Fix stdlib/disutils testing

[Avasam]

In hindsight the solution was quite simple: Just don't have setuptools installed in a test environment.
Turned on stubtest for distutils for python 3.10+
Fixed enough stubtest to get the CI green.

Missing variables and functions were added with stubgen. Then filled in necessary types for pyright to pass.

Solves one of the two concerns raised in #9520 (namely distutils being a mess to test, and untestable on 3.10+)
As for the second one: distutils shouldn't be imported with setuptools>=60 installed. But even if we wanted to support that use-case, we could.

[Avasam]

In its current state, this PR shows that the tests are run correctly on all version. Which I wanted to demo.
But I could split off stubtest allowlist entries fixes into a different PR, and add a few entries for now just to get the CI green.

[AlexWaygood]

LGTM. @hauntsaninja, does this seem like a reasonable approach to you? It involves adding a little bit of complexity to stubtest_stdlib.py, but I think it's worth it.

[hauntsaninja]

(only looked at stubtest_stdlib)

Approach looks fine, but let's remove the return code logic from the pip commands. We only care about preserving the stubtest return code, we don't care about pip's return codes. I'd also not capture output from those installs.

[Avasam]

I see pip's output is verbose enough printing our own error message is redundant

[github-actions]

Diff from mypy_primer, showing the effect of this PR on open source code:

pip (https://github.com/pypa/pip)
+ src/pip/_internal/locations/_distutils.py:60: error: Unused "type: ignore" comment

[hauntsaninja]

Thanks, changes to stubtest_stdlib.py lgtm!

Requested changes were made

[AlexWaygood]

Thank you!

[AlexWaygood]

I'd also not capture output from those installs.

Hmm, are you sure about that @hauntsaninja? The output from pip is now pretty noisy when I run stubtest_stdlib.py locally (which I do quite a lot):

(.venv) C:\Users\alexw\coding\typeshed>python tests/stubtest_stdlib.py
Collecting mypy==1.0
  Using cached mypy-1.0.0-cp311-cp311-win_amd64.whl (8.8 MB)
Collecting typing-extensions>=3.10
  Using cached typing_extensions-4.5.0-py3-none-any.whl (27 kB)
Collecting mypy-extensions>=0.4.3
  Using cached mypy_extensions-1.0.0-py3-none-any.whl (4.7 kB)
Installing collected packages: typing-extensions, mypy-extensions, mypy
Successfully installed mypy-1.0.0 mypy-extensions-1.0.0 typing-extensions-4.5.0

[notice] A new release of pip available: 22.3.1 -> 23.0.1
[notice] To update, run: C:\Users\alexw\AppData\Local\Temp\tmppzy1r9wi\Scripts\python.exe -m pip install --upgrade pip
Found existing installation: setuptools 65.5.0
Uninstalling setuptools-65.5.0:
  Successfully uninstalled setuptools-65.5.0
C:\Users\alexw\AppData\Local\Temp\tmppzy1r9wi\Scripts\python.exe -m mypy.stubtest --check-typeshed --custom-typeshed-dir . --allowlist tests\stubtest_allowlists\py3_common.txt --allowlist tests\stubtest_allowlists\py311.txt --allowlist tests\stubtest_allowlists\win32.txt --allowlist tests\stubtest_allowlists\win32-py311.txt
Success: no issues found in 484 modules
stubtest succeeded

I'd find the output easier to parse if we suppressed pip's prints to the terminal (other than when pip errors out, obviously).