fastapi-request-pipeline 0.1.3

A composable request processing pipeline for FastAPI

fastapi-request-pipeline

A composable, type-safe request processing pipeline for FastAPI. Build clean, maintainable APIs with reusable components for authentication, permissions, rate limiting, and more.

Features

• Composable Architecture - Build complex request flows from simple, reusable FlowComponent instances
• Type-Safe - Full type hints with strict mypy checking
• Built-in Components - Authentication (JWT, API Key, Cookie), permissions, rate limiting, pagination, filters
• Flow Composition - Layer and merge flows at app, router, and route levels
• OpenAPI Integration - Automatic OpenAPI schema enrichment with security requirements
• Debug Mode - Detailed execution traces for development
• Extensible - Easy to create custom FlowComponent subclasses
• Production Ready - Comprehensive test suite and production usage

Quick Start

pip install fastapi-request-pipeline

from fastapi import FastAPI, Depends
from fastapi_request_pipeline import (
    Flow,
    JWTAuthentication,
    HasRole,
    RateLimit,
    RequestContext,
    flow_dependency,
    enrich_openapi,
)

app = FastAPI()

# Define your flow
async def decode_jwt(token: str) -> dict:
    # Your JWT decoding logic
    return {"sub": "user123", "roles": ["admin"]}

admin_flow = Flow(
    JWTAuthentication(decode=decode_jwt),
    HasRole("admin"),
    RateLimit(rate=100, window_seconds=60),
)

# Use as dependency
@app.get("/admin/dashboard")
async def admin_dashboard(
    ctx: RequestContext = Depends(flow_dependency(admin_flow))
):
    return {"user": ctx.user, "message": "Welcome, admin!"}

# Enrich OpenAPI schema
enrich_openapi(app)

Core Concepts

Flow

A Flow is an ordered container of FlowComponent instances that process requests sequentially. FlowComponent instances are automatically sorted by ComponentCategory to ensure proper execution order (authentication → permissions → feature flags → throttling → filters → pagination → custom).

flow = Flow(
    JWTAuthentication(decode=decode_jwt),
    HasPermission("posts:write"),
    RateLimit(rate=100, window_seconds=60),
)

Built-in FlowComponent Classes

Built-in FlowComponent classes:

Authentication

• JWTAuthentication - Bearer token authentication
• APIKeyAuthentication - API key in headers
• CookieAuthentication - Session cookies
• AllowAnonymous - Skip authentication

Permissions

• Authenticated - Require authenticated user
• HasRole - Role-based access control
• HasPermission - Permission-based access control

Rate Limiting

• RateLimit - Configurable rate limiting with pluggable backends
• InMemoryThrottleBackend - Default in-memory backend
• Custom backends (e.g., Redis) via ThrottleBackend protocol

Filters & Pagination

• QueryFilter - Extract selected query parameters into RequestContext.state
• LimitOffset - Offset-based pagination

Flow Composition

Compose flows at different levels with merge_flows():

from fastapi_request_pipeline import merge_flows, OverrideFlow, DisableFlow

# Application-wide defaults
app_flow = Flow(
    JWTAuthentication(decode=decode_jwt),
    RateLimit(rate=1000, window_seconds=3600)
)

# Router-specific additions
admin_flow = Flow(HasRole("admin"))

# Route-specific overrides
public_flow = Flow(OverrideFlow(AllowAnonymous()))

# Merge with last-writer-wins per category
final_flow = merge_flows(app_flow, admin_flow, public_flow)

Custom FlowComponent Subclasses

Create custom FlowComponent subclasses by extending the base class:

from fastapi_request_pipeline import FlowComponent, ComponentCategory

class AuditLog(FlowComponent):
    category = ComponentCategory.CUSTOM

    async def resolve(self, ctx: RequestContext) -> None:
        await log_request(ctx.user, ctx.request.url.path)

Documentation

• User Guide - Comprehensive guide with examples
• API Reference - Complete API documentation
• CI/CD and Release Guide - Versioning rules, GitHub Releases, TestPyPI/PyPI publish pipeline
• Examples - Working code examples
  • Basic Usage
  • Authentication
  • Permissions
  • Rate Limiting
  • Flow Composition
  • Custom FlowComponent Subclasses
  • Real-World App

Requirements

• Python 3.11+
• FastAPI 0.100+

Development

# Install with dev dependencies
uv sync --extra dev

# Run tests
uv run pytest

# Run tests with coverage
uv run pytest --cov

# Lint
uv run ruff check .

# Format
uv run ruff format .

# Type check
uv run mypy --strict src/

Why Use This Library?

Instead of this:

@app.get("/posts")
async def get_posts(request: Request):
    # Authentication
    token = request.headers.get("Authorization")
    if not token or not token.startswith("Bearer "):
        raise HTTPException(401)
    user = decode_jwt(token[7:])

    # Permission check
    if "posts:read" not in user.permissions:
        raise HTTPException(403)

    # Rate limiting
    if not check_rate_limit(user.id):
        raise HTTPException(429)

    # Pagination
    limit = int(request.query_params.get("limit", 20))
    offset = int(request.query_params.get("offset", 0))

    # Business logic
    return get_posts_from_db(limit, offset)

Write this:

posts_flow = Flow(
    JWTAuthentication(decode=decode_jwt),
    HasPermission("posts:read"),
    RateLimit(rate=100, window_seconds=60),
    LimitOffset(default_limit=20),
)

@app.get("/posts")
async def get_posts(ctx: RequestContext = Depends(flow_dependency(posts_flow))):
    pagination = ctx.state["pagination"]
    return get_posts_from_db(pagination["limit"], pagination["offset"])

Benefits:

• Separation of concerns - Security logic separated from business logic
• Reusability - Define flows once, use across multiple endpoints
• Composability - Mix and match FlowComponent instances, override at any level
• Maintainability - Changes to auth/permissions in one place
• Type safety - Full type hints and IDE support
• Testability - FlowComponent instances are easy to unit test
• Documentation - OpenAPI schema automatically reflects security requirements

License

MIT