Loading Datetime Objects

Maybe this is not so much of an issue but a misunderstanding on my side, so I’d think the following pytest-style unit test should not raise a validation error, since a datetime.datetime instance should satisfy the property of being a a datetime already:

from marshmallow import Schema, fields
import datetime


class SomeSchema(Schema):
    element = fields.DateTime()


def test_datetime_loading_accepts_datetime_object():
    schema = SomeSchema(strict=True)
    res = schema.load({"element": datetime.datetime.now()})

I do however get a validation error marshmallow.exceptions.ValidationError: {'element': ['Not a valid datetime.']}

[JulienPalard]

I’m temporarily using this so deserialization accepts datetimes instances:

class MyDateTimeField(fields.DateTime):
    def _deserialize(self, value, attr, data):
        if isinstance(value, datetime):
            return value
        return super()._deserialize(value, attr, data)

[aklaver]

If are serializing(dump) a DateTime field then you need to provide a datetime object:

import datetime
from marshmallow import Schema, fields
from dateutil import tz
class UserSchema(Schema):
    name = fields.Str()
    email = fields.Email()
    created_at = fields.DateTime()
usr_sch = UserSchema()
usr = {'name': "Adrian Klaver", 'email': "[e-mail]", 'created_at': datetime.datetime(2017, 11, 22, 9, 58, tzinfo=tz.tzutc())}
usr_sch.dump(usr)
MarshalResult(data={'name': 'Adrian Klaver', 'created_at': '2017-11-22T09:58:00+00:00', 'email': '[e-mail]'}, errors={})

If you are going the other way deserializing(load) then you can use a string:

usr = {'name': "Adrian Klaver", 'email': "[e-mail]", 'created_at': '2017-11-22T09:58:00Z'}
usr_sch.load(usr)
UnmarshalResult(data={'name': 'Adrian Klaver', 'created_at': datetime.datetime(2017, 11, 22, 9, 58, tzinfo=tzutc()), 'email': '[e-mail]'}, errors={})