Skip to content

bpo-44653: Support typing types in parameter substitution in the union type. #27247

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jul 22, 2021
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 30 additions & 0 deletions Lib/test/test_types.py
Original file line number Diff line number Diff line change
Expand Up @@ -772,6 +772,36 @@ def test_union_parameter_chaining(self):
self.assertEqual((list[T] | list[S])[int, T], list[int] | list[T])
self.assertEqual((list[T] | list[S])[int, int], list[int])

def test_union_parameter_substitution(self):
def eq(actual, expected):
self.assertEqual(actual, expected)
self.assertIs(type(actual), type(expected))

T = typing.TypeVar('T')
S = typing.TypeVar('S')
NT = typing.NewType('NT', str)
x = int | T | bytes

eq(x[str], int | str | bytes)
eq(x[list[int]], int | list[int] | bytes)
eq(x[typing.List], int | typing.List | bytes)
eq(x[typing.List[int]], int | typing.List[int] | bytes)
eq(x[typing.Hashable], int | typing.Hashable | bytes)
eq(x[collections.abc.Hashable],
int | collections.abc.Hashable | bytes)
eq(x[typing.Callable[[int], str]],
int | typing.Callable[[int], str] | bytes)
eq(x[collections.abc.Callable[[int], str]],
int | collections.abc.Callable[[int], str] | bytes)
eq(x[typing.Tuple[int, str]], int | typing.Tuple[int, str] | bytes)
eq(x[typing.Literal['none']], int | typing.Literal['none'] | bytes)
eq(x[str | list], int | str | list | bytes)
eq(x[typing.Union[str, list]], typing.Union[int, str, list, bytes])
eq(x[str | int], int | str | bytes)
eq(x[typing.Union[str, int]], typing.Union[int, str, bytes])
eq(x[NT], int | NT | bytes)
eq(x[S], int | S | bytes)

def test_union_parameter_substitution_errors(self):
T = typing.TypeVar("T")
x = int | T
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
Support :mod:`typing` types in parameter substitution in the union type.
25 changes: 12 additions & 13 deletions Objects/unionobject.c
Original file line number Diff line number Diff line change
Expand Up @@ -446,23 +446,22 @@ union_getitem(PyObject *self, PyObject *item)
return NULL;
}

// Check arguments are unionable.
PyObject *res;
Py_ssize_t nargs = PyTuple_GET_SIZE(newargs);
for (Py_ssize_t iarg = 0; iarg < nargs; iarg++) {
PyObject *arg = PyTuple_GET_ITEM(newargs, iarg);
int is_arg_unionable = is_unionable(arg);
if (is_arg_unionable <= 0) {
Py_DECREF(newargs);
if (is_arg_unionable == 0) {
PyErr_Format(PyExc_TypeError,
"Each union argument must be a type, got %.100R", arg);
if (nargs == 0) {
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I am not sure that this situation is possible now, but just for the case...

res = make_union(newargs);
}
else {
res = PyTuple_GET_ITEM(newargs, 0);
Py_INCREF(res);
for (Py_ssize_t iarg = 1; iarg < nargs; iarg++) {
PyObject *arg = PyTuple_GET_ITEM(newargs, iarg);
Py_SETREF(res, PyNumber_Or(res, arg));

This comment was marked as resolved.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, it has quadratic complexity. But I think that most union types has less than 10 args. If it be a problem (many thousands args in real code) we can add some optimizations, but for now I prefer simplicity.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good point! Old union de-duplication also had quadratic complexity. So I think it's ok for now. Especially since this is only for substitution, which isn't a common use case either.

if (res == NULL) {
break;
}
return NULL;
}
}

PyObject *res = make_union(newargs);

Py_DECREF(newargs);
return res;
}
Expand Down