Add precommit and run precommit on all files

[tomelm]

Adds pre-commit and runs it on all the current files. I skipped a few validations so I'm not making too many big changes and also set the max-line-length to 120:

(venv) flask-graphql|add-precommit ⇒ git commit -am "Add precommit and run precommit on all files"
autopep8 wrapper.........................................................Passed
Check for added large files..............................................Passed
Check python ast.........................................................Passed
Check for byte-order marker..............................................Passed
Check for case conflicts.................................................Passed
Check docstring is first.................................................Passed
Check JSON...........................................(no files to check)Skipped
Check for merge conflicts................................................Passed
Check Yaml...............................................................Passed
Debug Statements (Python)................................................Passed
Fix double quoted strings................................................Passed
Fix End of Files.........................................................Passed
Flake8...................................................................Failed
hookid: flake8

tests/test_graphqlview.py:38:1: E731 do not assign a lambda expression, use a def
tests/test_graphqlview.py:39:1: E731 do not assign a lambda expression, use a def
tests/test_graphqlview.py:452:1: F811 redefinition of unused 'test_supports_pretty_printing' from line 328

Fix python encoding pragma...............................................Passed
Tests should end in _test.py.............................................Failed
hookid: name-tests-test

tests/schema.py does not match pattern ".*_test.py"
tests/test_graphqlview.py does not match pattern ".*_test.py"
tests/test_graphiqlview.py does not match pattern ".*_test.py"
tests/app.py does not match pattern ".*_test.py"

Pretty format JSON...................................(no files to check)Skipped
Fix requirements.txt.....................................................Passed
Trim Trailing Whitespace.................................................Passed
pyupgrade................................................................Passed
Reorder python imports...................................................Passed

[coveralls]

Coverage increased (+0.1%) to 96.629% when pulling 93ca54f on tomelm:add-precommit into 724695a on graphql-python:master.

[coveralls]

Coverage increased (+0.1%) to 96.629% when pulling 93ca54f on tomelm:add-precommit into 724695a on graphql-python:master.

[coveralls]

Coverage increased (+0.1%) to 96.629% when pulling 93ca54f on tomelm:add-precommit into 724695a on graphql-python:master.

[chriskuehl]

lgtm. often we add pre-commit run --all-files to the tests somewhere which helps to make sure it doesn't regress, but it's up to you if you want to do that or not.

[chriskuehl]

I'd probably call this file requirements-dev.txt since you don't actually need pre-commit to be installed in production (and that's what requirements.txt implies, even if you're not actually doing it in this case).

[chriskuehl]

Also, we usually don't pin specific versions for our dev dependencies, but that's totally up to you / the project's stance on that.

[syrusakbary]

Maybe adding something like the following in the README will actually add more visibility into Yelp's pre-commit project and provide a better understanding to the developer, rather than using a requirements.txt file that might confuse collaborators :)

---

Collaborate

This project is using pre-commit to help the developer run pre-commit hooks easily, allowing to autoformat Python files, reorder the imports, and other various checks.

You can easily start using it via pip install pre-commit.

---

Thoughts?

[syrusakbary]

I've been using isort for checking the import ordering: https://github.com/graphql-python/flask-graphql/blob/master/tox.ini#L30.
Also being used in the automatic linters in graphql-core https://github.com/graphql-python/graphql-core/blob/master/bin/autolinter#L7, and graphene https://github.com/graphql-python/graphene/blob/master/bin/autolinter#L7.

I kind of prefer it over render-python-imports as it seems it does a smarter import grouping, not requiring a import statement per imported var.

If we could use isort instead of render-python-imports would be great! :)

[syrusakbary]

Other than the comments, everything is looking great... thanks for the work @tomelm! :)