Colon slice behaves differently from built-in slice function

[gao462]

Bug Report

Slice using colon notation has different behavior from slice using built-in slice function.

To Reproduce

Use colon notation will have error message.

import numpy as onp
import numpy.typing as onpt
imin: onpt.NDArray[onp.int64] = onp.array(0)
imax: onpt.NDArray[onp.int64] = onp.array(2)
M = onp.array([1.0, 2.0, 3.0, 4.0])
print(M[imin:imax])

error: Slice index must be an integer or None  [misc]
Found 1 error in 1 file (checked 1 source file)

But using built-in slice function has no problem.

import numpy as onp
import numpy.typing as onpt
imin: onpt.NDArray[onp.int64] = onp.array(0)
imax: onpt.NDArray[onp.int64] = onp.array(2)
M = onp.array([1.0, 2.0, 3.0, 4.0])
islice = slice(imin, imax)
print(M[islice])

Success: no issues found in 1 source file

Expected Behavior

Colon slice should has no error message as built-in slice function.

Actual Behavior

Colon slice has error message which is unexpected.

Your Environment

• Mypy version used: 0.910
• Mypy command-line flags:
• Mypy configuration options from mypy.ini (and other config files):

[tool.mypy]
plugins = "numpy.typing.mypy_plugin"
check_untyped_defs = true
disallow_any_generics = true
ignore_missing_imports = true
no_implicit_optional = true
show_error_codes = true
strict_equality = true
warn_redundant_casts = true
warn_return_any = true
warn_unreachable = true
warn_unused_configs = true
no_implicit_reexport = false

• Python version used: 3.9.7
• Operating system and version: Linux

[sobolevn]

CC @Zac-HD

It might be related to the recent feature you were interested in 🙂

[Zac-HD]

Nope, not related - my PR was type-level slicing, this is value-level.

[erictraut]

It looks like this bug has been fixed. I'm not able to repro the error using mypy 1.5.

[hauntsaninja]

I fixed this in #14738