improve StrEnum

[pmeier]

This adds a from_str class method to StrEnum to clean up the interface. This seems to be fully scriptable

import torch
from torchvision.prototype.utils._internal import StrEnum


class Foo(StrEnum):
    BAR = StrEnum.auto()
    BAZ = 42


@torch.jit.script
def check_str_enum(input: torch.Tensor, foo: Foo) -> torch.Tensor:
    if foo == Foo.BAR:
        return input.abs()
    elif foo == Foo.BAZ:
        return input + 1
    else:
        raise ValueError("This should not happen!")


input = torch.tensor(-1)

print("should be 1: ", check_str_enum(input, Foo.BAR))

print("should be 0: ", check_str_enum(input, Foo.from_str("BAZ")))

check_str_enum(input, Foo.from_str("CAR"))

should be 1:  tensor(1)
should be 0:  tensor(0)
ValueError: Unknown value 'CAR' for Foo. Did you mean 'BAR'?

[facebook-github-bot]

💊 CI failures summary and remediations

As of commit d990996 (more details on the Dr. CI page):

---

💚 💚 Looks good so far! There are no failures yet. 💚 💚

---

This comment was automatically generated by Dr. CI (expand for details).

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

[datumbox]

Looks good overall but it's blocked by the dependency on the internal of prototype. See comment below. Thoughts?

[github-actions]

Hey @pmeier!

You merged this PR, but no labels were added. The list of valid labels is available at https://github.com/pytorch/vision/blob/main/.github/process_commit.py