This project is for easy gathering of statistics relating to typeshed's stubs. As well as being a CLI tool and library, it also powers a website where stats about typeshed's stubs are uploaded twice a day.
This project was created by Alex Waygood. It is not necessarily endorsed by any of the other typeshed maintainers.
Some examples of things you can do from the command line:
- Create a
.csv
file with stats on all typeshed stubs:typeshed-stats --typeshed-dir <PATH_TO_TYPESHED_CLONE> --to-file stats.csv
(the.csv
file extension will be automatically detected by the script to identify the format required). - Pretty-print stats on typeshed stubs for emoji and redis to the terminal, in JSON format:
typeshed-stats --typeshed-dir <PATH_TO_TYPESHED_CLONE> --to-json emoji redis
- Generate a MarkDown file detailing stats on typeshed's stubs for protobuf and the stdlib:
typeshed-stats --typeshed-dir <PATH_TO_TYPESHED_CLONE> --to-file stats.md stdlib protobuf
Example usage of the Python-level API:
from typeshed_stats.gather import tmpdir_typeshed, gather_stats
with tmpdir_typeshed() as typeshed:
stats_on_all_packages = gather_stats_on_multiple_packages(typeshed_dir=typeshed)
- Run
pip install typeshed-stats[rich]
to install the package - Run
typeshed-stats --help
for information about various options
Are there any examples of things this script can produce, other than the website?
I'm glad you asked! They're in the examples/
folder in this repo.
(These examples are generated using the regenerate.py
script in the scripts/
directory.)
- Clone the repo and
cd
into it - Create and activate a virtual environment
- Run
pip install -e .[dev]
- Either run the linters/tests individually (see the
.github/workflows
directory for details about what's run in CI) or use thescripts/runtests.py
convenience script to run them all in succession.