ngettext has different annotations to ngettext_lazy

[delfick]

Hello,

I've noticed that the number argument to ngettext is different between the lazy and non lazy versions

django-stubs/django-stubs/utils/translation/__init__.pyi

Lines 12 to 20 in 789d40f

	def ngettext(singular: str, plural: str, number: float) -> str: ...
	def pgettext(context: str, message: str) -> str: ...
	def npgettext(context: str, singular: str, plural: str, number: int) -> str: ...
	# lazy evaluated translation functions
	def gettext_lazy(message: str) -> _StrPromise: ...
	def pgettext_lazy(context: str, message: str) -> _StrPromise: ...
	def ngettext_lazy(singular: str, plural: str, number: int | str | None = None) -> _StrPromise: ...
	def npgettext_lazy(context: str, singular: str, plural: str, number: int | str | None = None) -> _StrPromise: ...

In the non lazy version it's float but in the lazy version it's int | str | None

Should they both be int | str | None?

[federicobond]

The str part is intentional, it's actually documented here: https://docs.djangoproject.com/en/6.1/topics/i18n/translation/#lazy-translations-and-plural.

Interesting find! I did not know about that functionality until today.

Edit: the float / int difference does appear to be an oversight.

[delfick]

huh, thanks @federicobond !