Aggregate Ruff configuration data.

See astral-sh/ruff#3365.

Do e.g. pip install -e . to install the package in a virtualenv. Use the [histogram] extra to also get, well, histograms.

The major workflow is:

1. Find files to scan.
   • ruff-usage-aggregate scan-github-search (or other data sources, to be implemented) to find possible candidate TOML files.
     • To use this, you'll need to set the RUA_GITHUB_TOKEN environment variable to a GitHub API token. You can also place it in a file called .env in the working directory.
     • It will output a github_search_* JSONL file that can be parsed later.
   • There is an "unofficial" suite of scraper scripts for the Ruff repository's GitHub dependents page in aux/; "unofficial" because it's not using the API and may break at any time. (You can still try make scrape-dependents.)
   • There's also a data/known-github-tomls.jsonl file in the repository, which contains a list of known TOML files.
   • You can use the ruff-usage-aggregate combine command to combine github search files, CSV and JSONL files to a new known-github-tomls.jsonl file.
2. Download the files.
   • Run e.g. ruff-usage-aggregate download-tomls -o tomls/ < data/known-github-tomls.jsonl to download TOML files to the tomls/ directory.
3. Aggregate data from downloaded files.
   • ruff-usage-aggregate scan-tomls -i tomls -o json will dump aggregate data to stdout in JSON format.
   • ruff-usage-aggregate scan-tomls -i tomls -o markdown will dump aggregate data to stdout in a pre-formatted Markdown format.

ruff-usage-aggregate is distributed under the terms of the MIT license.