Alex Waygood
cb7742e12c
tempfile: Fix TypeVar usage (#7939 )
...
Co-authored-by: Jelle Zijlstra <jelle.zijlstra@gmail.com >
2022-05-25 20:06:46 -07:00
Shantanu
c8f9abf41d
threading.Semaphore: add _value ( #7952 )
...
I've had occasion to use this multiple times. We already have _value for
asyncio.locks.Semaphore See also
https://github.com/python/cpython/issues/93213
Co-authored-by: hauntsaninja <>
2022-05-25 14:24:09 -07:00
Henry Schreiner
afb512f38b
sre_parse.SubPattern.getwidth: return a tuple of ints (#7951 )
2022-05-25 17:34:51 +01:00
Graham Bleaney
cb5b31cf15
Add LiteralString overloads to path module ( #7727 )
...
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com >
2022-05-25 08:52:52 -07:00
Alex Waygood
ac9efd8573
stubtest_stdlib: run on PRs updating requirements-tests.txt (#7949 )
2022-05-25 15:29:05 +02:00
Sebastian Rittau
301caad709
Update mypy to 0.960 ( #7948 )
2022-05-25 14:44:22 +02:00
Jelle Zijlstra
d5bc48d29b
builtins, collections: Fix unconstrained overloads for container constructors ( #7944 )
...
See https://github.com/microsoft/pyright/issues/3501#issuecomment-1135979479
Related to #7928
2022-05-25 10:07:23 +01:00
Jaakko Lappalainen
773ddb15bb
Pillow: use union type for Image.paste (#7893 )
2022-05-25 10:04:59 +01:00
Jelle Zijlstra
796a689fc1
mock: Fix unconstrained TypeVar ( #7945 )
...
The first overload covers the case where `new` is not given.
Part of #7928
2022-05-25 06:52:40 +01:00
Jelle Zijlstra
070e6c2a2d
invoke: Fix unconstrained TypeVar ( #7943 )
...
Part of #7928
f34c6c9413/invoke/tasks.py (L328)
2022-05-25 06:48:49 +01:00
Jelle Zijlstra
bb119544d8
pkg_resources: Fix unconstrained TypeVars ( #7941 )
...
499c468a57/pkg_resources/__init__.py (L1049)
Part of #7928
2022-05-24 19:49:22 -07:00
Alex Waygood
54e11a1ec4
mypy_extensions: fix TypeVar usage (#7937 )
...
#7928
2022-05-24 11:26:35 -07:00
L
1e097a75cb
markdown: Annotate methods of Registry ( #7926 )
...
Co-authored-by: Shantanu <12621235+hauntsaninja@users.noreply.github.com >
2022-05-24 10:23:48 -07:00
Jelle Zijlstra
d950ec37ca
fileinput: Fix TypeVar usage (#7934 )
...
* fileinput: Fix TypeVar usage
Fixes #7922 , part of #7928 .
2022-05-24 17:55:35 +01:00
Petter Friberg
c1aba8059f
Extend digest argument type for hmac.digest ( #7931 )
...
Now allows a `Callable` and `ModuleType` in addition to `str`
2022-05-24 11:38:00 +01:00
Jelle Zijlstra
158293257e
optparse: fix unconstrained TypeVar ( #7936 )
...
Part of #7928
2022-05-23 19:39:31 -07:00
Amy Pircher
8a07787f48
Add precise overloads for SSLContext.get_ca_certs ( #7933 )
2022-05-23 17:06:09 -07:00
Jelle Zijlstra
b4518bfb2b
argparse: remove incorrect default in overload ( #7929 )
...
Part of #7928
2022-05-23 18:12:19 +01:00
Shantanu
8968eb4fc9
bdb: update for py311 ( #7927 )
2022-05-23 09:20:48 +02:00
Jelle Zijlstra
9660ee97ee
codecs: add missing function, reflect runtime import * ( #7918 )
2022-05-22 16:06:21 -07:00
Jelle Zijlstra
18433202ba
concurrent.futures: new params in 3.11 ( #7919 )
...
fdc0e09c33
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com >
2022-05-22 15:51:09 -07:00
Shantanu
ceda368abc
shutil: update rmtree for py311 ( #7925 )
2022-05-22 15:40:06 -07:00
Shantanu
c0aafb9a34
termios: update for py311 ( #7924 )
...
Co-authored-by: hauntsaninja <>
2022-05-22 15:36:50 -07:00
Shantanu
5b1ef8b119
mmap: add MADV_FREE on darwin ( #7923 )
...
Co-authored-by: hauntsaninja <>
2022-05-22 15:30:09 -07:00
Rens Oliemans
6e18441a2c
Update emoji stubs for version 1.7 ( #7884 )
2022-05-22 11:44:01 +01:00
Jelle Zijlstra
54c2e23a55
3.11 allowlist: move inspect._ParameterKind.description to wontfix section ( #7920 )
2022-05-22 12:04:11 +03:00
Peter Pentchev
ef887b7beb
More setuptools.command.easy_install definitions. ( #7145 )
...
Co-authored-by: Sebastian Rittau <srittau@rittau.biz >
Co-authored-by: Akuli <akuviljanen17@gmail.com >
2022-05-22 00:31:31 +03:00
Jelle Zijlstra
3dcfbf3745
urllib.request: add code 308 handler for 3.11 ( #7917 )
...
c379bc5ec9
2022-05-21 14:13:17 -07:00
Shantanu
f2ea124e82
random: explicitly mention int for seed ( #7916 )
...
https://github.com/python/typeshed/pull/7906#discussion_r878750404
I felt this better documents how seed is used in practice
Co-authored-by: hauntsaninja <>
2022-05-21 13:00:05 -07:00
Jelle Zijlstra
1220a99399
unicodedata: add UCD.is_normalized ( #7915 )
2022-05-21 12:59:44 -07:00
Sebastian Rittau
a412ddb437
Update pyright ( #7914 )
...
Required for #7865
2022-05-21 19:15:06 +01:00
eggplants
6d372ffc09
Add docopt.printable_usage ( #7913 )
2022-05-21 18:33:04 +02:00
Jelle Zijlstra
ddc6eda10a
symtable: has_exec was removed in 3.9 ( #7911 )
2022-05-21 17:50:27 +02:00
Jelle Zijlstra
39c777f70c
pydoc: 3.11 changes ( #7910 )
2022-05-21 17:49:39 +02:00
Jelle Zijlstra
237b1bb189
binascii.a2b_base64: new param in 3.11 ( #7909 )
...
35b98e38b6
2022-05-21 08:36:03 -07:00
Jelle Zijlstra
9c0c631e46
gettext.install: names is kw-only in 3.11 ( #7908 )
2022-05-21 08:22:34 -07:00
Jelle Zijlstra
8b58371278
subprocess: new params in 3.11 ( #7907 )
2022-05-21 16:07:50 +01:00
David Robertson
e5594aac60
Annotations for psycopg2.ConnectionInfo ( #7834 )
...
* Annotations for psycopg2.ConnectionInfo
These annotations come from the documentation here:
https://www.psycopg.org/docs/extensions.html#psycopg2.extensions.ConnectionInfo
If there was doubt, I referred to the libpq documentation cited by
psycopg2's docs.
I wasn't completely sure about `dsn_parameters`. Psycopg2's docs list it
as an `dict`, and the example suggests it's a `dict[str, str]` at that.
From psycopg2's source I found
1d3a89a0bb/psycopg/conninfo_type.c (L183-L206)
which is implemented here:
1d3a89a0bb/psycopg/utils.c (L251-L279)
I'm no expert in CPython's API, but this looks to me like it's building
a `dict[str, str]`.
Additionally, the libpq docs
https://www.postgresql.org/docs/current/libpq-connect.html#LIBPQ-PQCONNINFO
https://www.postgresql.org/docs/current/libpq-connect.html#LIBPQ-PQCONNDEFAULTS
show that the underlying data just consists of strings.
Additionally, I'm pretty sure from this chunk of source
1d3a89a0bb/psycopg/conninfo_type.c (L581-L598)
That `ConnectionInfo.__init__` takes one positional-only argument, which
must be a `psycopg2.connection`. But I don't think users are intended to
be constructing this type, so I've not added that annotation.
* Annotate `connection.info` and related attributes
* Make ConnectionInfo attributes properties
According to 1d3a89a0bb/psycopg/conninfo_type.c (L534-L563)
* Mark connection attributes as readonly
according to 8ef195f2ff/psycopg/connection_type.c (L1244)
* Explain why some properties aren't `T | None`
2022-05-21 07:38:52 -07:00
Jelle Zijlstra
0198d75153
types: 3.11 (and earlier) fixes ( #7900 )
...
- New attributes in 3.11
- Arguments to CodeType.__init__ are always positional (have fun)
- co_consts can use object instead of Any
2022-05-21 07:37:11 -07:00
Alex Waygood
76a4bd796b
Simplify and correct many numeric unions ( #7906 )
...
Unblocks PyCQA/flake8-pyi#222
2022-05-21 15:25:00 +01:00
Jelle Zijlstra
7576805aee
zipfile: 3.11 fixes ( #7898 )
...
* zipfile: 3.11 fixes
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com >
2022-05-21 09:22:32 +01:00
Alex Waygood
aa70cc8abb
CI: Don't install tomli for stubtest_stdlib ( #7901 )
...
The script doesn't import tomli
2022-05-20 19:33:06 -07:00
Jelle Zijlstra
585517b20e
stubtest: put all the incorrect __get__ sigs together ( #7899 )
2022-05-20 09:47:02 -07:00
Alex Waygood
d54d018e2d
CI: Fix stubtest workflow; make paths for other workflows more specific ( #7897 )
2022-05-20 16:37:02 +01:00
Alex Waygood
e2cf6c21d5
CI: Don't run stubtest_stdlib on PRs that don't touch the stdlib ( #7895 )
...
* Don't run stubtest_stdlib on PRs that don't touch the stdlib
2022-05-20 15:56:29 +01:00
Jukka Lehtosalo
cd039c9b73
Ignore mypy errors in Python 2 builtins and typing ( #7894 )
...
Similar errors are already ignored in Python 3 stubs. The
errors break Python 2 tests of mypy.
See https://github.com/python/typeshed/issues/7367#issuecomment-1131472225
for context.
2022-05-20 15:27:09 +01:00
Sebastian Rittau
adf48f1f96
Only run CI on pushes to main/master and pull requests ( #7883 )
...
Currently, when a maintainer opens a pull request from a branch in the
typeshed repository (instead of a fork), CI is run twice: once for the
PR and once for the push. This not only wastes CI minutes (and it seems
we're running into some limits lately), but also makes CI output twice
as large and more difficult to process.
The disadvantage is that a PR needs to be opened to run CI.
Alternatively, CI can be triggered manually.
2022-05-20 14:14:25 +02:00
Jelle Zijlstra
e7cbde0928
xml: stubtest fixes ( #7890 )
2022-05-20 13:03:45 +02:00
Jaakko Lappalainen
611fa18b5c
Use bytes for resp param of imaplib.Internaldate2tuple ( #7892 )
2022-05-20 12:45:20 +02:00
Jelle Zijlstra
26fa57ab78
traceback: update for py311 ( #7886 )
2022-05-19 19:14:12 -07:00