Edit

Schema registry

By default, all the resolvers created using the @Resolver decorator, are registered to the "default" schema in the Schema Registry.

What is the aim of the Schema Registry?

The Schema Registry is an advanced use-case of Tartiflette, used by the developers who want to expose multiple Engines from the same codebase.

The Schema Registry will allow you to assign a Resolver to a specific Schema, which you can then choose during the initialization process of the Engine.

By default, every Resolver is assigned to the default schema. Moreover, every Engine is attached to the default schema.

How to use multiple Schemas and Engines from the same codebase?

The following code sample will create 2 schemas in the Schema Registry.

• "default"
• "proof_of_concept"

import asyncio

from tartiflette import Resolver, create_engine

@Resolver("Query.hello") # Will be assigned to the 'default' Schema
async def resolver_hello(parent, args, ctx, info):
    return "hello " + args["name"]


@Resolver("Query.hello", "proof_of_concept") # Will be assigned to the 'proof_of_concept' Schema
async def resolver_hello(parent, args, ctx, info):
    return "Hey " + args["name"]


async def run():
    tftt_engine = await create_engine(
        """
        type Query {
            hello(name: String): String
        }
        """
    ) # The engine created will be attached to the SDL of the 'default' Schema.

    result = await tftt_engine.execute(
        query='query { hello(name: "Chuck") }'
    )

    # the result will be
    # {
    #     "data": {
    #         "hello": "Hello Chuck"
    #     }
    # }

    tftt_proof_of_concept = await create_engine(
        """
        type Query {
            hello(name: String): String
        }
        """,
        schema_name="proof_of_concept"
    ) # This Engine will be attached to the SDL of the 'proof_of_concept' Schema.

    result_poc = await tftt_proof_of_concept.execute(
        query='query { hello(name: "Chuck") }'
    )

    # the `result_poc` will be
    # {
    #     "data": {
    #         "hello": "Hey Chuck"
    #     }
    # }