marshmallow-pynamodb

PynamoDB integration with the marshmallow (de)serialization library.

Installation

From PyPi:

$ pip install marshmallow-pynamodb

From GitHub:

$ pip install git+https://github.com/mathewmarcus/marshmallow-pynamodb#egg=marshmallow_pynamodb

Declare your models

from pynamodb.models import Model
from pynamodb.attributes import UnicodeAttribute

class User(Model):
    class Meta:
        table_name = "user"
    email = UnicodeAttribute(null=True)
    first_name = UnicodeAttribute(range_key=True)
    last_name = UnicodeAttribute(hash_key=True)

Generate marshmallow schemas

from marshmallow_pynamodb import ModelSchema

class UserSchema(ModelSchema):
    class Meta:
        model = User

user_schema = UserSchema()

(De)serialize your data

user = User(last_name="Smith", first_name="John")

user_schema.dump(user).data
# {u'first_name': u'John', u'last_name': u'Smith', u'email': None}

user_schema.load({"last_name": "Smith", "first_name": "John"}).data
# user<Smith>

Nested models? No problem

from marshmallow_pynamodb.schema import ModelSchema

from pynamodb.models import Model
from pynamodb.attributes import ListAttribute, MapAttribute, NumberAttribute, UnicodeAttribute

class Location(MapAttribute):
    latitude = NumberAttribute()
    longitude = NumberAttribute()
    name = UnicodeAttribute()


class Person(MapAttribute):
    firstName = UnicodeAttribute()
    lastName = UnicodeAttribute()
    age = NumberAttribute()


class OfficeEmployeeMap(MapAttribute):
    office_employee_id = NumberAttribute()
    person = Person()
    office_location = Location()


class Office(Model):
    class Meta:
        table_name = 'OfficeModel'

    office_id = NumberAttribute(hash_key=True)
    address = Location()
    employees = ListAttribute(of=OfficeEmployeeMap)


class OfficeSchema(ModelSchema):
    class Meta:
        model = Office


OfficeSchema().load({'office_id': 789,
                     'address': {'latitude': 6.98454,
                                 'longitude': 172.38832,
                                 'name': 'some_location'},
                     'employees': [{'office_employee_id': 123,
                                    'person': {'firstName': 'John',
                                               'lastName': 'Smith',
                                               'age': 45},
                                    'office_location': {'latitude': -24.0853,
                                                        'longitude': 144.87660,
                                                        'name': 'other_location'}},
                                   {'office_employee_id': 456,
                                    'person': {'firstName': 'Jane',
                                               'lastName': 'Doe',
                                               'age': 33},
                                    'office_location': {'latitude': -20.57989,
                                                        'longitude': 92.30463,
                                                        'name': 'yal'}}]}).data
# Office<789>

License

MIT licensed. See the bundled LICENSE file for more details.