Michael J. Sullivan
Michael J. Sullivan
A `mypy.ini` in the current directory will be picked up implicitly, even if no config file is specified. This has a bad interaction with how the command line arguments are...
Since #5699, mypy now rejects the following program (minimized from a few failures at dropbox): ``` from typing import TypeVar, Callable, Tuple T = TypeVar('T') def f(x: Callable[..., T]) ->...
We provide a `--strict` flag that is strict enough that we can not use it in mypy itself, which is one of the most carefully typed projects I know of....
Currently, global variables are always just stored as module attributes and accessed using the C API. It would be nice to be able to access them more directly. The snag...
When I've profiled mypyc-generated code I've used a combination of `perf`, `cProfile`, and some patches to mypyc to log stats on what getattrs we do. It would be good to...
Catching `KeyError` from a failed dict lookup in a tight loop is about twice as slow under mypyc as it is in cpython!
We should support interpreted subclasses of traits. I think that most approaches to accomplishing that should also make it possible to directly create instances of traits.
There are a bunch of async features we don't support - [x] async with/async for (https://www.python.org/dev/peps/pep-0492/) - [x] async comprehensions (https://www.python.org/dev/peps/pep-0530/) - [ ] async generators (https://www.python.org/dev/peps/pep-0525/) - [ ]...
We currently don't support code like `x = foo() + await bar()`, since the result of calling `foo()` will be stored in a temporary that is not spilled to the...
We emit code that just directly includes the names of python identifiers, which can be unicode. clang seems OK with this but gcc gets extremely upset. We should probably generate...