We welcome contributions of all kinds including improvements to the core code, addition of new dataset scripts to add new datasets to the Retriever, improvements to the documentation, bug reports, or anything else you can think of. We strive to be supportive of anyone who wants to contribute, so don't be shy, give it a go, and we'll do our best to help. One way to ease into contributing is to add datasets to the Retriever.
We use a standard GitHub flow for development and reviewing contributions. Fork the repository. Make changes to a branch of your fork and then submit a pull request.
We use pytest for testing. To run the tests first install nose using pip:
pip install pytest
Then from the root of the repository install the Retriever:
python setup.py install
and run the tests:
pytest
You should see a bunch of output from the Retriever showing the test results.
Tests for MySQL and PostgreSQL require properly configured database management systems for testing.
Requires that the postgres user has permissions on a database named testdb_retriever
from localhost. This login information should be stored in the postgreSQL
password file.
Requires that the travis user has permissions on a database named testdb_retriever
from localhost.
We use Travis CI for continuous integration testing. All pull requests will automatically report whether the tests are passing.
Our postgreSQL tests fail idiosyncratically on Travis, so if only the postgres tests are failing it on Travis, and they are passing locally, it is likely not a problem.