Merge pull request #6 from mroeschke/bug/to_datetime/format_range

Fix to_datetime with format allowing out-of-range values

python/cudf/cudf/core/column/string.py

@@ -5665,16 +5665,37 @@ def _as_datetime_or_timedelta_column(self, dtype, format):
         if (self == "None").any():
             raise ValueError("Could not convert `None` value to datetime")
 
+        is_nat = self == "NaT"
+        if dtype.kind == "M":
+            without_nat = self.apply_boolean_mask(is_nat.unary_operator("not"))
+            all_same_length = (
+                libstrings.count_characters(without_nat).distinct_count(
+                    dropna=True
+                )
+                == 1
+            )
+            if not all_same_length:
+                # Unfortunately disables OK cases like:
+                # ["2020-01-01", "2020-01-01 00:00:00"]
+                # But currently incorrect for cases like (drops 10):
+                # ["2020-01-01", "2020-01-01 10:00:00"]
+                raise NotImplementedError(
+                    "Cannot parse date-like strings with different formats"
+                )
+            valid_ts = str_cast.istimestamp(self, format)
+            valid = valid_ts | is_nat
+            if not valid.all():
+                raise ValueError(f"Column contains invalid data for {format=}")
+
         casting_func = (
             str_cast.timestamp2int
             if dtype.type == np.datetime64
             else str_cast.timedelta2int
         )
         result_col = casting_func(self, dtype, format)
 
-        boolean_match = self == "NaT"
-        if (boolean_match).any():
-            result_col[boolean_match] = None
+        if is_nat.any():
+            result_col[is_nat] = None
 
         return result_col
 

python/cudf/cudf/tests/test_datetime.py

@@ -804,6 +804,16 @@ def test_to_datetime_format(data, format, infer_datetime_format):
     assert_eq(actual, expected)
 
 
+def test_to_datetime_data_out_of_range_for_format():
+    with pytest.raises(ValueError):
+        cudf.to_datetime("2015-02-99", format="%Y-%m-%d")
+
+
+def test_to_datetime_different_formats_notimplemented():
+    with pytest.raises(NotImplementedError):
+        cudf.to_datetime(["2015-02-01", "2015-02-01 10:10:10"])
+
+
 def test_datetime_can_cast_safely():
 
     sr = cudf.Series(

python/cudf/cudf/tests/test_index.py

@@ -2272,7 +2272,7 @@ def test_index_nan_as_null(data, nan_idx, NA_idx, nan_as_null):
         [],
         ["this", "is"],
         [0, 19, 13],
-        ["2019-01-01 04:00:00", "2019-01-01 06:00:00", "2018-03-02"],
+        ["2019-01-01 04:00:00", "2019-01-01 06:00:00", "2018-03-02 10:00:00"],
     ],
 )
 def test_isin_index(data, values):