Make numpydoc lint --ignore ... parsing more robust

[stefanv]

Before, the only way to specify ignores is as:

  numpydoc lint filename --ignore EX01 SA01

and multiple invocations of the ignore flag overwrote one another.

Instead, ignored checks are now specified as:

  numpydoc lint --ignore=EX01,SA01 filename

The comma can also be whitespace:

  numpydoc lint --ignore="EX01 SA01" filename

Multiple ignore flags work correctly:

  numpydoc lint --ignore EX01 --ignore SA01 filename

---

@stefmolin Would you be willing to take a look?

[rossbar]

LGTM, though @stefmolin 's opinion would be the most valuable. Thanks @stefanv !

[stefanv]

I should note that this is backward incompatible (but I don't think we've made a release with the numpydoc lint interface yet?). Specifically:

numpydoc lint foo.py --ignore EX01 SA01

needs to be rewritten as any of the following:

numpydoc lint --ignore EX01,SA01 foo.py  # --ignore can be before or after foo.py
numpydoc lint --ignore=EX01,SA01 foo.py
numpydoc lint --ignore EX01 --ignore SA01 foo.py

[stefmolin]

I'm traveling now, but I will take a look when I'm back later this week.

[stefmolin]

Functionality works as expected. I think we are missing a test case for this though (I left a suggestion).

[stefanv]

@stefmolin I addressed your feedback, thank you. This should be ready now.

[larsoner]

LGTM, @stefmolin feel free to look again if you want. If not, no sweat -- it seems like the review comment about needing a test has been addressed so I'm happy to click the green button!