Adds GraphQL support to your aiohttp app.
Switch branches/tags
Nothing to show
Clone or download
dfee Merge pull request #1 from Hellzed/master
Update to the latest GraphiQL
Latest commit bb6d6ef Dec 15, 2017
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
aiohttp_graphql Switch to new jsdelivr urls Dec 13, 2017
tests init Sep 23, 2017
.gitignore init Sep 23, 2017
.pylintrc init Sep 23, 2017
LICENSE.txt updated README, license Sep 23, 2017
MANIFEST.in init Sep 23, 2017
README.md updated README, license Sep 23, 2017
setup.cfg made pypi ready Sep 23, 2017
setup.py made pypi ready Sep 23, 2017

README.md

aiohttp-graphql

Adds GraphQL support to your aiohttp application.

Based on flask-graphql by Syrus Akbary and sanic-graphql by Sergey Porivaev.

Usage

Just use the GraphQLView view from aiohttp_graphql

from aiohttp_graphql import GraphQLView

GraphQLView.attach(app, schema=Schema, graphiql=True)

# Optional, for adding batch query support (used in Apollo-Client)
GraphQLView.attach(app, schema=Schema, batch=True)

This will add a /graphql endpoint to your app (customizable by passing route_path='/mypath' to GraphQLView.attach).

Note: GraphQLView.attach is just a convenience function, and the same functionality can be acheived with

gql_view = GraphQLView(schema=Schema, **kwargs)
app.router.add_route('*', gql_view, name='graphql')

It's worth noting that the the "view function" of GraphQLView is contained in GraphQLView.__call__. So, when you create an instance, that instance is callable with the request object as the sole positional argument. To illustrate:

gql_view = GraphQLView(schema=Schema, **kwargs)
gql_view(request)  # <-- the instance is callable and expects a `aiohttp.web.Request` object.

Supported options

  • schema: The GraphQLSchema object that you want the view to execute when it gets a valid request.
  • executor: The Executor that you want to use to execute queries. If an AsyncioExecutor instance is provided, performs queries asynchronously within executor’s loop.
  • root_value: The root_value you want to provide to executor.execute.
  • context: A value to pass as the context to the graphql() function. By default is set to dict with request object at key request.
  • pretty: Whether or not you want the response to be pretty printed JSON.
  • graphiql: If True, may present GraphiQL when loaded directly from a browser (a useful tool for debugging and exploration).
  • graphiql_version: The version of the provided graphiql package.
  • graphiql_template: Inject a Jinja template string to customize GraphiQL.
  • middleware: Custom middleware for graphql-python.
  • batch: Set the GraphQL view as batch (for using in Apollo-Client or [ReactRelayNetworkLayer])
  • jinja_env: Sets jinja environment to be used to process GraphiQL template. If Jinja’s async mode is enabled (by enable_async=True), uses Template.render_async instead of Template.render. If environment is not set, fallbacks to simple regex-based renderer.
  • max_age: sets the response header Access-Control-Max-Age for preflight requests
  • encoder: the encoder to use for responses (sensibly defaults to graphql_server.json_encode)
  • error_formatter: the error formatter to use for responses (sensibly defaults to graphql_server.default_format_error)
  • enable_async: whether async mode will be enabled.

Testing

Testing is done with pytest.

git clone https://github.com/dfee/aiohttp-graphql
cd aiohttp-graphql
# Create a virtualenv
python3.6 -m venv env && source env/bin/activate  # for example
pip install -e .[test]
pytest

The tests, while modeled after sanic-graphql's tests, have been entirely refactored to take advantage of pytest-asyncio, conform with PEP-8, and increase readability with pytest fixtures. For usage tests, please check them out.

License

Copyright for portions of project aiohttp-graphql are held by Syrus Akbary as part of project flask-graphql and sanic-graphql as part of project Sergey Porivaev. All other claims to this project aiohttp-graphql are held by Devin Fee.

This project is licensed under the MIT License.