带有Flask的石墨烯不会加载架构

Question

I'm trying to run this example for using Graphene with Flask. I make the models.py , schema.py , and app.py verbatim from that page, and put them all in the flask_sqlalchemy folder, as instructed. Then I make and populate the database with the instructions listed at the bottom.

Everything was going ok until I tried to run app.py . When I did that it gave me this error:

Traceback (most recent call last):
  File "./app.py", line 6, in <module>
    from schema import schema, Department
  File "C:\Users\asdf\Python\GraphQL\flask_sqlalchemy\schema.py", line 7, in <module>
    schema = graphene.Schema()
  File "C:\Users\asdf\Envs\GraphQL\lib\site-packages\graphene\types\schema.py", line 27, in __init__
    ).format(query)
AssertionError: Schema query must be Object Type but got: None.

So basically it failed on the line in schema.py that says:

schema = graphene.Schema()

And indeed, if I open a command prompt and do this it fails the same way:

>>> import graphene
>>> s  = graphene.Schema()
Traceback... (same traceback)

I'm using Python 3.5 on Windows 10. The only difference from the tutorial is that I use virtualenvwrapper-win instead of the regular virtualenv.

This is my first experience with graphene or graphql in general, and I'm sure it ends up being a stupid mistake. Any help would be greatly appreciated!

Thanks a lot, Alex

Answer 1

You need to provide at least a query object to your schema object, else there is nothing to "expose" through GraphQL.

class Employee(SQLAlchemyObjectType):

    class Meta:
        model = EmployeeModel


class Query(graphene.ObjectType):

    employee = graphene.Field(
        Employee, 
        employee_id=graphene.Argument(graphene.Integer)
    )

    def resolve_employee(self, args, context, info):
        """Resolves `employee` object on the root query"""
        employee_id = args.get('employee_id')
        employee = EmployeeModel.query.get(employee_id)
        return employee

# Provide the root query to schema
schema = graphene.Schema(query=query)

Answer 2

Thank you Ivan! Your response put me on the right track and I'd mark that as the correct answer, but the solution was even simpler.

schema.py in the tutorial is this:

# flask_sqlalchemy/schema.py
import graphene
from graphene import relay
from graphene_sqlalchemy import SQLAlchemyObjectType, SQLAlchemyConnectionField
from models import db_session, Department as DepartmentModel, Employee as EmployeeModel

schema = graphene.Schema()

class Department(SQLAlchemyObjectType):
    class Meta:
        model = DepartmentModel
        interfaces = (relay.Node, )

class Employee(SQLAlchemyObjectType):
    class Meta:
        model = EmployeeModel
        interfaces = (relay.Node, )

class Query(graphene.ObjectType):
    node = relay.Node.Field()
    all_employees = SQLAlchemyConnectionField(Employee)

schema.query = Query

But as Ivan pointed out, the graphene.Schema() function needs to be handed a query, so it fails at schema = graphene.Schema() . The easy solution is to move that line to the end and hand it the Query class, so change the file to:

# flask_sqlalchemy/schema.py
import graphene
from graphene import relay
from graphene_sqlalchemy import SQLAlchemyObjectType, SQLAlchemyConnectionField
from models import db_session, Department as DepartmentModel, Employee as EmployeeModel

class Department(SQLAlchemyObjectType):
    class Meta:
        model = DepartmentModel
        interfaces = (relay.Node, )

class Employee(SQLAlchemyObjectType):
    class Meta:
        model = EmployeeModel
        interfaces = (relay.Node, )

class Query(graphene.ObjectType):
    node = relay.Node.Field()
    all_employees = SQLAlchemyConnectionField(Employee)

schema = graphene.Schema(query = Query)

And then the example appears to work as intended. I think this is an error in the tutorial and should be changed on the Graphene website.