Collects test coverage data from your Python test suite and sends it to Code Climate's hosted, automated code review service.
Code Climate - https://codeclimate.com
The codeclimate-test-reporter
is compatible with coverage.py coverage
reports. By default, coverage.py will generate a .coverage
file in the current
directory. codeclimate-test-reporter
, run without arguments, will
look for a coverage report at this default location.
Note: The codeclimate-test-reporter
requires a repo token from
codeclimate.com, so if you don't have one, the first step is to signup
and configure your repo. Then:
You can place the repo token in the environment under the key
CODECLIMATE_REPO_TOKEN
or pass the token as a CLI argument:
$ CODECLIMATE_REPO_TOKEN=[token] codeclimate-test-reporter
Submitting payload to https://codeclimate.com... done!
$ codeclimate-test-reporter --token [token]
Submitting payload to https://codeclimate.com... done!
We recommend configuring the repo token in the environment through your CI settings which will hide the value during runs. The token should be considered a scoped password. Anyone with the token can submit test coverage data to your Code Climate repo.
# CODECLIMATE_REPO_TOKEN already set in env
$ codeclimate-test-reporter
Submitting payload to https://codeclimate.com... done!
To generate a coverage report with pytest, you can use the pytest-cov plugin:
$ py.test --cov=your_package tests/
TOTAL 284 27 90%
======================== 14 passed in 0.75 seconds ========================
To generate a coverage report with nose, you can use the nose cover plugin:
$ nosetests --with-coverage --cover-erase --cover-package=your_package
TOTAL 284 27 90%
----------------------------------------------------------------------
Ran 14 tests in 0.743s
OK
By default, coverage.py will create the test coverage report at ./.coverage
.
If you configure coverage.py to generate a coverage report at an alternate
location, pass that to the codeclimate-test-reporter
:
$ codeclimate-test-reporter --file ./alternate/location/.coverage
Submitting payload to https://codeclimate.com... done!
You can install the codeclimate-test-reporter using pip:
$ pip install codeclimate-test-reporter
Successfully installed codeclimate-test-reporter-[version]
Or you can add the reporter to your requirements.txt
file and install it along
with other project dependencies:
# requirements.txt
codeclimate-test-reporter
$ pip install -r requirements.txt
Successfully installed codeclimate-test-reporter-[version]
Across the many different testing frameworks, setups, and environments, there are lots of variables at play. Before setting up test coverage, it's important to understand what we do and do not currently support:
-
Single payload: We currently only support a single test coverage payload per commit. If you run your tests in multiple steps, or via parallel tests, Code Climate will only process the first payload that we receive. If you are using a CI, be sure to check if you are running your tests in a parallel mode.
Note: If you've configured Code Climate to analyze multiple languages in the same repository (e.g., Python and JavaScript), we can nonetheless only process test coverage information for one of these languages. We'll process the first payload that we receive.
-
Invalid File Paths: By default, our test reporters expect your application to exist at the root of your repository. If this is not the case, the file paths in your test coverage payload will not match the file paths that Code Climate expects.
If you're having trouble setting up or working with our test coverage feature, see our detailed help doc, which covers the most common issues encountered.
If the issue persists, feel free to open up an issue here or contact help with debug information from the reporter:
$ codeclimate-test-reporter --debug
codeclimate-test-reporter [version]
Requests 2.9.1
Python 3.5.1 (default, Jan 22 2016, 08:54:32)
[GCC 4.2.1 Compatible Apple LLVM 7.0.2 (clang-700.1.81)]
/usr/local/opt/python3/bin/python3.5
Darwin 15.4.0
Patches, bug fixes, feature requests, and pull requests are welcome.
A Dockerfile
is included for developer convenience. Simply run make test
to
run the test suite and coverage report.
To release a new version, first run ./bin/prep-release [version]
to bump the
version and open a PR on GitHub.
Once the PR is merged, run make release
to publish the new version to pypy and
create a tag on GitHub. This step requires pypy credentials at ~/.pypirc
. You
can copy this repo's .pypirc.sample
file as a starting point.
See LICENSE.txt