gh-101819: Adapt _io._BufferedIOBase_Type methods to Argument Clinic #104355
Conversation
…linic Make sure the defining class is passed to all methods, so we can easily fetch module state from them in the future.
|
@AlexWaygood, would you mind taking a look at the docstrings. I exploited this possibility to tweak them a little bit (consistent mood, etc.) |
|
|
||
| cls: defining_class | ||
| / | ||
| *args: object |
There was a problem hiding this comment.
IMO the signature is wrong and should be fixed. It should have 1 optional positional argument, maybe called size. Look at other read() functions.
There was a problem hiding this comment.
Yes, I did think about that. I also think AC should be able to attach Py_UNUSED to unused arguments (I'm working on that.) I'll create a follow-up PR for fixing the clinic input. (This is not the only problematic usage of *args in _io.)
| return all data until EOF. | ||
|
|
||
| If the argument is positive, and the underlying raw stream is | ||
| not 'interactive', multiple raw reads may be issued to satisfy |
There was a problem hiding this comment.
It would be nice to clarify what "interactive" means. I suppose that it's related to "is it a TTY?" test.
The doc uses the same phrasing: https://docs.python.org/dev/library/io.html#io.BufferedIOBase.read
There was a problem hiding this comment.
Yeah. Let's improve that in a separate issue/PR. Good observation, though!
Make sure the defining class is passed to all methods,
so we can easily fetch module state from them in the future.
_ioextension module #101819