README.rst

Graphene-Django-Extras

This package add some extra functionalities to graphene-django to facilitate the graphql use without Relay:

1. Allows pagination and filtering on Queries.
2. Allows to define DjangoRestFramework serializers based Mutations.
3. Allows use Directives on Queries and Fragments.

NOTE: Subscription support was moved to graphene-django-subscriptions

Installation:

For installing graphene-django-extras, just run this command in your shell:

pip install "graphene-django-extras"

Documentation:

Extra functionalities:

Fields:

1. DjangoObjectField
2. DjangoFilterListField
3. DjangoFilterPaginateListField
4. DjangoListObjectField (Recommended for Queries definition)

Mutations:

1. DjangoSerializerMutation (Recommended for Mutations definition)

Types:

1. DjangoListObjectType (Recommended for Types definition)
2. DjangoInputObjectType
3. DjangoSerializerType (Recommended for quick queries and mutations definitions)

Paginations:

1. LimitOffsetGraphqlPagination
2. PageGraphqlPagination

Queries and Mutations examples:

This is a basic example of graphene-django-extras package use. You can configure global params for DjangoListObjectType classes pagination definitions on settings.py like this:

GRAPHENE_DJANGO_EXTRAS = {
    'DEFAULT_PAGINATION_CLASS': 'graphene_django_extras.paginations.LimitOffsetGraphqlPagination',
    'DEFAULT_PAGE_SIZE': 20,
    'MAX_PAGE_SIZE': 50,
    'CACHE_ACTIVE': True,
    'CACHE_TIMEOUT': 300    # seconds
}

1- Types Definition:

from django.contrib.auth.models import User
from graphene_django_extras import DjangoListObjectType, DjangoSerializerType, DjangoObjectType
from graphene_django_extras.paginations import LimitOffsetGraphqlPagination

from .serializers import UserSerializer


class UserType(DjangoObjectType):
    class Meta:
        model = User
        description = " Type definition for a single user object "
        filter_fields = {
            'id': ['exact', ],
            'first_name': ['icontains', 'iexact'],
            'last_name': ['icontains', 'iexact'],
            'username': ['icontains', 'iexact'],
            'email': ['icontains', 'iexact']
        }


class UserListType(DjangoListObjectType):
    class Meta:
        description = " Type definition for users objects list "
        model = User
        pagination = LimitOffsetGraphqlPagination()


class UserModelType(DjangoSerializerType):
    """ With this type definition it't necessary a mutation definition for models """

    class Meta:
        description = " User's model type definition "
        serializer_class = UserSerializer
        pagination = LimitOffsetGraphqlPagination(default_limit=25, ordering="-username") # ordering can be: string, tuple or list
        filter_fields = {
            'id': ['exact', ],
            'first_name': ['icontains', 'iexact'],
            'last_name': ['icontains', 'iexact'],
            'username': ['icontains', 'iexact'],
            'email': ['icontains', 'iexact'],
            'is_staff': ['exact']
        }

2- You can to define InputTypes for use on mutations:

from graphene_django_extras import DjangoInputObjectType


class UserInput(DjangoInputObjectType):
    class Meta:
        description = " User InputType definition to use as input on an Arguments class on traditional Mutations "
        model = User

3- Defining Mutations:

You can define traditional mutations that use InputTypes or Mutations based on DRF serializers:

import graphene
from .serializers import UserSerializer
from graphene_django_extras import DjangoSerializerMutation
from .types import UserType
from .input_types import UserInputType


class UserSerializerMutation(DjangoSerializerMutation):
    """
        DjangoSerializerMutation auto implement Create, Delete and Update functions
    """
    class Meta:
        description = " DRF serializer based Mutation for Users "
        serializer_class = UserSerializer


class UserMutation(graphene.Mutation):
    """
        On traditional mutation classes definition you must implement the mutate function
    """

    user = graphene.Field(UserType, required=False)

    class Arguments:
        new_user = graphene.Argument(UserInput)

    class Meta:
        description = " Graphene traditional mutation for Users "

    @classmethod
    def mutate(cls, root, info, *args, **kwargs):
        ...

4- Defining schemes:

import graphene
from graphene_django_extras import DjangoObjectField, DjangoListObjectField, DjangoFilterPaginateListField, DjangoFilterListField, LimitOffsetGraphqlPagination
from .types import UserType, UserListType, UserModelType
from .mutations import UserMutation, UserSerializerMutation


class Queries(graphene.ObjectType):
    # Possible User list queries definitions
    users = DjangoListObjectField(UserListType, description=_('All Users query'))
    users1 = DjangoFilterPaginateListField(UserType, pagination=LimitOffsetGraphqlPagination())
    users2 = DjangoFilterListField(UserType)
    users3 = DjangoListObjectField(UserListType, filterset_class=UserFilter, description=_('All Users query'))

    # Defining a query for a single user
    # The DjangoObjectField have a ID type input field, that allow filter by id and is't necessary to define resolve function
    user = DjangoObjectField(UserType, description=_('Single User query'))

    # Another way to define a query to single user
    user1 = UserListType.RetrieveField(description=_('User List with pagination and filtering'))

    # Exist two ways to define single or list user queries with DjangoSerializerType
    user_retrieve1, user_list1 = UserModelType.QueryFields(description='Some description message for both queries',
                                                           deprecation_reason='Some deprecation message for both queries')
    user_retrieve2 = UserModelType.RetrieveField(description='Some description message for retrieve query',
                                                 deprecation_reason='Some deprecation message for retrieve query')
    user_list2 = UserModelType.ListField(description='Some description message for list query',
                                         deprecation_reason='Some deprecation message for list query')


class Mutations(graphene.ObjectType):
    user_create = UserSerializerMutation.CreateField(deprecation_reason='Some one deprecation message')
    user_delete = UserSerializerMutation.DeleteField()
    user_update = UserSerializerMutation.UpdateField()

    # Exist two ways to define mutations with DjangoSerializerType
    user_create1, user_delete1, user_update1 = UserModelType.MutationFields(
        description='Some description message for create, delete and update mutations',
        deprecation_reason='Some deprecation message for create, delete and update mutations')

    user_create2 = UserModelType.CreateField(description='Description message for create')
    user_delete2 = UserModelType.DeleteField(description='Description message for delete')
    user_update2 = UserModelType.UpdateField(description='Description message for update')

    traditional_user_mutation = UserMutation.Field()

5- Directives settings:

For use Directives you must follow two simple steps: 1. You must add 'graphene_django_extras.ExtraGraphQLDirectiveMiddleware' to your GRAPHENE dict config on your settings.py:

# settings.py

GRAPHENE = {
    'SCHEMA_INDENT': 4,
    'MIDDLEWARE': [
        'graphene_django_extras.ExtraGraphQLDirectiveMiddleware'
    ]
}

2. You must add the directives param with yours custom directives to your schema definition. This module come with some common directives for you, this directives allow to you format strings, numbers, lists, and dates (optional), and you can load like this:

# schema.py
from graphene_django_extras import all_directives

schema = graphene.Schema(
    query=RootQuery,
    mutation=RootMutation,
    directives=all_directives
)

NOTE: Date directive depends of dateutil module, so if you do not have installed it, this directive will not be available. You can install dateutil module manually:

pip install python-dateutil

or like this:

pip install graphene-django-extras[date]

That's all !!!

6- Complete Directive list:

FOR NUMBERS:

1. FloorGraphQLDirective: Floors value. Supports both String and Float fields.
2. CeilGraphQLDirective: Ceils value. Supports both String and Float fields.

FOR LIST:

1. ShuffleGraphQLDirective: Shuffle the list in place.
2. SampleGraphQLDirective: Take a 'k' int argument and return a k length list of unique elements chosen from the taken list

FOR DATE:

1. DateGraphQLDirective: Take a optional 'format' string argument and format the date from resolving the field by dateutil module with the 'format' format. Default format is: 'DD MMM YYYY HH:mm:SS' equivalent to '%d %b %Y %H:%M:%S' python format.

FOR STRING:

1. DefaultGraphQLDirective: Take a 'to' string argument. Default to given value if None or "".
2. Base64GraphQLDirective: Take a optional ("encode" or "decode") 'op' string argument(default='encode'). Encode or decode the string taken.
3. NumberGraphQLDirective: Take a 'as' string argument. String formatting like a specify Python number formatting.
4. CurrencyGraphQLDirective: Take a optional 'symbol' string argument(default="$"). Prepend the symbol to taken string and format it like a currency.
5. LowercaseGraphQLDirective: Lowercase the taken string.
6. UppercaseGraphQLDirective: Uppercase the taken string.
7. CapitalizeGraphQLDirective: Return the taken string with its first character capitalized and the rest lowered.
8. CamelCaseGraphQLDirective: CamelCase the taken string.
9. SnakeCaseGraphQLDirective: SnakeCase the taken string.
10. KebabCaseGraphQLDirective: SnakeCase the taken string.
11. SwapCaseGraphQLDirective: Return the taken string with uppercase characters converted to lowercase and viceversa.
12. StripGraphQLDirective: Take a optional 'chars' string argument(default=" "). Return the taken string with the leading and trailing characters removed. The 'chars' argument is not a prefix or suffix; rather, all combinations of its values are stripped.
13. TitleCaseGraphQLDirective: Return the taken string titlecased, where words start with an uppercase character and the remaining characters are lowercase.
14. CenterGraphQLDirective: Take a 'width' string argument and a optional 'fillchar' string argument(default=" "). Return the taken string centered with the 'width' argument as new length. Padding is done using the specified 'fillchar' argument. The original string is returned if 'width' argument is less than or equal to taken string length.
15. ReplaceGraphQLDirective: Take two strings arguments 'old' and 'new', and a optional integer argument 'count'. Return the taken string with all occurrences of substring 'old' argument replaced by 'new' argument value. If the optional argument 'count' is given, only the first 'count' occurrences are replaced.

7- Queries's examples:

{
    allUsers(username_Icontains:"john"){
        results(limit:5, offset:5){
            id
            username
            firstName
            lastName
        }
        totalCount
    }

    allUsers1(lastName_Iexact:"Doe", limit:5, offset:0){
        id
        username
        firstName
        lastName
    }

    allUsers2(firstName_Icontains: "J"){
        id
        username
        firstName
        lastName
    }

    user(id:2){
        id
        username
        firstName
    }

    user1(id:2){
        id
        username
        firstName
    }
}

8- Mutations's examples:

mutation{
    userCreate(newUser:{password:"test*123", email: "[email protected]", username:"test"}){
        user{
            id
            username
            firstName
            lastName
        }
        ok
        errors{
            field
            messages
        }
    }

    userDelete(id:1){
        ok
        errors{
            field
            messages
        }
    }

    userUpdate(newUser:{id:1, username:"John"}){
        user{
            id
            username
        }
        ok
        errors{
            field
            messages
        }
    }
}

9- Directives's examples:

Let's suppose that we have this query:

query{
    allUsers{
        result{
            id
            firstName
            lastName
            dateJoined
            lastLogin
        }
    }
}

And return this data:

{
  "data": {
    "allUsers": {
      "results": [
        {
            "id": "1",
            "firstName": "JOHN",
            "lastName": "",
            "dateJoined": "2017-06-20 09:40:30",
            "lastLogin": "2017-08-05 21:05:02"
        },
        {
            "id": "2",
            "firstName": "Golden",
            "lastName": "GATE",
            "dateJoined": "2017-01-02 20:36:45",
            "lastLogin": "2017-06-20 10:15:31"
        },
        {
            "id": "3",
            "firstName": "Nike",
            "lastName": "just do it!",
            "dateJoined": "2017-08-30 16:05:20",
            "lastLogin": "2017-12-05 09:23:09"
        }
      ]
    }
  }
}

As we see, some data it's missing or just not have the format that we like it, so let's go to format the output data that we desired:

query{
    allUsers{
        result{
            id
            firstName @capitalize
            lastName @default(to: "Doe") @title_case
            dateJoined @date(format: "DD MMM YYYY HH:mm:SS")
            lastLogin @date(format: "time ago")
        }
    }
}

And we get this output data:

{
  "data": {
    "allUsers": {
      "results": [
        {
            "id": "1",
            "firstName": "John",
            "lastName": "Doe",
            "dateJoined": "20 Jun 2017 09:40:30",
            "lastLogin": "4 months, 12 days, 15 hours, 27 minutes and 58 seconds ago"
        },
        {
            "id": "2",
            "firstName": "Golden",
            "lastName": "Gate",
            "dateJoined": "02 Jan 2017 20:36:45",
            "lastLogin": "5 months, 28 days, 2 hours, 17 minutes and 53 seconds ago"
        },
        {
            "id": "3",
            "firstName": "Nike",
            "lastName": "Just Do It!",
            "dateJoined": "30 Aug 2017 16:05:20",
            "lastLogin": "13 days, 3 hours, 10 minutes and 31 seconds ago"
        }
      ]
    }
  }
}

As we see, the directives is a easy way to format output data on queries, and it's can be put together like a chain.

List of possible date's tokens: "YYYY", "YY", "WW", "W", "DD", "DDDD", "d", "ddd", "dddd", "MM", "MMM", "MMMM", "HH", "hh", "mm", "ss", "A", "ZZ", "z".

You can use this shortcuts too:

1. "time ago"
2. "iso": "YYYY-MMM-DDTHH:mm:ss"
3. "js" or "javascript": "ddd MMM DD YYYY HH:mm:ss"

Change Log:

v0.3.7:

1. Improved DjangoListType and DjangoObjectType to share the filterset_class between the two class.

v0.3.6:

1. Improve DjangoSerializerMutation resolvers.

v0.3.5:

1. Fixed minor bug on ExtraGraphQLDirectiveMiddleware.
2. Fixed error with DRF 3.8 Compatibility.
3. Updated List's Fields to pass info.context to filterset as request, this allow filtering by request data.
4. Added new feature to ordering paginated queries.

v0.3.4-alpha2:

1. Fixed minor bug on DjangoListObjectType.

v0.3.4-alpha1:

1. Added filterset_class to the listing types as default filter.
2. Changed getOne by RetrieveField on DjangoListObjectType.

v0.3.3:

1. Added filterset_class to DjangoObjectType.
2. Fixed minor bug on factory_types function.

v0.3.3-alpha1:

1. Fixed minor bug on queryset_factory function.

v0.3.2:

1. Updated Date directive format function for better string format combinations.
2. Updated custom Time, Date and DateTime base types to be used with Date directive.
3. Fixed bug with caching Introspection queries on ExtraGraphQLView.

v0.3.1:

1. Fixed bug with default Date directive format.

v0.3.0:

1. Added Binary graphql type. A BinaryArray is used to convert a Django BinaryField to the string form.
2. Added 'CACHE_ACTIVE' and 'CACHE_TIMEOUT' config options to GRAPHENE_DJANGO_EXTRAS settings for activate cache queries result and define a expire time. Default values are: CACHE_ACTIVE=False, CACHE_TIMEOUT=300 (5 minutes).
3. Updated Date directive for use with Django TimeField, DateField, and DateTimeField.
4. Updated ExtraGraphQLView and AuthenticatedGraphQLView to allow use subscription requests on graphene-django >=2.0
5. Updated setup dependence to graphene-django>=2.0.

v0.2.2:

1. Fixed performance bug on some queries when request nested ManyToMany fields.

v0.2.1:

1. Fixed bug with default PaginationClass and DjangoFilterPaginateListField.

v0.2.0:

1. Added some useful directives to use on queries and fragments.
2. Fixed error on DjangoFilterPaginateListField resolve function.

v0.1.6:

1. Fixed bug on create and update function on serializer mutation.

v0.1.3:

1. Fixed some minors bugs.

v0.1.2:

1. Added ok field and errors field to DjangoSerializerType like on DjangoSerializerMutation.
2. Added possibility of filtering in those queries fields that return a list of objects.
3. Updated DRF compatibility.
4. Fixed bug with filters when use global DEFAULT_PAGINATION_CLASS.

v0.1.1:

1. Fixed error with JSONField reference on Django==1.8.x installations.

v0.1.0:

1. Added DjangoSerializerType for quick Django's models types definition (See documentation).
2. Moved support for Subscriptions to graphene-django-subscriptions packages for incompatibility with graphene-django>=2.0.
3. Fixed bug on DjangoFilterPaginateListField's pagination.

v0.1.0-alpha12:

1. Added new settings param: MAX_PAGE_SIZE, to use on GRAPHENE_DJANGO_EXTRAS configuration dict for better customize DjangoListObjectType's pagination.
2. Added support to Django's field: GenericRel.
3. Improve model's fields calculation for to add all possible related and reverse fields.
4. Improved documentation translation.

v0.1.0-alpha11:

1. Improved ordering for showed fields on graphqli's IDE.
2. Added better descriptions for auto generated fields.

v0.1.0-alpha10:

1. Improve converter.py file to avoid create field for auto generate OneToOneField product of an inheritance.
2. Fixed bug in Emun generation for fields with choices of model inheritance child.

v0.1.0-alpha9:

1. Fixed bug on GenericType and GenericInputType generations for Queries list Type and Mutations.

v0.1.0-alpha6:

1. Fixed with exclude fields and converter function.

v0.1.0-alpha5:

1. Updated to graphene-django>=2.0.
2. Fixed minor bugs on queryset_builder performance.

v0.1.0-alpha4:

1. Add queryset options to DjangoListObjectType Meta class for specify wanted model queryset.
2. Add AuthenticatedGraphQLView on graphene_django_extras.views for use 'permission', 'authorization' and 'throttle' classes based on the DRF settings. Special thanks to @jacobh for this comment.

v0.1.0-alpha3:

1. Fixed bug on subscriptions when not specified any field in "data" parameter to bean return on notification message.

v0.1.0-alpha2:

1. Fixed bug when subscribing to a given action (create, update or delete).
2. Added intuitive and simple web tool to test notifications of graphene-django-extras subscription.

v0.1.0-alpha1:

1. Added support to multiselect choices values for models.CharField with choices attribute, on queries and mutations. Example: Integration with django-multiselectfield package.
2. Added support to GenericForeignKey and GenericRelation fields, on queries and mutations.
3. Added first approach to support Subscriptions with Channels, with subscribe and unsubscribe operations. Using channels-api package.
4. Fixed minors bugs.

v0.0.4:

1. Fix error on DateType encode.

v0.0.3:

1. Implement custom implementation of DateType for use converter and avoid error on Serializer Mutation.

v0.0.2:

1. Changed dependency of DRF to 3.6.4 on setup.py file, to avoid an import error produced by some changes in new version of DRF=3.7.0 and because DRF 3.7.0 dropped support to Django versions < 1.10.

v0.0.1:

1. Fixed bug on DjangoInputObjectType class that refer to unused interface attribute.
2. Added support to create nested objects like in DRF <http://www.django-rest-framework.org/api-guide/serializers/#writable-nested-representations>, it's valid to SerializerMutation and DjangoInputObjectType, only is necessary to specify nested_fields=True on its Meta class definition.
3. Added support to show, only in mutations types to create objects and with debug=True on settings, inputs autocomplete ordered by required fields first.
4. Fixed others minors bugs.

v0.0.1-rc.2:

1. Make queries pagination configuration is more friendly.

v0.0.1-rc.1:

1. Fixed a bug with input fields in the converter function.

v0.0.1-beta.10:

1. Fixed bug in the queryset_factory function because it did not always return a queryset.

v0.0.1-beta.9:

1. Remove hard dependence with psycopg2 module.
2. Fixed bug that prevented use queries with fragments.
3. Fixed bug relating to custom django_filters module and ordering fields.

v0.0.1-beta.6:

1. Optimizing imports, fix some minors bugs and working on performance.

v0.0.1-beta.5:

1. Repair conflict on converter.py, by the use of get_related_model function with: OneToOneRel, ManyToManyRel and ManyToOneRel.

v0.0.1-beta.4:

1. First commit