Adding --validate option __main__ and run new validation

[datapythonista]

Closes #213

python -m numpydoc keeps existing functionality, but when called with --validate runs the new validation:

(sklearn-dev) [mgarcia@xps numpydoc]$ PYTHONPATH=. python -m numpydoc --validate sklearn.linear_model.LinearRegression
sklearn.linear_model.LinearRegression:GL02:Closing quotes should be placed in the line after the last text in the docstring (do not close the quotes in the same line as the text, or leave a blank line between the last text and the quotes)
sklearn.linear_model.LinearRegression:GL03:Double line break found; please use only one blank line to separate sections or paragraphs, and do not leave blank lines at the end of docstrings
sklearn.linear_model.LinearRegression:ES01:No extended summary found
sklearn.linear_model.LinearRegression:PR06:Parameter "fit_intercept" type should use "bool" instead of "boolean"
sklearn.linear_model.LinearRegression:PR08:Parameter "fit_intercept" description should start with a capital letter
sklearn.linear_model.LinearRegression:PR06:Parameter "normalize" type should use "bool" instead of "boolean"
sklearn.linear_model.LinearRegression:PR06:Parameter "copy_X" type should use "bool" instead of "boolean"
sklearn.linear_model.LinearRegression:SA01:See Also section not found
sklearn.linear_model.LinearRegression:EX01:No examples section found

[larsoner]

Sounds good to me. This broke the test_main.py test. It would also be good to add a test for this to .travis.yml, maybe by calling numpydoc --validate numpydoc.<something> and asserting that it works for something good (zero exit code and no output) and for something bad (non-zero exit code and some output)

[datapythonista]

Tests are passing now. Not sure exactly what was the goal of mocking modules in the original tests, but I reimplemented all tests in a way that seems simpler to me, and that everything is being tested. I also added calls to python -m numpydoc to travis directly to check the status codes.

I don't think the coverage warning is relevant, I think it's being caused by the lack of tests of the argparse stuff in if __name__ == '__main__':. I'm sure we don't want to test that.

[datapythonista]

@larsoner did you have a chance to look at this? I think this should be ready, let me know if it's not.

[larsoner]

Looks good to me.

Deal with --ignore in a separate PR?

[datapythonista]

Yes, I think ignoring is complex enough to go into a separate PR. We probably want to specify it in setup.cfg too, and whether is better to whitelist or blacklist error codes may not be obvious.

[larsoner]

Okay let's get this in and keep iterating, thanks @datapythonista