff for code formatting, linting, and import sorting. If you've installed the project with the instructions above, you should have access to the ruff binary.
The repo comes with a .vscode directory that contains a recommended ruff extension, as well as settings to set ruff as your Python formatter and to format code and sort imports on save. If you're not using VSCode, you can run ruff format from the project root directory to format all Python code.
There is a GitHub actions "check" for code formatting, which will fail if you have unformatted code in your PR.