[Reverted] Make os.stat_result and friends NamedTuples #1103
Conversation
stdlib/2/os/__init__.pyi
Outdated
| ('f_namemax', int)]) | ||
|
|
||
| class stat_result(NamedTuple('stat_result', [ | ||
| ('st_mode', int), # protection bits, |
There was a problem hiding this comment.
I don't think we need all of those comments. Typeshed is for types, not documentation.
stdlib/2/os/__init__.pyi
Outdated
| # st_ctime. More items may be added at the end by some implementations. | ||
|
|
||
| if sys.version_info >= (3, 3): | ||
| st_atime_ns = 0 # time of most recent access, in nanoseconds |
There was a problem hiding this comment.
Better to write st_atime_ns: int (or st_atime_ns = ... # type: int, but I think we've established that all users of typeshed obey PEP 526 now).
stdlib/3/os/__init__.pyi
Outdated
| @@ -325,8 +321,11 @@ if sys.version_info >= (3, 5): | |||
| def scandir(path: str = ...) -> Iterator[DirEntry[str]]: ... | |||
| @overload | |||
| def scandir(path: bytes) -> Iterator[DirEntry[bytes]]: ... | |||
| @overload | |||
| def stat_float_times(newvalue: bool = ...) -> None: ... | |||
There was a problem hiding this comment.
I don't think this should have a default value; that will make it overlap with the other overload.
|
Thanks! Merged. |
|
Hmm... This caused two types of breakage with our internal code. I think it's better to roll back. @JelleZijlstra, @euresti What do you guys think? Here are the errors (same as I reported for #1111): This one's caused by the (newly introduced) difference between os.stat_result and posix.stat_result: This one's triggered by Who agrees with a rollback? |
|
We can fix the first one pretty easily by making stdlib/2/posix.pyi re-export os.stat_result (the one in stdlib/3 already does that). I don't think the second one can be fixed in the stubs, but maybe mypy is being too strict there (it could just allow that code and make it return Any). I'm OK with reverting for now, but it would be good to get some form of this PR back in eventually so |
|
I filed python/mypy#3078 for the latter. I'll revert if you don't beat me to it. |
This reverts commit 6c3e175.
gvanrossum
changed the title
|
Ack! Sorry about the mess. Please revert and I'll take a look later.
…On Mar 29, 2017 2:25 PM, "Guido van Rossum" ***@***.***> wrote:
I filed python/mypy#3078 <python/mypy#3078> for
the latter. I'll revert if you don't beat me to it.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1103 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/ABCCArqkmnpXuGWPU2HwHDJRGq8GTrETks5rqqIlgaJpZM4MppiS>
.
|
The elements can be accessed as indexes or through attributes, so I chose to make them NamedTuples.