I propose changingPyDict_Copy() API to always return adict. Previously, it returned afrozendict if it was called on afrozendict. IMO it's less surprising and more convenient that PyDict_Copy() always return a copy (as adict).

Example intypeobject.c:

staticPyTypeObject*type_new_init(type_new_ctx*ctx){PyObject*dict=PyDict_Copy(ctx->orig_dict);if (dict==NULL) {        gotoerror;    }        ...set_tp_dict(type,dict);    ...}

The function acceptsdict orfrozendict but the type dictionary (dict variable) must be adict.

Example inModules/_elementtree.c:

staticPyObject*get_attrib_from_keywords(PyObject*kwds){    ...if (attrib) {        ...Py_SETREF(attrib,PyDict_Copy(attrib));    }else {attrib=PyDict_New();    }if (attrib!=NULL&&PyDict_Update(attrib,kwds)<0) {Py_DECREF(attrib);returnNULL;    }returnattrib;}

(With my pending change PRgh-145508), the function acceptsdict orfrozendict but the result must be adict (for example, we callPyDict_Update() on it).

cc@corona10@ZeroIntensity@eendebakpt

Currently,PyDict_Copy(frozendict) returns afrozendict copy which is something used internally tomodify an immutablefrozendict! But I don't think that we should leak such implementation details to the public C API.

Existing C extensions callingPyDict_Copy() expect it to return adict. If it can return afrozendict on Python 3.15, it can trigger bugs.

Always returningdict avoids these issues.