objtuple: Go out of the way to support comparison of subclasses.
Two things are handled here: allow to compare native subtypes of tuple,
e.g. namedtuple (TODO: should compare type too, currently compared
duck-typedly by content). Secondly, allow user sunclasses of tuples
(and its subtypes) be compared either. "Magic" I did previously in
objtype.c covers only one argument (lhs is many), so we're in trouble
when lhs is native type - there's no other option besides handling
rhs in special manner. Fortunately, this patch outlines approach with
fast path for native types.
diff --git a/py/objtuple.c b/py/objtuple.c
index 7d4e877..ca65b28 100644
--- a/py/objtuple.c
+++ b/py/objtuple.c
@@ -99,12 +99,20 @@
// Don't pass MP_BINARY_OP_NOT_EQUAL here
STATIC bool tuple_cmp_helper(int op, mp_obj_t self_in, mp_obj_t another_in) {
- assert(MP_OBJ_IS_TYPE(self_in, &mp_type_tuple));
- if (!MP_OBJ_IS_TYPE(another_in, &mp_type_tuple)) {
- return false;
+ mp_obj_type_t *self_type = mp_obj_get_type(self_in);
+ if (self_type->getiter != tuple_getiter) {
+ assert(0);
}
+ mp_obj_type_t *another_type = mp_obj_get_type(another_in);
mp_obj_tuple_t *self = self_in;
mp_obj_tuple_t *another = another_in;
+ if (another_type->getiter != tuple_getiter) {
+ // Slow path for user subclasses
+ another = mp_instance_cast_to_native_base(another, &mp_type_tuple);
+ if (another == MP_OBJ_NULL) {
+ return false;
+ }
+ }
return mp_seq_cmp_objs(op, self->items, self->len, another->items, another->len);
}