denis-ryzhkov / aiographql

asyncio + graphql = fast and simple api

aiographql

asyncio + graphql = fast and simple api

• asyncio - explicit concurrency to reduce race conditions
  https://docs.python.org/3/library/asyncio.html
  https://glyph.twistedmatrix.com/2014/02/unyielding.html
• graphql - all you need and nothing more in one request +auto docs of your api
  http://graphql.org/
  http://graphene-python.org/
• uvloop, protocol - top performance
  https://magic.io/blog/uvloop-blazing-fast-python-networking/
  https://github.com/MagicStack/uvloop#performance
• minimal http - unlike REST frameworks that are waste of time for /graphql endpoint
• pluggable context - for auth, logging, etc
• exception handling - at all levels, with default or custom handler

Usage

pip install aiographql

cat <<'END' >serve.py
import asyncio, aiographql, graphene

class User(graphene.ObjectType):
    id = graphene.ID(required=True)
    name = graphene.String()

class Query(graphene.ObjectType):
    me = graphene.Field(User)

    async def resolve_me(self, info):
        await asyncio.sleep(1)  # DB
        return User(id=42, name='John')

schema = graphene.Schema(query=Query, mutation=None)

aiographql.serve(schema, listen=[
    dict(protocol='tcp', port=25100),
    dict(protocol='unix', path='/tmp/worker0'),
])
END

python3 serve.py

curl http://localhost:25100/ --data-binary \
'{"query": "{
    me {
        id
        name
    }
}", "variables": null}'

# OR:
curl --unix-socket /tmp/worker0 http:/ --data-binary ...

# Result:
# 1 second async await for DB and then:
{"data":{"me":{"id":"42","name":"John"}}}

See more examples and tests about JWT auth, concurrent slow DB queries, remote_addr, etc:
https://github.com/academicmerit/aiographql/tree/master/tests

Config

import aiographql; help(aiographql.serve)

serve(schema, listen, get_context=None, exception_handler=None, enable_uvloop=True, run=True)
    Configure the stack and start serving requests

• schema: graphene.Schema - GraphQL schema to serve
• listen: list - one or more endpoints to listen for connections:
  • dict(protocol='tcp', port=25100, ...) - https://docs.python.org/3/library/asyncio-eventloop.html#asyncio.AbstractEventLoop.create_server
  • dict(protocol='unix', path='/tmp/worker0', ...) - https://docs.python.org/3/library/asyncio-eventloop.html#asyncio.AbstractEventLoop.create_unix_server
• get_context: None or [async] callable(loop, context: dict): mixed - to produce GraphQL context like auth from input unified with exception_handler()
• exception_handler: None or callable(loop, context: dict) - default or custom exception handler as defined in
  https://docs.python.org/3/library/asyncio-eventloop.html#asyncio.AbstractEventLoop.set_exception_handler +
  • headers: bytes or None - HTTP headers, if known
  • request: dict or bytes or None - accumulated HTTP request before content length is known, then accumulated content, then GraphQL request
• enable_uvloop: bool - enable uvloop for top performance, unless you have a better loop
• run: bool - if True, run the loop; False is good for tests
• return servers: Servers - await servers.close() to close listening sockets - good for tests

TODO

• Support GET and Content-Type: application/graphql
  http://graphql.org/learn/serving-over-http/#get-request
  http://graphql.org/learn/serving-over-http/#post-request
• Support GZIP
  http://graphql.org/learn/best-practices/#json-with-gzip
  http://nginx.org/en/docs/http/ngx_http_gzip_module.html
• Support backpressure protection, etc
  https://vorpus.org/blog/some-thoughts-on-asynchronous-api-design-in-a-post-asyncawait-world/
• Meet high quality standards and join https://github.com/aio-libs

License

aiographql version 0.2.1
Created and maintained by Denis Ryzhkov <[e-mail]> and other aiographql authors
Copyright (C) 2018 by AcademicMerit LLC (dba FineTune)
MIT License, see https://opensource.org/licenses/MIT