Skip to content

Library to write servers for GraphQL-like query languages

License

Notifications You must be signed in to change notification settings

evo-company/hiku

Repository files navigation

Hiku

project documentation version tag license

Hiku is a library to implement Graph APIs. Essential GraphQL support included.

Installation

$ pip3 install hiku

Bug fixes and new features are frequently published via release candidates:

$ pip3 install --upgrade --pre hiku

Documentation

Read documentation

Optional dependencies

  • graphql-core - for GraphQL support
  • sqlalchemy - for SQLAlchemy support as a data-source
  • aiopg - for async PostgreSQL support with aiopg
  • asyncpg - for async PostgreSQL support with asyncpg
  • prometheus-client - for Prometheus metrics support
  • sentry-sdk - for Sentry tracing support

Highlights

  • Not coupled to a single specific query language
  • Flexibility in result serialization
  • Natively uses normalized result representation, without data duplication
  • All concurrency models supported: coroutines, threads
  • Parallel query execution
  • No data under-fetching or over-fetching between client<->server and between server<->database
  • No N+1 problems by design
  • Introduces a concept of Two-Level Graph in order to decouple data-sources and business-logic

Quick example

Graph definition:

from hiku.graph import Graph, Root, Node, Field, Link
from hiku.types import String, Sequence, TypeRef

def characters_data(fields, ids):
    data = {
        1: {'name': 'James T. Kirk', 'species': 'Human'},
        2: {'name': 'Spock', 'species': 'Vulcan/Human'},
        3: {'name': 'Leonard McCoy', 'species': 'Human'},
    }
    return [[data[i][f.name] for f in fields] for i in ids]

def characters_link():
    return [1, 2, 3]

GRAPH = Graph([
    Node('Character', [
        Field('name', String, characters_data),
        Field('species', String, characters_data),
    ]),
    Root([
        Link('characters', Sequence[TypeRef['Character']],
             characters_link, requires=None),
    ]),
])

Query:

from hiku.schema import Schema
from hiku.builder import Q, build
from hiku.executors.sync import SyncExecutor

schema = Schema(SyncExecutor(), GRAPH)

result = schema.execute_sync(build([
    Q.characters[
        Q.name,
        Q.species,
    ],
]))

# use result in your code
for character in result.data["characters"]:
    print(character["name"], '-', character["species"])

Output:

James T. Kirk - Human
Spock - Vulcan/Human
Leonard McCoy - Human

Contributing

Use Tox in order to test and lint your changes.