xonsh/.travis.yml

language: python
python:
  - 3.4
  - 3.5
  - "nightly"
install:
  - pip install -r requirements-tests.txt
script:
  - coverage run --source xonsh -m py.test
after_success:
  - codecov
a first approach to a travis file * to run tests with travis-ci.org * prerun test, pip install ply and nosetests * discover and run test with nosetests 2015-03-15 23:35:10 -03:00			`language: python`
			`python:`
			`- 3.4`
try travis 3.5 2015-09-26 22:31:48 -04:00			`- 3.5`
Travis test on nightly 2016-06-03 10:24:33 -07:00			`- "nightly"`
a first approach to a travis file * to run tests with travis-ci.org * prerun test, pip install ply and nosetests * discover and run test with nosetests 2015-03-15 23:35:10 -03:00			`install:`
fix travis call to pip 2016-04-01 00:16:37 -04:00			`- pip install -r requirements-tests.txt`
a first approach to a travis file * to run tests with travis-ci.org * prerun test, pip install ply and nosetests * discover and run test with nosetests 2015-03-15 23:35:10 -03:00			`script:`
update travis and appveyor to run pytest (hopefully) 2016-06-22 16:17:04 -04:00			`- coverage run --source xonsh -m py.test`
Enable (some) coverage. Not sure exactly how xonsh is tested completely, so I expect most of the coverage process to be wrong (for now) Codecov have [browser extensions](https://github.com/codecov/browser-extension) that allow you to see the coverage directly in the pull requests in GitHub. We might have to give codecov some limited access to the repo (webhook), and tell the codecov bot to shut its mouth once it's enabled. But it's super non intrusive, and kind-of useful when you refactor to see that your code is tested. Also set some sensible where coverage would fail only if it drops by 10% in 1 PR. 2016-06-06 19:43:21 -07:00			`after_success:`
			`- codecov`