- I wrote a thing: https://github.com/asottile/dead
- wanted to try it out, there's lots of false positives and I didn't look
through all the things it pointed out but here's some
When rewriting assertions, pytest makes a call to
`__name__` on each object in a comparision. If one of
the objects has reimplemented `__getattr__`, they could
fail trying to fetch `__name__` with an error other than
`AttributeError`, which is what `hasattr` catches.
In this case, the stack trace for the failed `__getattr__`
call will show up in the pytest output, even though
it isn't related to the test failing.
This change fixes that by catching exceptions
that `hasattr` throws.
Edited the changelog for extra clarity, and to fire off auto-formatting
Oddly enough, keeping `filename='{filename!r}'` caused an error while
collecting tests, but getting rid of the single ticks fixed it
Hopefully closes#3191
Unfortunately we need to get a `py.path.local` object to perform the fnmatch
operation, it is different from the standard `fnmatch` module because it
implements its own custom logic. So we need to use `py.path` to perform
the fnmatch for backward compatibility reasons.
Ideally we should be able to use a "pure path" in `pathlib` terms (a path
not bound to the file system), but we don't have those in pylib.
Fix#3973
What happens is that atomic_write on Python 2.7 on Windows will try
to convert the paths to unicode, but this triggers the import of
the encoding module for the file system codec, which in turn triggers
the rewrite, which in turn again tries to import the module, and so on.
This short-circuits the cases where we try to import another file when
writing a pyc file; I don't expect this to affect anything because
the only modules that could be affected are those imported by
atomic_writes.
Fix#3506
Daniel Hahler
Anthony Sottile
Anton Lodder
Bruno Oliveira
Tomer Keren
Ronny Pfannschmidt
iwanb
Fabio Zadrozny