On Fri, Dec 24, 2021 at 01:54:35PM +1100, Chris Angelico wrote: > My reading of this is that a function IS the type of a function with > that signature, just like how None means the type NoneType. Is that > correct?
That's not the status quo, but I think the idea is that it will be. Except that I think that the preferred terminology is that it is a *function prototype* rather than a type, since it is a prototype, not an actual class you can instantiate. A template, if you will: "Any callable with the same signature (modulo parameter names) as this template / protocol is acceptable." > Or putting it another way: is this (silly) example legal? [snip example] I think maybe it should be, but right now, mypy reports it as illegal: [steve ~]$ mypy ~/func_prototype.py /home/steve/func_prototype.py:7: error: Function "func_prototype.repeat_string" is not valid as a type If we allow this, I see no reason why functions used as prototypes should be required to have an empty body. (Ellipsis or pass.) I think that actual, useful functions with an implementation should be allowed, as in your example. -- Steve _______________________________________________ Python-Dev mailing list -- python-dev@python.org To unsubscribe send an email to python-dev-le...@python.org https://mail.python.org/mailman3/lists/python-dev.python.org/ Message archived at https://mail.python.org/archives/list/python-dev@python.org/message/JU2JCG5TLKA35AGD636OJH7NWQ75HFMX/ Code of Conduct: http://python.org/psf/codeofconduct/
