Make sure code is PEP8 compliant

[dhermes]

Related to #178

In cases where we actively violate PEP8, we should have our own custom pylintc (e.g.).

[dhermes]

Both pep8ify and pyfmt will rewrite code to adhere.

They both seem to be unsure of themselves when it comes to maximum_line_length.

After running, it'd be smart to check with git diff --color-words locally. Luckily GitHub also now has nice word highlighting in diffs.

[dhermes]

As an experiment I tried to get datastore/entity.py to conform. Notice that #188 has come up as an issue in f7de067 since the linter is fooled about the type of self.key().

@tseaver Can you take a look and let me know what you think?

Some notes:

• Neither pep8ify nor pyfmt properly indent docstrings

• pyfmt is much slower than pep8ify (at least on the first run)

• pylint -E <file> only checks errors

• pylint --reports=n <file> only shows the offending lines

• All PyLint codes: http://pylint-messages.wikidot.com/all-codes

• To get just the errors and the global evaluation score

  pylint --disable=RP0001,RP0002,RP0003,RP0101,RP0401,RP0402,RP0701,RP0801 <file>

• Adding color pylint --output-format=colorized

[dhermes]

I spent way too much time hacking on the ast but have a decent hack which I think will do the part missing in pep8ify, which is correct indenting of docstrings:
https://gist.github.com/dhermes/85c3a3a464d2cff312ea

[dhermes]

I have two more branches:
https://github.com/dhermes/gcloud-python/tree/pep8ify-gcloud-dir
https://github.com/dhermes/gcloud-python/tree/pep8ify-setup

UPDATE:
I destroyed these branches by changing the gcloud-dir one into an upstream branch of the four indents PR (#207) and merging the setup one into the same PR.

[kleyow]

@dhermes I'm in the process of cleaning as well pyflakes/pep8 just double checking, we using 2 space instead of 4 right?

[tseaver]

We decided to move to 4 spaces (to ease community contributions).

[dhermes]

@kleyow Which files are you working on? Hoping to avoid double work.

Also, for tests without docstrings, are we OK with a global # pylint: disable=missing-docstring at the top?

[dhermes]

Sorry, just saw #200