GRAPHQL CONDITIONAL FRAGMENT HOW TO
GRAPHQL CONDITIONAL FRAGMENT CODE

This GraphQL schema is generated at request time and is always up-to-date with the current status of the space. Each Contentful space comes with a GraphQL schema based on its content model. In the next steps, we will implement some specific behavior in this directive. The GraphQL Content API provides a GraphQL API interface to the content from Contentful.

GRAPHQL CONDITIONAL FRAGMENT CODE

Info : 'ResolveInfo', ) - > Any : # Add your code here # return await next_directive (introspected_element, ctx, info ) Info : 'ResolveInfo', ) - > Any : # Add your code here # return await next_directive (value, ctx, info ) async def on_introspection ( Info : 'ResolveInfo', ) - > Any : # Add your code here # return await next_resolver (parent, args, ctx, info ) async def on_pre_output_coercion ( Parent_node, argument_definition_node, argument_node, value, ctxĬtx : Optional, ) - > Any : # Add your code here # return await next_directive (parent_node, value, ctx ) async def on_field_execution ( I am trying to run a query returning a union type (e.g. Parent_node : Union ,Īrgument_definition_node : 'InputValueDefinitionNode' ,Īrgument_node : Optional ,Ĭtx : Optional, ) - > Any : # Add your code here # return await next_directive ( In a raw GraphQL call this would be achieved using conditional fragments like so: query a.

GRAPHQL CONDITIONAL FRAGMENT HOW TO

How to declare a new directive? from typing import Any, Callable, Dict, Optional, Unionįrom tartiflette import ( 'rateLimiting' ) class RateLimiting : async def on_argument_execution ( dynamic introspection based on access rights) To ask for a field on a specific object type, you need to use an inline fragment: Learn more about this in the inline fragments section in the query guide.

on_introspection: during an introspection query, allows you to wrap the schema fields to add metadata to fields or even to remove objects (e.g.

on_pre_output_coercion: allows you to mutate the result of the resolved value of a field before returning it.

resolve the field remotly, apply a specific rate limit on a field.)

on_field_execution: allows you to wrap the field execution.

on_post_input_coercion: allows you to hook the execution flow right after an input value/variable has beed coerced (appears in version 0.10.0).

access rights, validate the input format.)

on_argument_execution: allows you to wrap the argument execution.

A directive will allow you to execute some logic at any of the following stages: Tartiflette gets most of its extensibility by directives. For instance, in the following SDL, we will define a Recipe type with a deprecated title field: directive ( reason : String = 'No longer supported' ) on FIELD_DEFINITION | ENUM_VALUE type Recipe How to implement?

allows you to hide elements of your SDL from the introspection queryĪs explain above, directive looks like a decorator in Python and are used the same way.

allows for conditional inclusion of fields, fragment spreads, and inline fragments during execution as described by the if argument.

allows for conditional exclusion of fields, fragment spreads, and inline fragments during execution as described by the if argument.

allows you to mark a field or an enum value as deprecated (this introspection will flag them as deprecated).

Tartiflette provides a bunch of built-ins directives: The possible applications are numerous, access permissions, limiting the introspection, rate limiting, auto-generating resolver function etc. It looks like a decorator in our favorite language, Python. So it's this damned if I do, damned if I don't quandary and my lack of experience with graphql and apollo really shines here.A directive is a way of describing a behavior you can apply to different objects in your SDL. If I omit the files from the fragment, the component works for all users.but then the uploader can't get their draft files.

The query has a fragment for the 'drafts' data, including the conditionally unwanted files.

While non-edit privilege users don't see the draft, the app still queries its entire fileset, which makes fetching the dataset page really slow. The draft files should only be queried if the uploader is logged in. The uploader has a 'draft' version that they can edit in the UI, then they publish snapshots of the draft, which all users can see. I'm working on an app that allows users to upload large datasets.

Graphql conditional fragment