W191: erroneously checks indentation within triple-quoted strings

[petere]

The following code

if True:
    print("""
    tab at start of this line
""")

produces

test.py:3:1: E101 indentation contains mixed spaces and tabs
test.py:3:1: W191 indentation contains tabs

Both of these are wrong, I believe, because the string content is not part of the program indentation.

python -t does not warn about this either.

[myint]

I think this is a duplicate of #45.

[petere]

Hmm, I suppose, but I don't understand why that issue was closed, since the problem wasn't fixed.

[sigmavirus24]

@petere if you read the issue history in #45, you'll see that you can use # noqa at the end of the string to disable pep8 checks on it. That was added by #243 and that PR can probably be extended to this purpose but would at that point probably not work so well.

[IanLee1521]

Hi @petere -- If you had a patch to change this, I'd be happy to review it.

[petere]

Apparently, #45 was closed because # noqa is considered to "fix" this issue. But that seems wrong, because it would also hide other issues in the same statement. Trivial example:

if True:
    print("""
    tab at start of this line
""");  # noqa

I think this issue (or another one like it) should be kept open until a proper fix is found.

[IanLee1521]

Sure. I can tell you why it happens (those checks are done on physical rather than logical lines) but I haven't looked into or thought about exactly how to fix it. Again, I'd be happy to review a patch if you were to put one together. 😄

[arthall]

This appears to no longer be an issue. I ran pycodestyle as of commit fd4f1c2 on the example and no style violations were identified.

[IanLee1521]

@arthall -- I created 8a93219 which as least highlights how the current behavior works (it does emit the errors as described in the original post).

[arthall]

Thank you @IanLee1521 . I see where I went wrong. I copied the example into my editor to create a test and the editor changed the tab to 4 spaces. I will keep be extra careful to check for situations like this in the future.

[IanLee1521]

That's the same thing that happened to me. No worries!

[arthall]

I believe I have a patch for this issue, but it causes 5 tests to fail. In the first test below, the W191 is no longer emitted for the string containing a line which starts with a tab. The next two tests no longer emit E101 or W191, which is my understanding of what @petere desires as the resolution of this issue.

What should I do at this point? Should I submit a pull request with the tests failing?

First failing test

#: E101 E122 W191 W191
if True:
    pass

change_2_log = \
"""Change 2 by slamb@testclient on 2006/04/13 21:46:23

    creation
"""

p4change = {
    2: change_2_log,
}


class TestP4Poller(unittest.TestCase):
    def setUp(self):
        self.setUpGetProcessOutput()
        return self.setUpChangeSource()

    def tearDown(self):
        pass

Second failing test

#: E101 W191
if True:
    print("""
    tab at start of this line
""")

Third failing test

#: E101 W191
'''multiline string
    with tabs
   and spaces
'''