Don't use 'is' with literals

It causes a SyntaxWarning on install, and possibly undefined test results

Author: Rebecca N. Palmer <[email protected]>
Bug-Debian: https://bugs.debian.org/956021
Forwarded: pandas-dev/pandas#33322


Gbp-Pq: Name fix_is_warnings.patch

Author: Debian Science Team

pandas/tests/frame/test_alter_axes.py

@@ -236,9 +236,16 @@ def test_set_index_pass_arrays_duplicate(
 
         # need to adapt first drop for case that both keys are 'A' --
         # cannot drop the same column twice;
-        # use "is" because == would give ambiguous Boolean error for containers
+        # plain == would give ambiguous Boolean error for containers
         first_drop = (
-            False if (keys[0] is "A" and keys[1] is "A") else drop  # noqa: F632
+            False
+            if (
+                type(keys[0]) == str
+                and keys[0] == "A"
+                and type(keys[1]) == str
+                and keys[1] == "A"
+            )
+            else drop
         )
         # to test against already-tested behaviour, we add sequentially,
         # hence second append always True; must wrap keys in list, otherwise