clarify pyarrow tests

Author: MarcoGorelli

pandas/_libs/tslibs/parsing.pyx

@@ -1101,13 +1101,15 @@ cdef void _maybe_warn_about_dayfirst(format: str, bint dayfirst):
         if (day_index > month_index) and dayfirst:
             warnings.warn(
                 f"Parsing dates in {format} format when dayfirst=True was specified. "
-                "Pass `dayfirst=False` or `format='{format}'` to silence this warning.",
+                f"Pass `dayfirst=False` or explicitly specify a format to silence "
+                "this warning.",
                 stacklevel=find_stack_level(inspect.currentframe()),
             )
         if (day_index < month_index) and not dayfirst:
             warnings.warn(
                 f"Parsing dates in {format} format when dayfirst=False was specified. "
-                "Pass `dayfirst=True` or `format='{format}'` to silence this warning.",
+                f"Pass `dayfirst=True` or explicitly specify a format to silence "
+                "this warning.",
                 stacklevel=find_stack_level(inspect.currentframe()),
             )
 

pandas/io/parsers/base_parser.py

@@ -1234,7 +1234,9 @@ def _isindex(colspec):
                     continue
                 # Pyarrow engine returns Series which we need to convert to
                 # numpy array before converter, its a no-op for other parsers
-                data_dict[colspec] = converter(np.asarray(data_dict[colspec]))
+                inp = np.asarray(data_dict[colspec])
+                res = converter(inp)
+                data_dict[colspec] = res
             else:
                 new_name, col, old_names = _try_convert_dates(
                     converter, colspec, data_dict, orig_names

pandas/tests/io/parser/test_parse_dates.py

@@ -1245,7 +1245,10 @@ def test_bad_date_parse(all_parsers, cache_dates, value):
     parser = all_parsers
     s = StringIO((f"{value},\n") * 50000)
 
-    if "PyArrowParser" in parser.__str__():
+    if parser.engine == "pyarrow":
+        # None in input gets converted to 'None', for which
+        # pandas tries to guess the datetime format, triggering
+        # the warning. TODO: parse dates directly in pyarrow, see
         # https://github.com/pandas-dev/pandas/issues/48017
         warn = UserWarning
     else:
@@ -1266,12 +1269,15 @@ def test_bad_date_parse(all_parsers, cache_dates, value):
 @pytest.mark.parametrize("value", ["0"])
 def test_bad_date_parse_with_warning(all_parsers, cache_dates, value):
     # if we have an invalid date make sure that we handle this with
-    # and w/o the cache properly
+    # and w/o the cache properly. TODO: parse dates directly in pyarrow, see
+    # https://github.com/pandas-dev/pandas/issues/48017
+
     parser = all_parsers
     s = StringIO((f"{value},\n") * 50000)
 
-    if "PyArrowParser" in parser.__str__():
-        # https://github.com/pandas-dev/pandas/issues/48017
+    if parser.engine == "pyarrow":
+        # pyarrow reads "0" as 0 (of type int64), and so
+        # pandas doesn't try to guess the datetime format
         warn = None
     else:
         warn = UserWarning