* add typing.ContextManager for 3.6+ only
This fixes the easier part of #655.
Would it make sense to add a generic typing.ContextManager that exists in any Python version?
* update comment
* fix argument types for ContextManager.__exit__
* add AsyncContextManager
* add @asynccontextmanager
* typing.ContextManager now always exists
* back out async-related changes
Will submit those in a separate PR later
* fix import order
* AbstractContextManager only exists in 3.6+
* AbstractContextManager -> ContextManager
Put this in stdlib/3 because it changed pretty radically in 3.2. We might be
able to move it to 2and3 later, but I want to focus on completing the Python 3
standard library first.
* Fix incorrect usage of AnyStr
- sqlite3 was using Union[bytes, AnyStr], which doesn't make sense
- The urllib functions I changed accept either bytes or str for their "safe"
argument
- Also added supports for PathLike to pstats
- Remove some unused imports of AnyStr
* pstats: python 2 accepts unicode
* improve stubs for _codecs
I noticed that many of these used AnyStr in a context where it doesn't make
sense. Found a few more things after reviewing _codecsmodule.c:
- Some Windows-only functions were omitted
- A few argument counts and names and return types were wrong
The Python 2 _codecs stub has many of the same issues, but the module is also
pretty similar between Python 2 and 3. I'm planning to send another PR after this
one to merge the two and move the stub into 2and3.
* correct platform name
Fixespython/mypy#3247
Also reviewed some related logic:
- __mod__ was added to both bytes and bytearray with PEP 461, which went into 3.5
- str.__mod__ takes only one argument, and the signature calls it "value"
- In Python 2, bytearray doesn't have __mod__ (and typeshed correctly omits it)
This is the other half of fixing #1115.
Also fixed the stub for JSONDecodeError and made it only exist in 3.5+.
Didn't otherwise review the exact attributes and types in the package.
None is the default value for these two arguments to logging.Logger._log (which
most other stuff in logging delegates to).
I was getting errors in my codebase because mypy now distinguishes between Any
and Optional[Any].
* add stubs for tracemalloc
This one has especially nice documentation, even including type annotations.
* no PEP 526 so I can run tests locally
* fixes
* another one
* move cgi to 2and3 and add types
This is pretty stable between Python 2 and 3, and not very well documented.
* fix CI failures
* adjust comment
(want to force Travis to re-run)
* give up on overriding values
* add compileall stubs to 3 and add types to the ones in 2
Didn't merge the stubs because all functions have additional parameters since 3.2,
so there would be no shared code between 2 and 3.
* add comment
- Add sys.version_info checks (mypy now supports them)
- Went over Python 3 docs and corrected a few things
- Reexport things that are imported from opcode
This works towards fixing #924. To make the changes easier to review, I'm planning to
fix the issue in two steps:
- This commit makes the 3 stubs compatible with py2
- The next PR will move the 3 stubs to 2and3 and remove the 2 stubs
I wrote this code by diffing the files in third_party/{2,3}/requests and making
appropriate changes to the Python 3 stubs (mostly str/Text stuff). I verified
that the 3 stubs now pass mypy when parsed as 2.7, but I don't have an annotated
requests-using codebase to test on.
CONTRIBUTING.md says to prefer ... Not the most impactful change but fixing
these will allow us to lint for it in the future and get a consistent style.
"> (3,)" works but looks like the code is checking for Python 4.
"<= (3, 5)" was intended to check for versions up to and including 3.5, and probably works that
way in current type checkers. However, sys.version_info is actually a 5-tuple that is greater
than (3, 5), so a hypothetical type checker that uses the full version info would interpret
this check incorrectly.
This ensures that all version_info comparisons use <, >=, ==, or !=.
Closes#953.
I reviewed the documentation at https://docs.python.org/3/library/asyncio-eventloop.html#asyncio.AbstractEventLoop and added missing methods and missing @coroutine decorators.
I ran mypy on the sample file from the issue report to confirm that mypy handles the combination of @abstractmethod and @coroutine correctly.
Also fixed a number of types in this file that are annotated as Any but could be given more precise types (e.g., sockets and protocol factories).
Both mypy and pytype only use the major and minor version in type checking. Using
checks like "sys.version_info >= (3, 4, 4)" won't actually work properly for
people type checking their code using version 3.4, because (3, 4) >= (3, 4, 4) will
always be false (at least in mypy's approach; not sure if pytype is different).
Fixes#975
And a few more things I noticed while reading the tempfile docs.
In 3.5, most functions in tempfile were changed to accept either str or bytes in their prefix and suffix arguments, and return bytes or str accordingly. This seemed like a case for AnyStr.
We were missing tempdirb and tempprefixb, added in 3.5.
TemporaryFile and others were declared as returning BinaryIO, but they actually return either binary or text IO depending on the mode passed in. I changed the return type to IO[Any], similar to builtins.open.
I've seen some stubs like
```
class X:
def __init__(self, x: str) -> None:
self.x = x
self.y = 0
```
I think this should be written instead as
```
class X:
x: str
y: int
def __init__(self, x: str) -> None: ...
```
A previous PR led mypy to refuse to instantiate ChainMap,
because it had unimplemented abstract methods.
This PR adds the abstract methods to the stub,
which is enough to persuade mypy to allow instantiating
ChainMap.
The code for this method starts as follows:
```
@coroutine
def communicate(self, input=None):
if input is not None:
stdin = self._feed_stdin(input)
```
Jelle Zijlstra