BUG: GH36928 Allow dict_keys to be used as column names by read_csv

Author: abmyii

pandas/_libs/lib.pyx

@@ -1019,6 +1019,8 @@ cdef inline bint c_is_list_like(object obj, bint allow_sets) except -1:
         and not (util.is_array(obj) and obj.ndim == 0)
         # exclude sets if allow_sets is False
         and not (allow_sets is False and isinstance(obj, abc.Set))
+        # allow dict_keys objects
+        or isinstance(obj, abc.KeysView)
     )
 
 

pandas/tests/io/parser/test_common.py

@@ -2221,3 +2221,14 @@ def test_read_table_delim_whitespace_non_default_sep(all_parsers):
     )
     with pytest.raises(ValueError, match=msg):
         parser.read_table(f, delim_whitespace=True, sep=",")
+
+
+def test_dict_keys_as_names(all_parsers):
+    data = "a,b\n1,2"
+
+    keys = {'a': int, 'b': int}.keys()
+    parser = all_parsers
+
+    parser.read_csv(
+        StringIO(data), names=keys
+    )