It's a widely used (but perhaps not official?) convention to consider a metaclass's methods to be class methods and to use cls as the first argument (rather than self).

Use the simple heuristic of "class inherits from type" to identify metaclasses and mark all of their methods as CLASSMETHOD.

The pep8 entry was quite confusing to me and I don't think it will ever show up with a newish flake8 version. Also update naming to the latest release.

(Unfortunately) some libraries in the python standard library do not follow pep8. This means that people using them are forced to write non pep8-compliant code. This code should be accepted by the pep8-naming plugin. For example unittest.TestCase is meant to be overridden so setUp and tearDown can be implemented. The same goes for logging.Logger.

When I run flake8 with the new pep8-naming version 0.11.0, I get an AttributeError from pep8ext_naming.py

$ .venv/bin/flake8
multiprocessing.pool.RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/path/python/envs/3.7.6/lib/python3.7/multiprocessing/pool.py", line 121, in worker
    result = (True, func(*args, **kwds))
  File "/path/python/envs/3.7.6/lib/python3.7/multiprocessing/pool.py", line 44, in mapstar
    return list(map(*args))
  File "/path/.venv/lib/python3.7/site-packages/flake8/checker.py", line 655, in _run_checks
    return checker.run_checks()
  File "/path/.venv/lib/python3.7/site-packages/flake8/checker.py", line 589, in run_checks
    self.run_ast_checks()
  File "/path/.venv/lib/python3.7/site-packages/flake8/checker.py", line 496, in run_ast_checks
    for (line_number, offset, text, _) in runner:
  File "/path/.venv/lib/python3.7/site-packages/pep8ext_naming.py", line 195, in visit_tree
    for error in self.visit_tree(child):
  File "/path/.venv/lib/python3.7/site-packages/pep8ext_naming.py", line 191, in visit_tree
    for error in self.visit_node(node):
  File "/path/.venv/lib/python3.7/site-packages/pep8ext_naming.py", line 201, in visit_node
    self.tag_class_functions(node)
  File "/path/.venv/lib/python3.7/site-packages/pep8ext_naming.py", line 242, in tag_class_functions
    iter_child_nodes(cls_node), ismetaclass, late_decoration)
  File "/path/.venv/lib/python3.7/site-packages/pep8ext_naming.py", line 259, in set_function_nodes_types
    name = d.func.id if isinstance(d, ast.Call) else d.id
AttributeError: 'Attribute' object has no attribute 'id'
"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File ".venv/bin/flake8", line 8, in <module>
    sys.exit(main())
  File "path/.venv/lib/python3.7/site-packages/flake8/main/cli.py", line 22, in main
    app.run(argv)
  File "/path/.venv/lib/python3.7/site-packages/flake8/main/application.py", line 360, in run
    self._run(argv)
  File "/path/.venv/lib/python3.7/site-packages/flake8/main/application.py", line 348, in _run
    self.run_checks()
  File "/path/.venv/lib/python3.7/site-packages/flake8/main/application.py", line 262, in run_checks
    self.file_checker_manager.run()
  File "/path/.venv/lib/python3.7/site-packages/flake8/checker.py", line 323, in run
    self.run_parallel()
  File "/path/.venv/lib/python3.7/site-packages/flake8/checker.py", line 289, in run_parallel
    for ret in pool_map:
  File "/path/python/envs/3.7.6/lib/python3.7/multiprocessing/pool.py", line 354, in <genexpr>
    return (item for chunk in result for item in chunk)
  File "/path/python/envs/3.7.6/lib/python3.7/multiprocessing/pool.py", line 748, in next
    raise value
AttributeError: 'Attribute' object has no attribute 'id'

My guess is that it is not only Python 3.7, but that is all I've tried.

I believe it's common practice to shorten long imports from selenium like it is done in Selenium docs: http://selenium-python.readthedocs.io/waits.html

But pep8-naming detects error:

from selenium.webdriver.support import expected_conditions as EC
^
1     N812 lowercase 'EC' imported as non lowercase
1

I think the intention of N812 was to avoid importing lowercase as CamelCase, but this one is actually UPPERCASE, so I think it's false positive.

I have occasionally found out that it is possible to use any case for naming class attributes:

class Test(object):
    camelCase = 1
    snake_case = 2
    UPPER_CASE = 3

flake8 and pep8-naming won't complain about it.

Since https://www.python.org/dev/peps/pep-0008 does not say a word about class attributes naming style, I believe that it might be a hot discussion.

In my opinion class attributes should be using snake_case by default. What do you think? Thanks!

pep8-naming tests fail for flake8 v3/v4; avoid suggesting that it works

Additionally, v6 added one more required kwarg. Let's add that one too.

Signed-off-by: Stavros Ntentos [email protected]

After upgrading to flake>=3.9.1 the following test failure came up.

Traceback (most recent call last):
  File "run_tests.py", line 117, in <module>
    main()
  File "run_tests.py", line 34, in main
    errors += test_file(filename, testcase, code, options)
  File "run_tests.py", line 99, in test_file
    parse_options(checker, options)
  File "run_tests.py", line 90, in parse_options
    checker.parse_options(processed_options)
  File "/nix/store/9fggrr07qvlrqqqf2lpy9ynfn30v1ans-python3.8-pep8-naming-0.11.1/lib/python3.8/site-packages/pep8ext_naming.py", line 181, in parse_options
    engine = style_guide.DecisionEngine(options)
  File "/nix/store/mr2dci4kaf4illh63hb264lhn4vp6s1w-python3.8-flake8-3.9.1/lib/python3.8/site-packages/flake8/style_guide.py", line 171, in __init__
    ).union(options.extended_default_ignore)
AttributeError: 'Values' object has no attribute 'extended_default_ignore'

Seems related to this change: https://github.com/PyCQA/flake8/pull/1317

Here's an interesting one (see last line of traceback):

Traceback (most recent call last):
  File "/usr/local/Cellar/python/3.6.5/Frameworks/Python.framework/Versions/3.6/lib/python3.6/multiprocessing/pool.py", line 119, in worker
    result = (True, func(*args, **kwds))
  File "/usr/local/lib/python3.6/site-packages/flake8/checker.py", line 648, in _run_checks
    return checker.run_checks()
  File "/usr/local/lib/python3.6/site-packages/flake8/checker.py", line 579, in run_checks
    self.run_ast_checks()
  File "/usr/local/lib/python3.6/site-packages/flake8/checker.py", line 493, in run_ast_checks
    for (line_number, offset, text, check) in runner:
  File "/usr/local/lib/python3.6/site-packages/pep8ext_naming.py", line 146, in visit_tree
    for error in self.visit_tree(child):
  File "/usr/local/lib/python3.6/site-packages/pep8ext_naming.py", line 142, in visit_tree
    for error in self.visit_node(node):
  File "/usr/local/lib/python3.6/site-packages/pep8ext_naming.py", line 152, in visit_node
    self.tag_class_functions(node)
  File "/usr/local/lib/python3.6/site-packages/pep8ext_naming.py", line 185, in tag_class_functions
    ismetaclass = any(name for name in cls_node.bases if name.id == 'type')
  File "/usr/local/lib/python3.6/site-packages/pep8ext_naming.py", line 185, in <genexpr>
    ismetaclass = any(name for name in cls_node.bases if name.id == 'type')
AttributeError: 'Attribute' object has no attribute 'id'

It is interesting because before I started whittling down the flake8 errors (all I was doing was adding documentation really), pep8-naming was doing OK. Then out of the blue this crash!

I found a similar issue which was resolved by checking if id was available.

Before I start working on a PR that changes any xxx.id to be getattr(xxx, "id", None), since there are a lot more uses of .id in pep8-naming, anything special to look out for?

The file being checked here is actually deliberately breaking pep8-naming conventions, but I was working on the other flake8 errors before I #noqa the classes that intentionally break things. I wonder if that is somehow related (the fact that there are actual errors causing this issue)?

This may be related to #45:

def foo_():
    pass


def foo__():
    pass


def test___main__():
    """Test python -m functionality."""
    with pytest.raises(SystemExit) as excinfo:
        with unittest.mock.patch('sys.argv', []):
            # pylint: disable=unused-variable, redefined-outer-name
            import backlog.__main__  # noqa: F401
    assert excinfo.value.code == 0

/home/david/Projects/backlog/tests/test_cli.py:14:5: N802 function name 'foo__' should be lowercase xxx
/home/david/Projects/backlog/tests/test_cli.py:18:5: N802 function name 'test___main__' should be lowercase xxx

• Add classmethod-decorator and staticmethod-decorator config parameters
• Add support for parsing CLI-style flake8-config parameters in testing framework.
• Remove support for multiple codes in a testcase, since it was never used.
• Add tests for late-decorated methods

Fixes #38

These are differences in behavior based on existing and new test cases:

statement | before | now -- | -- | -- import os as OS | N812 lowercase 'os' imported as non lowercase 'OS' | N810 package or module 'os' imported as non lowercase 'OS' import GOOD as good | N811 constant 'GOOD' imported as non constant 'good' | Okay (because GOOD is a package or module) import good as BAD | N812 lowercase 'good' imported as non lowercase 'BAD' | N810 package or module 'good' imported as non lowercase 'BAD' import good as Bad | N812 lowercase 'good' imported as non lowercase 'Bad' | N810 package or module 'good' imported as non lowercase 'Bad' import GOOD as Γ | Okay | N810 package or module 'GOOD' imported as non lowercase 'Γ'

Hope it makes sense based on the discussion in #201.

For example, this code:

import SalesforcePy.commons as sfc

fails with:

./test.py:1:2: N813 camelcase 'SalesforcePy.commons' imported as lowercase 'sfc'

Shouldn't pep-naming just look at the package name being imported (commons) instead of looking at the parent package names too, and then not fail in this case?

I'm assuming this rule exists for when a class name is being imported, as discussed in https://stackoverflow.com/a/37590149/2654518. In this case:

from SalesforcePy.commons import BaseRequest as baserequest

sure should fail with:

./test.py:1:2: N813 camelcase 'BaseRequest' imported as lowercase 'baserequest'

The built-in base HTTP request handler class in http.server asks you to create methods named (for example) do_GET, do_POST, etc. Rule N802 warns about these methods.

$ cat -n a.py
     1  # pylint: disable=missing-class-docstring, missing-module-docstring
     2  
     3  class Foo(type):
     4      pass
     5  
     6  
     7  class Bar(Foo):
     8      def __init__(cls):
     9          super().__init__()
$ flake8 --ignore=D100,D101,D107 a.py 
a.py:8:19: N805 first argument of a method should be named 'self'

PEP8: Always use cls for the first argument to class methods.

Note that pylint is correct here:

$ pylint a.py

--------------------------------------------------------------------
Your code has been rated at 10.00/10 (previous run: 10.00/10, +0.00)

Prosed fix for #178

Add specific exemption from N815 for all subclasses of TypedDict because class variable naming conventions should not apply to dictionary keys.

I've tried to reuse code from the rule for Exceptions (N818) as per the suggestion on #178.

This is my first time working with the ast module or with any flake8 plugin so please bear that in mind when reviewing this PR. Any advice on whether it's reasonable to tag every ClassDef with the names of its superclasses, and potential performance concerns around that, would be very welcome. On balance, I thought it may be preferable to do this rather than determining the superclasses again for every variable in a class.

Thanks :)