fenrir-framework 2.2.2__tar.gz → 2.3.3__tar.gz

{fenrir_framework-2.2.2/fenrir_framework.egg-info → fenrir_framework-2.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fenrir-framework
-Version: 2.2.2
+Version: 2.3.3
 Summary: A hybrid Python web framework combining Flask, FastAPI, Bottle, Falcon, and Sanic
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
@@ -24,11 +24,11 @@ Dynamic: license-file
 [![PyPI version](https://img.shields.io/pypi/v/fenrir-framework.svg?color=blueviolet)](https://pypi.org/project/fenrir-framework/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
 [![Python Version](https://img.shields.io/badge/Python-3.8%2B-blue.svg)](https://www.python.org/)
-[![Tests](https://img.shields.io/badge/Tests-528%20Passed-brightgreen.svg)](https://github.com/IshikawaUta/fenrir/actions)
+[![Tests](https://img.shields.io/badge/Tests-568%20Passed-brightgreen.svg)](https://github.com/IshikawaUta/fenrir/actions)
 [![CI](https://github.com/IshikawaUta/fenrir/actions/workflows/test.yml/badge.svg)](https://github.com/IshikawaUta/fenrir/actions/workflows/test.yml)
 [![Performance](https://img.shields.io/badge/Performance-High--Speed%20ASGI-orange.svg)]()
 
-**Fenrir** is a state-of-the-art, high-performance, hybrid Python web framework built on top of modern ASGI specifications. It elegantly merges the best programming paradigms from Python's most popular web frameworks (**Flask**, **FastAPI**, **Sanic**, **Falcon**, and **Bottle**) into a single unified workspace, powered locally by the premium **Asteri v2.2.2** application server.
+**Fenrir** is a state-of-the-art, high-performance, hybrid Python web framework built on top of modern ASGI specifications. It elegantly merges the best programming paradigms from Python's most popular web frameworks (**Flask**, **FastAPI**, **Sanic**, **Falcon**, and **Bottle**) into a single unified workspace, powered locally by the premium **Asteri** application server.
 
 Whether you prefer the automatic Pydantic validation of FastAPI, the seamless context-locals of Flask, the raw class-based speed of Falcon, or the robust background task model of Sanic, **Fenrir** allows you to leverage them all simultaneously in the same codebase.
 
@@ -42,6 +42,12 @@ Install directly from **PyPI**:
 pip install fenrir-framework
 ```
 
+Or install with Redis support for distributed sessions and rate limiting:
+
+```bash
+pip install fenrir-framework[redis]
+```
+
 Or install in development mode by cloning the repository:
 
 ```bash
@@ -55,6 +61,7 @@ pip install -e .
 ## 🌟 Key Features
 
 *   **⚡ High-Speed ASGI Core**: Extremely low-overhead routing and handler pipeline, achieving massive request throughput.
+*   **🔺 Trie-Based Routing**: O(k) route matching where k = path depth, instead of O(n) linear scan. Handles 1000+ routes efficiently.
 *   **🧩 Framework Hybridization**:
     *   **FastAPI Paradigm**: Native Pydantic v2 data validation, `Annotated` type decorators, automated parameter resolution (`Query`, `Path`, `Header`, `Cookie`, `Body`), dynamic dependency injection (`Depends`), and automated `response_model` serialization.
     *   **Flask Paradigm**: Thread/Task-safe context locals (`request`, `g`, `session`), Jinja2 template rendering (`render_template`), and request teardown hooks.
@@ -62,7 +69,13 @@ pip install -e .
     *   **Sanic Paradigm**: Global `sys.modules` patching (`install_sanic_compat()`), standard response helpers (`json`, `text`, `html`, `raw`, `redirect`), lifecycle listeners (`before_server_start`, etc.), and a background event scheduler (`app.add_task`).
     *   **Bottle Paradigm**: Built-in WSGI-to-ASGI wrapper and legacy mount adapter (`app.mount_wsgi()`) to run old WSGI applications at ASGI speeds.
 *   **📖 Auto-Generated OpenAPI Docs**: Interactive **Swagger UI** (`/docs`) and **ReDoc** (`/redoc`) instantly generated from your Pydantic schemas and route metadata.
-*   **🔌 Modern Communications**: Out-of-the-box support for **WebSockets** and **Server-Sent Events (SSE)**.
+*   **🔌 Modern Communications**: Out-of-the-box support for **WebSockets** (with authentication) and **Server-Sent Events (SSE)**.
+*   **🔐 WebSocket Authentication**: `WebSocketTokenAuth` dependency for token-based WebSocket authentication via headers or query parameters.
+*   **🗄️ Connection Pooling**: Built-in generic `ConnectionPool` and `DatabasePool` with health checks, retry logic, and automatic connection recycling.
+*   **🌐 HTTP/2 Push**: `HTTP2Push` utility for server push with Link headers and auto-push decorators.
+*   **⏱️ Advanced Rate Limiting**: Per-IP or per-user rate limiting with optional Redis backend for distributed deployments.
+*   **📦 Streaming Request Body**: `stream_body()` method for memory-efficient processing of large uploads without buffering.
+*   **🗜️ Optimized GZip Compression**: Default compression level 6 (optimal CPU/ratio trade-off) instead of level 9.
 *   **🛠️ Premium CLI Tooling**: Visual route tables, interactive app shell, in-memory benchmarking suite, project scaffolding, and environment system inspection.
 *   **🐍 Python 3.8–3.13 Compatible**: Full backward compatibility ensured via `typing_extensions` polyfills for `Annotated`, `get_origin`, `get_args`; and a `contextvars`-aware `asyncio.to_thread` shim.
 
@@ -85,7 +98,7 @@ logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger("demo")
 
 # Initialize the Hybrid App
-app = Fenrir(title="Fenrir Hybrid Demo", version="2.2.2")
+app = Fenrir(title="Fenrir Hybrid Demo", version="2.3.2")
 
 # --- 1. FastAPI-Style Validation & DI ---
 class UserRegister(BaseModel):
@@ -129,12 +142,158 @@ if __name__ == "__main__":
 
 ---
 
+## 🔺 Trie-Based Routing
+
+Fenrir v2.3.2 uses a trie-based routing index for O(k) route matching, where k is the path depth. This is significantly faster than linear O(n) matching when you have many routes.
+
+```python
+from fenrir import Fenrir
+
+app = Fenrir()
+
+# These routes are indexed in a trie for fast lookup
+@app.get("/api/v1/users")
+async def list_users(): ...
+
+@app.get("/api/v1/users/<int:user_id>")
+async def get_user(user_id: int): ...
+
+@app.get("/api/v1/posts/<int:post_id>/comments")
+async def get_comments(post_id: int): ...
+
+# Route matching is O(k) where k = number of path segments
+# /api/v1/users/42 → checks: api → v1 → users → 42 (parametric)
+```
+
+---
+
+## 🔐 WebSocket Authentication
+
+Authenticate WebSocket connections using tokens from headers or query parameters:
+
+```python
+from fenrir import Fenrir, WebSocket, Depends
+from fenrir.security import WebSocketTokenAuth
+
+app = Fenrir()
+auth = WebSocketTokenAuth()
+
+@app.websocket("/ws")
+async def websocket_handler(websocket: WebSocket, token: str = Depends(auth)):
+    await websocket.accept()
+    await websocket.send_text(f"Authenticated with token: {token}")
+    while True:
+        data = await websocket.receive_text()
+        await websocket.send_text(f"Echo: {data}")
+```
+
+---
+
+## 🗄️ Connection Pooling
+
+Built-in connection pooling for databases and external services:
+
+```python
+from fenrir import Fenrir
+from fenrir.pool import ConnectionPool
+
+app = Fenrir()
+
+# Create a connection pool
+pool = ConnectionPool(
+    create_func=lambda: create_engine("sqlite:///db.sqlite3"),
+    close_func=lambda engine: engine.dispose(),
+    min_size=2,
+    max_size=10,
+)
+
+@app.get("/users")
+async def list_users():
+    async with pool.acquire() as conn:
+        result = conn.execute("SELECT * FROM users")
+        return {"users": [dict(row) for row in result]}
+```
+
+---
+
+## 🌐 HTTP/2 Push
+
+Proactively push resources to clients before they request them:
+
+```python
+from fenrir import Fenrir
+from fenrir.http2 import HTTP2Push
+
+app = Fenrir()
+push = HTTP2Push()
+
+@app.get("/")
+async def index():
+    return push.push(
+        "<html><link rel='stylesheet' href='/static/style.css'></html>",
+        push_paths=["/static/style.css", "/static/app.js"],
+    )
+```
+
+---
+
+## ⏱️ Advanced Rate Limiting
+
+Per-IP or per-user rate limiting with optional Redis backend:
+
+```python
+from fenrir import Fenrir
+from fenrir.middleware import RateLimitMiddleware
+
+app = Fenrir()
+
+# Per-IP rate limiting
+app.add_middleware(RateLimitMiddleware, max_requests=100, window_seconds=60)
+
+# Per-user rate limiting
+def user_key(scope):
+    for k, v in scope.get("headers", []):
+        if k == b"x-user-id":
+            return v.decode("latin-1")
+    client = scope.get("client")
+    return client[0] if client else "unknown"
+
+app.add_middleware(RateLimitMiddleware, key_func=user_key)
+
+# Distributed rate limiting with Redis
+import redis.asyncio as aioredis
+redis_client = aioredis.Redis()
+app.add_middleware(RateLimitMiddleware, redis_client=redis_client)
+```
+
+---
+
+## 📦 Streaming Request Body
+
+Process large uploads efficiently without buffering the entire body:
+
+```python
+from fenrir import Fenrir, Request
+
+app = Fenrir()
+
+@app.post("/upload")
+async def upload(request: Request):
+    total_bytes = 0
+    async for chunk in request.stream_body(chunk_size=65536):
+        total_bytes += len(chunk)
+        # Process each chunk without loading entire body into memory
+    return {"bytes_received": total_bytes}
+```
+
+---
+
 ## 💻 CLI Command Reference
 
 Fenrir comes packed with a high-fidelity, visually rich command-line tool. Start the CLI by executing `fenrir` or `python -m fenrir.cli`.
 
 ### 1. `fenrir run`
-Serve your application locally. Powered by **Asteri v2.2.2**, supporting dynamic multiprocessing, worker management, and live hot-reloading.
+Serve your application locally. Powered by **Asteri**, supporting dynamic multiprocessing, worker management, and live hot-reloading.
 ```bash
 fenrir run demo_app:app --port 8000 --dev
 ```
@@ -180,7 +339,7 @@ fenrir info demo_app:app
 
 ## 🧪 Comprehensive Test Suite
 
-Fenrir is thoroughly covered by an automated test suite comprising **528 tests** validating every single component, compat namespace, file upload, routing detail, and CLI functionality. The suite runs automatically via **GitHub Actions** on every push across Python **3.8 – 3.13**.
+Fenrir is thoroughly covered by an automated test suite comprising **568 tests** validating every single component, including the new trie-based routing, streaming body, connection pooling, HTTP/2 push, WebSocket authentication, and rate limiting features. The suite runs automatically via **GitHub Actions** on every push across Python **3.8 – 3.13**.
 
 Run the test suite locally:
 
@@ -190,13 +349,46 @@ PYTHONPATH=. pytest -v
 
 ### Output:
 ```text
-======================= 528 passed, 1 skipped in 3.4s ========================
+=============================== 568 passed, 1 skipped in 6.72s ===============================
 ```
 
 ---
 
 ## 🔄 Changelog
 
+### v2.3.2 — Architecture & Performance Upgrade
+
+Major architecture improvements, new features, and performance optimizations:
+
+**Architecture Improvements**
+- **Trie-Based Routing**: Replaced O(n) linear route matching with O(k) trie-based routing. Route lookup now scales with path depth, not total route count.
+- **Context Vars Migration**: Removed `sys._fenrir_active_app` hack, replaced with proper `contextvars.ContextVar` for thread/async-task-safe app context.
+
+**New Components**
+- **Connection Pooling (`fenrir.pool`)**: Generic `ConnectionPool` and `DatabasePool` with health checks, retry logic, automatic connection recycling, and configurable pool sizes.
+- **HTTP/2 Push (`fenrir.http2`)**: `HTTP2Push` utility for server push with Link headers, auto-push decorators, and resource type guessing.
+- **WebSocket Authentication (`fenrir.security`)**: `WebSocketTokenAuth` dependency for token-based WebSocket authentication via headers or query parameters.
+
+**New Features**
+- **Streaming Request Body**: `request.stream_body()` method for memory-efficient processing of large uploads without buffering.
+- **Per-User Rate Limiting**: `key_func` parameter in `RateLimitMiddleware` for custom rate limiting keys (user ID, API key, etc.).
+- **Distributed Rate Limiting**: Redis backend support for `RateLimitMiddleware` using sliding window algorithm.
+
+**Performance Optimizations**
+- **GZip Compression Level**: Default `compresslevel` changed from 9 to 6 for optimal CPU/ratio trade-off.
+- **Redis Rate Limiter**: Uses `time.monotonic()` instead of `time.time()` for clock-safe operation, with unique IDs to prevent collisions.
+- **Deprecated API Fix**: Replaced deprecated `asyncio.get_event_loop()` with `asyncio.get_running_loop()` in WSGI adapter.
+
+**Bug Fixes**
+- Fixed missing `import sys` in `app.py` that silently broke root_path detection.
+- Fixed stale `sys._fenrir_active_app` references in `views.py` and `templating.py`.
+- Fixed inconsistent version strings across `pyproject.toml`, `__init__.py`, and `app.py`.
+- Fixed unused `import asyncio` in `falcon.py`.
+- Removed private `Semaphore._value` access from `Pool.stats`.
+
+**New Exports**
+- `RouteTrie`, `WebSocketTokenAuth`, `ConnectionPool`, `DatabasePool`, `HTTP2Push`
+
 ### v2.2.2 — Major Feature Update
 
 New middleware, session backends, pagination, and more:

{fenrir_framework-2.2.2 → fenrir_framework-2.3.3}/README.md

@@ -7,11 +7,11 @@
 [![PyPI version](https://img.shields.io/pypi/v/fenrir-framework.svg?color=blueviolet)](https://pypi.org/project/fenrir-framework/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
 [![Python Version](https://img.shields.io/badge/Python-3.8%2B-blue.svg)](https://www.python.org/)
-[![Tests](https://img.shields.io/badge/Tests-528%20Passed-brightgreen.svg)](https://github.com/IshikawaUta/fenrir/actions)
+[![Tests](https://img.shields.io/badge/Tests-568%20Passed-brightgreen.svg)](https://github.com/IshikawaUta/fenrir/actions)
 [![CI](https://github.com/IshikawaUta/fenrir/actions/workflows/test.yml/badge.svg)](https://github.com/IshikawaUta/fenrir/actions/workflows/test.yml)
 [![Performance](https://img.shields.io/badge/Performance-High--Speed%20ASGI-orange.svg)]()
 
-**Fenrir** is a state-of-the-art, high-performance, hybrid Python web framework built on top of modern ASGI specifications. It elegantly merges the best programming paradigms from Python's most popular web frameworks (**Flask**, **FastAPI**, **Sanic**, **Falcon**, and **Bottle**) into a single unified workspace, powered locally by the premium **Asteri v2.2.2** application server.
+**Fenrir** is a state-of-the-art, high-performance, hybrid Python web framework built on top of modern ASGI specifications. It elegantly merges the best programming paradigms from Python's most popular web frameworks (**Flask**, **FastAPI**, **Sanic**, **Falcon**, and **Bottle**) into a single unified workspace, powered locally by the premium **Asteri** application server.
 
 Whether you prefer the automatic Pydantic validation of FastAPI, the seamless context-locals of Flask, the raw class-based speed of Falcon, or the robust background task model of Sanic, **Fenrir** allows you to leverage them all simultaneously in the same codebase.
 
@@ -25,6 +25,12 @@ Install directly from **PyPI**:
 pip install fenrir-framework
 ```
 
+Or install with Redis support for distributed sessions and rate limiting:
+
+```bash
+pip install fenrir-framework[redis]
+```
+
 Or install in development mode by cloning the repository:
 
 ```bash
@@ -38,6 +44,7 @@ pip install -e .
 ## 🌟 Key Features
 
 *   **⚡ High-Speed ASGI Core**: Extremely low-overhead routing and handler pipeline, achieving massive request throughput.
+*   **🔺 Trie-Based Routing**: O(k) route matching where k = path depth, instead of O(n) linear scan. Handles 1000+ routes efficiently.
 *   **🧩 Framework Hybridization**:
     *   **FastAPI Paradigm**: Native Pydantic v2 data validation, `Annotated` type decorators, automated parameter resolution (`Query`, `Path`, `Header`, `Cookie`, `Body`), dynamic dependency injection (`Depends`), and automated `response_model` serialization.
     *   **Flask Paradigm**: Thread/Task-safe context locals (`request`, `g`, `session`), Jinja2 template rendering (`render_template`), and request teardown hooks.
@@ -45,7 +52,13 @@ pip install -e .
     *   **Sanic Paradigm**: Global `sys.modules` patching (`install_sanic_compat()`), standard response helpers (`json`, `text`, `html`, `raw`, `redirect`), lifecycle listeners (`before_server_start`, etc.), and a background event scheduler (`app.add_task`).
     *   **Bottle Paradigm**: Built-in WSGI-to-ASGI wrapper and legacy mount adapter (`app.mount_wsgi()`) to run old WSGI applications at ASGI speeds.
 *   **📖 Auto-Generated OpenAPI Docs**: Interactive **Swagger UI** (`/docs`) and **ReDoc** (`/redoc`) instantly generated from your Pydantic schemas and route metadata.
-*   **🔌 Modern Communications**: Out-of-the-box support for **WebSockets** and **Server-Sent Events (SSE)**.
+*   **🔌 Modern Communications**: Out-of-the-box support for **WebSockets** (with authentication) and **Server-Sent Events (SSE)**.
+*   **🔐 WebSocket Authentication**: `WebSocketTokenAuth` dependency for token-based WebSocket authentication via headers or query parameters.
+*   **🗄️ Connection Pooling**: Built-in generic `ConnectionPool` and `DatabasePool` with health checks, retry logic, and automatic connection recycling.
+*   **🌐 HTTP/2 Push**: `HTTP2Push` utility for server push with Link headers and auto-push decorators.
+*   **⏱️ Advanced Rate Limiting**: Per-IP or per-user rate limiting with optional Redis backend for distributed deployments.
+*   **📦 Streaming Request Body**: `stream_body()` method for memory-efficient processing of large uploads without buffering.
+*   **🗜️ Optimized GZip Compression**: Default compression level 6 (optimal CPU/ratio trade-off) instead of level 9.
 *   **🛠️ Premium CLI Tooling**: Visual route tables, interactive app shell, in-memory benchmarking suite, project scaffolding, and environment system inspection.
 *   **🐍 Python 3.8–3.13 Compatible**: Full backward compatibility ensured via `typing_extensions` polyfills for `Annotated`, `get_origin`, `get_args`; and a `contextvars`-aware `asyncio.to_thread` shim.
 
@@ -68,7 +81,7 @@ logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger("demo")
 
 # Initialize the Hybrid App
-app = Fenrir(title="Fenrir Hybrid Demo", version="2.2.2")
+app = Fenrir(title="Fenrir Hybrid Demo", version="2.3.2")
 
 # --- 1. FastAPI-Style Validation & DI ---
 class UserRegister(BaseModel):
@@ -112,12 +125,158 @@ if __name__ == "__main__":
 
 ---
 
+## 🔺 Trie-Based Routing
+
+Fenrir v2.3.2 uses a trie-based routing index for O(k) route matching, where k is the path depth. This is significantly faster than linear O(n) matching when you have many routes.
+
+```python
+from fenrir import Fenrir
+
+app = Fenrir()
+
+# These routes are indexed in a trie for fast lookup
+@app.get("/api/v1/users")
+async def list_users(): ...
+
+@app.get("/api/v1/users/<int:user_id>")
+async def get_user(user_id: int): ...
+
+@app.get("/api/v1/posts/<int:post_id>/comments")
+async def get_comments(post_id: int): ...
+
+# Route matching is O(k) where k = number of path segments
+# /api/v1/users/42 → checks: api → v1 → users → 42 (parametric)
+```
+
+---
+
+## 🔐 WebSocket Authentication
+
+Authenticate WebSocket connections using tokens from headers or query parameters:
+
+```python
+from fenrir import Fenrir, WebSocket, Depends
+from fenrir.security import WebSocketTokenAuth
+
+app = Fenrir()
+auth = WebSocketTokenAuth()
+
+@app.websocket("/ws")
+async def websocket_handler(websocket: WebSocket, token: str = Depends(auth)):
+    await websocket.accept()
+    await websocket.send_text(f"Authenticated with token: {token}")
+    while True:
+        data = await websocket.receive_text()
+        await websocket.send_text(f"Echo: {data}")
+```
+
+---
+
+## 🗄️ Connection Pooling
+
+Built-in connection pooling for databases and external services:
+
+```python
+from fenrir import Fenrir
+from fenrir.pool import ConnectionPool
+
+app = Fenrir()
+
+# Create a connection pool
+pool = ConnectionPool(
+    create_func=lambda: create_engine("sqlite:///db.sqlite3"),
+    close_func=lambda engine: engine.dispose(),
+    min_size=2,
+    max_size=10,
+)
+
+@app.get("/users")
+async def list_users():
+    async with pool.acquire() as conn:
+        result = conn.execute("SELECT * FROM users")
+        return {"users": [dict(row) for row in result]}
+```
+
+---
+
+## 🌐 HTTP/2 Push
+
+Proactively push resources to clients before they request them:
+
+```python
+from fenrir import Fenrir
+from fenrir.http2 import HTTP2Push
+
+app = Fenrir()
+push = HTTP2Push()
+
+@app.get("/")
+async def index():
+    return push.push(
+        "<html><link rel='stylesheet' href='/static/style.css'></html>",
+        push_paths=["/static/style.css", "/static/app.js"],
+    )
+```
+
+---
+
+## ⏱️ Advanced Rate Limiting
+
+Per-IP or per-user rate limiting with optional Redis backend:
+
+```python
+from fenrir import Fenrir
+from fenrir.middleware import RateLimitMiddleware
+
+app = Fenrir()
+
+# Per-IP rate limiting
+app.add_middleware(RateLimitMiddleware, max_requests=100, window_seconds=60)
+
+# Per-user rate limiting
+def user_key(scope):
+    for k, v in scope.get("headers", []):
+        if k == b"x-user-id":
+            return v.decode("latin-1")
+    client = scope.get("client")
+    return client[0] if client else "unknown"
+
+app.add_middleware(RateLimitMiddleware, key_func=user_key)
+
+# Distributed rate limiting with Redis
+import redis.asyncio as aioredis
+redis_client = aioredis.Redis()
+app.add_middleware(RateLimitMiddleware, redis_client=redis_client)
+```
+
+---
+
+## 📦 Streaming Request Body
+
+Process large uploads efficiently without buffering the entire body:
+
+```python
+from fenrir import Fenrir, Request
+
+app = Fenrir()
+
+@app.post("/upload")
+async def upload(request: Request):
+    total_bytes = 0
+    async for chunk in request.stream_body(chunk_size=65536):
+        total_bytes += len(chunk)
+        # Process each chunk without loading entire body into memory
+    return {"bytes_received": total_bytes}
+```
+
+---
+
 ## 💻 CLI Command Reference
 
 Fenrir comes packed with a high-fidelity, visually rich command-line tool. Start the CLI by executing `fenrir` or `python -m fenrir.cli`.
 
 ### 1. `fenrir run`
-Serve your application locally. Powered by **Asteri v2.2.2**, supporting dynamic multiprocessing, worker management, and live hot-reloading.
+Serve your application locally. Powered by **Asteri**, supporting dynamic multiprocessing, worker management, and live hot-reloading.
 ```bash
 fenrir run demo_app:app --port 8000 --dev
 ```
@@ -163,7 +322,7 @@ fenrir info demo_app:app
 
 ## 🧪 Comprehensive Test Suite
 
-Fenrir is thoroughly covered by an automated test suite comprising **528 tests** validating every single component, compat namespace, file upload, routing detail, and CLI functionality. The suite runs automatically via **GitHub Actions** on every push across Python **3.8 – 3.13**.
+Fenrir is thoroughly covered by an automated test suite comprising **568 tests** validating every single component, including the new trie-based routing, streaming body, connection pooling, HTTP/2 push, WebSocket authentication, and rate limiting features. The suite runs automatically via **GitHub Actions** on every push across Python **3.8 – 3.13**.
 
 Run the test suite locally:
 
@@ -173,13 +332,46 @@ PYTHONPATH=. pytest -v
 
 ### Output:
 ```text
-======================= 528 passed, 1 skipped in 3.4s ========================
+=============================== 568 passed, 1 skipped in 6.72s ===============================
 ```
 
 ---
 
 ## 🔄 Changelog
 
+### v2.3.2 — Architecture & Performance Upgrade
+
+Major architecture improvements, new features, and performance optimizations:
+
+**Architecture Improvements**
+- **Trie-Based Routing**: Replaced O(n) linear route matching with O(k) trie-based routing. Route lookup now scales with path depth, not total route count.
+- **Context Vars Migration**: Removed `sys._fenrir_active_app` hack, replaced with proper `contextvars.ContextVar` for thread/async-task-safe app context.
+
+**New Components**
+- **Connection Pooling (`fenrir.pool`)**: Generic `ConnectionPool` and `DatabasePool` with health checks, retry logic, automatic connection recycling, and configurable pool sizes.
+- **HTTP/2 Push (`fenrir.http2`)**: `HTTP2Push` utility for server push with Link headers, auto-push decorators, and resource type guessing.
+- **WebSocket Authentication (`fenrir.security`)**: `WebSocketTokenAuth` dependency for token-based WebSocket authentication via headers or query parameters.
+
+**New Features**
+- **Streaming Request Body**: `request.stream_body()` method for memory-efficient processing of large uploads without buffering.
+- **Per-User Rate Limiting**: `key_func` parameter in `RateLimitMiddleware` for custom rate limiting keys (user ID, API key, etc.).
+- **Distributed Rate Limiting**: Redis backend support for `RateLimitMiddleware` using sliding window algorithm.
+
+**Performance Optimizations**
+- **GZip Compression Level**: Default `compresslevel` changed from 9 to 6 for optimal CPU/ratio trade-off.
+- **Redis Rate Limiter**: Uses `time.monotonic()` instead of `time.time()` for clock-safe operation, with unique IDs to prevent collisions.
+- **Deprecated API Fix**: Replaced deprecated `asyncio.get_event_loop()` with `asyncio.get_running_loop()` in WSGI adapter.
+
+**Bug Fixes**
+- Fixed missing `import sys` in `app.py` that silently broke root_path detection.
+- Fixed stale `sys._fenrir_active_app` references in `views.py` and `templating.py`.
+- Fixed inconsistent version strings across `pyproject.toml`, `__init__.py`, and `app.py`.
+- Fixed unused `import asyncio` in `falcon.py`.
+- Removed private `Semaphore._value` access from `Pool.stats`.
+
+**New Exports**
+- `RouteTrie`, `WebSocketTokenAuth`, `ConnectionPool`, `DatabasePool`, `HTTP2Push`
+
 ### v2.2.2 — Major Feature Update
 
 New middleware, session backends, pagination, and more:

{fenrir_framework-2.2.2 → fenrir_framework-2.3.3}/fenrir/__init__.py

@@ -27,7 +27,7 @@ from fenrir.response import (
 from fenrir.templating import render_template, BaseTemplateRenderer, Jinja2Renderer
 from fenrir.views import View, MethodView
 from fenrir.helpers import url_for, send_file, send_from_directory, redirect
-from fenrir.routing import Router, Route, APIRouter
+from fenrir.routing import Router, Route, APIRouter, RouteTrie
 from fenrir.sse import EventSourceResponse
 from fenrir.security import (
     APIKeyCookie,
@@ -39,6 +39,7 @@ from fenrir.security import (
     OAuth2PasswordBearer,
     OAuth2AuthorizationCodeBearer,
     OpenIDConnect,
+    WebSocketTokenAuth,
 )
 from fenrir.background import BackgroundTasks, BackgroundTask
 from fenrir.compat import WsgiToAsgi, install_bottle_compat, install_falcon_compat, install_sanic_compat
@@ -53,6 +54,8 @@ from fenrir.middleware import (
     RequestIDMiddleware,
     RateLimitMiddleware,
 )
+from fenrir.pool import ConnectionPool, DatabasePool
+from fenrir.http2 import HTTP2Push
 from fenrir.pagination import PaginationParams, paginate, paginate_dict
 from fenrir.sessions import (
     RedisSessionInterface,
@@ -65,7 +68,7 @@ from fenrir.sessions import (
 # Re-export Annotated for convenient use with param markers
 from fenrir.compat import Annotated
 
-__version__ = "2.2.2"
+__version__ = "2.3.3"
 __all__ = [
     # Core app
     "Fenrir",
@@ -126,6 +129,7 @@ __all__ = [
     "Router",
     "Route",
     "APIRouter",
+    "RouteTrie",
     # SSE
     "EventSourceResponse",
     # Security
@@ -138,6 +142,7 @@ __all__ = [
     "OAuth2PasswordBearer",
     "OAuth2AuthorizationCodeBearer",
     "OpenIDConnect",
+    "WebSocketTokenAuth",
     # Background tasks
     "BackgroundTasks",
     "BackgroundTask",
@@ -165,6 +170,11 @@ __all__ = [
     "PaginationParams",
     "paginate",
     "paginate_dict",
+    # Connection Pooling
+    "ConnectionPool",
+    "DatabasePool",
+    # HTTP/2 Push
+    "HTTP2Push",
     # Server-side sessions
     "RedisSessionInterface",
     "InMemorySessionInterface",