Support timestamp characteristics

src/volur/sdk/v1alpha2/sources/csv/base.py

@@ -5,7 +5,7 @@
 from datetime import datetime
 from typing import Any, AsyncIterator, Literal
 
-from google.type.date_pb2 import Date
+from google.type.date_pb2 import Date, Datetime
 from volur.pork.materials.v1alpha3 import material_pb2
 from volur.pork.products.v1alpha3 import product_pb2
 from volur.pork.shared.v1alpha1 import characteristic_pb2, quantity_pb2
@@ -374,3 +374,79 @@ def get_value(
             raise ValueError(
                 f"provided value {_} in column {self.column_id} can not be interpreted as date characteristic"  # noqa: E501
             )
+
+
+@dataclass
+class CharacteristicColumnTimestamp(CharacteristicColumn):
+    timestamp_formats: list[str] = field(init=False)
+    default_timestamp_format: list[str] = field(
+        default_factory=lambda: [
+            "%d-%m-%Y %H:%M:%S",
+            "%Y-%m-%d %H:%M:%S",
+            "%Y/%m/%d %H:%M:%S",
+            "%d-%m-%Y %H:%M:%S",
+            "%d/%m/%Y %H:%M:%S",
+            "%d-%m-%Y %I:%M:%S %p",
+            "%Y-%m-%d %I:%M:%S %p",
+            "%Y/%m/%d %I:%M:%S %p",
+            "%d-%m-%Y %I:%M:%S %p",
+            "%d/%m/%Y %I:%M:%S %p",
+            "%d-%m-%Y %H:%M:%S.%f",
+            "%Y-%m-%d %H:%M:%S.%f",
+            "%Y/%m/%d %H:%M:%S.%f",
+            "%d-%m-%Y %H:%M:%S.%f",
+            "%d/%m/%Y %H:%M:%S.%f",
+        ],
+        init=False,
+    )
+    extra_timestamp_formats: list[str] = field(default_factory=list)
+
+    def __post_init__(
+        self: "CharacteristicColumnDate",
+        column_name: str | int,
+        characteristic_name: str,
+    ) -> None:
+        super().__post_init__(column_name, characteristic_name)
+        self.timestamp_formats = [
+            *self.default_timestamp_format,
+            *self.extra_timestamp_formats,
+        ]
+
+    def get_value(
+        self: "CharacteristicColumnDate",
+        data: dict[str | int, Any],
+    ) -> characteristic_pb2.Characteristic:
+        _ = data.get(self.column_id, None)
+        if _ is None or _ == "":
+            return characteristic_pb2.Characteristic(
+                name=self.characteristic_id,
+                value=characteristic_pb2.CharacteristicValue(),
+            )
+
+        elif isinstance(_, str):
+            for timestamp_format in self.timestamp_formats:
+                try:
+                    parsed_timestamp = datetime.strptime(_, timestamp_format)
+
+                    return characteristic_pb2.Characteristic(
+                        name=self.characteristic_id,
+                        value=characteristic_pb2.CharacteristicValue(
+                            value_date=Datetime(
+                                year=parsed_timestamp.year,
+                                month=parsed_timestamp.month,
+                                day=parsed_timestamp.day,
+                                hours=parsed_timestamp.hour,
+                                minutes=parsed_timestamp.minute,
+                                seconds=parsed_timestamp.second,
+                            )
+                        ),
+                    )
+                except ValueError:
+                    pass
+            raise ValueError(
+                f"provided value {_} in column {self.column_id} has invalid date format"
+            )
+        else:
+            raise ValueError(
+                f"provided value {_} in column {self.column_id} can not be interpreted as date characteristic"  # noqa: E501
+            )

tests/test_sources/csv/v1alpha2/test_characteristic_column_timestamp.py

@@ -0,0 +1,133 @@
+from typing import Any
+
+import pytest
+from google.type.date_pb2 import Datetime
+from volur.pork.shared.v1alpha1 import characteristic_pb2
+from volur.sdk.v1alpha2.sources.csv import CharacteristicColumnDate
+
+ids = [
+    "return-correct-characteristic-for-a-correct-date-value-in-column1",
+    "return-correct-characteristic-for-a-correct-date-value-in-column2",
+    "return-empty-chracteristic-because-value-in-column-is-none",
+    "raise-exception-because-value-is-not-valid-date",
+    "raise-exception-because-value-is-not-string",
+]
+
+test_data = [
+    (
+        {
+            "column_name": "15-06-2021 15:30:34",
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        characteristic_pb2.Characteristic(
+            name="characteristic_name",
+            value=characteristic_pb2.CharacteristicValue(
+                value_date=Datetime(year=2021, month=6, day=15, hours=15, minutes=30, seconds=34),
+            ),
+        ),
+        False,
+        None,
+    ),
+    (
+        {
+            "column_name": "10/06/2018 03:30:15 PM",
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        characteristic_pb2.Characteristic(
+            name="characteristic_name",
+            value=characteristic_pb2.CharacteristicValue(
+                value_date=Datetime(year=2018, month=6, day=10, hours=15, minutes=30, seconds=15),
+            ),
+        ),
+        False,
+        None,
+    ),
+    (
+        {
+            "column_name": "10/06/2018 03:10:34.45",
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        characteristic_pb2.Characteristic(
+            name="characteristic_name",
+            value=characteristic_pb2.CharacteristicValue(
+                value_date=Datetime(year=2018, month=6, day=10, hours=3, minutes=10, seconds=34),
+            ),
+        ),
+        False,
+        None,
+    ),
+    (
+        {
+            "column_name": "",
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        characteristic_pb2.Characteristic(
+            name="characteristic_name",
+            value=characteristic_pb2.CharacteristicValue(),
+        ),
+        False,
+        None,
+    ),
+    (
+        {
+            "column_name": "01-01-202",
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        None,
+        True,
+        "provided value 01-01-202 in column column_name has invalid date format",
+    ),
+    (
+        {
+            "column_name": 2020,
+        },
+        CharacteristicColumnDate(
+            column_name="column_name",
+            characteristic_name="characteristic_name",
+        ),
+        None,
+        True,
+        "provided value 2020 in column column_name can not be interpreted as date characteristic",  # noqa: E501
+    ),
+]
+
+
+@pytest.mark.parametrize(
+    argnames=[
+        "data",
+        "column",
+        "expected",
+        "should_raise_an_exception",
+        "exception_text",
+    ],
+    argvalues=test_data,
+    ids=ids,
+)
+def test_characteristic_column_date(
+    data: dict[str | int, Any],
+    column: CharacteristicColumnDate,
+    expected: characteristic_pb2.CharacteristicValue | None,
+    should_raise_an_exception: bool,
+    exception_text: str | None,
+) -> None:
+    if should_raise_an_exception:
+        with pytest.raises(ValueError, match=exception_text):
+            actual = column.get_value(data)
+    else:
+        actual = column.get_value(data)
+        assert actual == expected