weltenfw 0.1.0

WeltenHub Client Framework — typed REST client and Pydantic schemas for the WeltenHub Story Universe API

weltenfw

WeltenHub Client Framework — typed REST client and Pydantic v2 schemas for the WeltenHub Story Universe API.

Installation

pip install weltenfw                  # core (httpx + pydantic)
pip install weltenfw[django]          # + Django AppConfig + DjangoCache
pip install weltenfw[authoring]       # + authoringfw adapter

Quick Start

from weltenfw import WeltenClient
from weltenfw.schema.world import WorldCreateInput

with WeltenClient(
    base_url="https://weltenforger.com/api/v1",
    token="your-token",
) as client:
    # Create a world
    world = client.worlds.create(WorldCreateInput(
        name="Aldoria",
        setting_era="medieval",
        description="A kingdom of runes and magic.",
    ))
    print(world.id, world.name)

    # List with pagination
    for w in client.worlds.iter_all():
        print(w.name)

    # Lookup cache
    genres = client.lookups.genres()   # cached after first call

Async

async with WeltenClient(base_url=..., token=...) as client:
    world = await client.worlds.aget(world_id)
    worlds = await client.worlds.alist()

Django Integration

# settings.py
INSTALLED_APPS = [..., "weltenfw.django"]
WELTENHUB_URL   = "https://weltenforger.com/api/v1"
WELTENHUB_TOKEN = env("WELTENHUB_TOKEN")

# views.py / services.py
from weltenfw.django import get_client
client = get_client()   # lazy singleton, per worker

Tenant Provisioning (S2S)

from weltenfw.schema.tenant import ProvisionRequest

result = client.tenants.provision(ProvisionRequest(
    username="hugo",
    email="[e-mail]",
    display_name="Hugo",
))
# ProvisionResponse(token="...", tenant_id=UUID("..."), created=True)

Architecture

• 1 Client = 1 Token = 1 Tenant — Multi-tenant consumers instantiate one client per tenant token.
• Pydantic v2 — all responses validated into typed schemas (frozen=True).
• Separate Input schemas — *CreateInput for POST, *UpdateInput for PATCH.
• Lookup cache — pluggable CacheBackend (NullCache default, MemoryCache, DjangoCache for Redis).
• Trailing-slash normalization — prevents silent POST→GET method loss on Django's 301 redirect.

Links

• WeltenHub API Docs
• ADR-032 Architecture Decision
• Changelog