This will probably need tests before it gets accepted, but I could not successfully run the tests in Lib/test/test_format.py ...

I used cargo run using the absolute path and also cargo run --release using the absolute path. Am I going about this incorrectly?

I believe you can use the following command to test your changes locally:

# this runs all of the tests (not necessary if you know which test is affected)
cargo run --release --features ssl,jit -- -m test -j 1 -u all --slowest --fail-env-changed -v
# this runs only the test suite named "test_whatever" (usually located at `Lib/test/test_whatever.py`)
cargo run --release --features ssl,jit -- -m test -j 1 -u all --slowest --fail-env-changed -v test_whatever

For a list of all resources and more command-line options, you can execute cargo run --release --features ssl,jit -- -m test -h.

To format the Rust code, try running cargo fmt --all.

Something's not quite right with either the tests I added for the N case or the action. I'm not getting local test failures at least.

Something's not quite right with either the tests I added for the N case or the action. I'm not getting local test failures at least.

I think I got everything figured out. Thanks for your help. A lot of my issues were because of my dev environment but I got them sorted out.

The issue with the test was caused by my assuming Python 3.11 supported the N float formatter code but it doesn't. Removed the functionality for that and the test should yield the same Error Py 3.11 does when N is used.

there's a passing test now in test_enum.py so 🎉 :

UNEXPECTED SUCCESS: test_format_enum_float (test.test_enum.TestEnum.test_format_enum_float)

there's also a panic in the formatting, looking at the runner logs it seems to be due to something being off during a slicing operation.

This test doesn't run on Windows (hence why you did not probably see this error). It blows up in test_locale of test_format and my guess would be format(1234.5, "n"). Try running that locally in the REPL and see what it does.

there's a passing test now in test_enum.py so 🎉 :

UNEXPECTED SUCCESS: test_format_enum_float (test.test_enum.TestEnum.test_format_enum_float)

there's also a panic in the formatting, looking at the runner logs it seems to be due to something being off during a slicing operation.

This test doesn't run on Windows (hence why you did not probably see this error). It blows up in test_locale of test_format and my guess would be format(1234.5, "n"). Try running that locally in the REPL and see what it does.

Thank you for the feedback! I think I fixed it and I also unmarked the unexpected success

You also fixed more test

UNEXPECTED SUCCESS: test_locale (test.test_format.FormatTest.test_locale)

and another failure

======================================================================
FAIL: test_int__format__ (test.test_types.TypesTests.test_int__format__)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/RustPython/RustPython/pylib/Lib/test/test_types.py", line 373, in test_int__format__
    self.assertRaises(ValueError, 0 .__format__, format_spec)
AssertionError: ValueError not raised by __format__

You also fixed more test

UNEXPECTED SUCCESS: test_locale (test.test_format.FormatTest.test_locale)

and another failure

======================================================================
FAIL: test_int__format__ (test.test_types.TypesTests.test_int__format__)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/RustPython/RustPython/pylib/Lib/test/test_types.py", line 373, in test_int__format__
    self.assertRaises(ValueError, 0 .__format__, format_spec)
AssertionError: ValueError not raised by __format__

Sorry I keep missing things. Can't run the tests on my dev environment so I have a wonky setup at the moment. Think I got them sorted out.

I allowed the N case to function for int formatting functionality when that wasn't supposed to happen. I did not want to introduce "Case" to the Number enum but I wanted the error messages to line up properly with Python 3.11. Hope this is okay

Can't run the tests on my dev environment

Generally good idea to push here and wait for results since other platforms might have issues you can't catch, you only need to be able to parse the logs (which is straight-forward after doing it once or twice but can get slightly tedious).

Can't run the tests on my dev environment

Generally good idea to push here and wait for results since other platforms might have issues you can't catch, you only need to be able to parse the logs (which is straight-forward after doing it once or twice but can get slightly tedious).

Glad to hear that, hopefully I won't hit your GitHub action minute limit haha. Can't recall if there's only limitations in private repos or not.

Forgot to run cargo fmt so the workflow will have to run at least once more.

@JazzGlobal If you are interested in supporting locale, you can start with #4609

@JazzGlobal If you are interested in supporting locale, you can start with #4609

Thank you, and thank you all for guiding me through my first Rust contribution 😁