SpeedPatrolling

This tool helps Wikidata editors to patrol recent changes.

For more information, please see the tool’s on-wiki documentation page.

Toolforge setup

On Wikimedia Toolforge, this tool runs under the speedpatrolling tool name, using the Toolforge Components Service to coordinate building a container with the Toolforge Build Service and then deploying that for the webservice. The components configuration is in the toolforge.yaml file.

To start a new deployment, run the following command on Toolforge after becoming the tool account:

toolforge components deployment create

This should automatically kick off an image build and restart the webservice at the end.

Details and troubleshooting

To inspect the overall deployment status, run:

toolforge components deployment show

To debug the image build step, it may be useful to trigger an image build explicitly – you can add --ref=foobar to build from the foobar branch instead of the main branch:

toolforge build start https://gitlab.wikimedia.org/toolforge-repos/speedpatrolling

The web frontent is a Flask WSGI app using gunicorn, and runs as the speedpatrolling job, which you may inspect with commands like these:

toolforge jobs show speedpatrolling
toolforge jobs logs speedpatrolling
kubectl get deployment speedpatrolling
kubectl exec -it deployment/speedpatrolling -- bash

Configuration

The tool reads configuration from both the config.yaml file (if it exists) and from any environment variables starting with TOOL_*. The config file is more convenient for local development; the environment variables are used on Toolforge: list them with toolforge envvars list. Nested dicts are specified with envvar names where __ separates the key components, so that e.g. the following are equivalent:

toolforge envvars create TOOL_OAUTH__CONSUMER_KEY 271b735e0cf895694f2ee7a3ae7a2dbc

yaml
OAUTH:
    CONSUMER_KEY: 271b735e0cf895694f2ee7a3ae7a2dbc

For the available configuration variables, see the config.yaml.example file.

Update

The tool should automatically be updated on every push to the main branch. To trigger a manual update, run toolforge components deployment create as described above.

Local development setup

You can also run the tool locally, which is much more convenient for development (for example, Flask will automatically reload the application any time you save a file).

git clone https://gitlab.wikimedia.org/toolforge-repos/speedpatrolling.git
cd speedpatrolling
pip3 install -r requirements.txt -r dev-requirements.txt
flask --debug run

If you want, you can do this inside some virtualenv too.

Note that your possibilities to work on this tool are rather limited unless you request your own OAuth consumer and configure it in a config.yaml file – without OAuth credentials, the tool cannot even load a list of unpatrolled changes.

Contributing

To send a patch, you can submit a merge request on GitLab or a pull request on GitHub. (E-mail / patch-based workflows are also acceptable.)

License

The code in this repository is released under the AGPL v3, as provided in the LICENSE file.