Introduction

MicroPie is a fast, lightweight, modern Python web framework built on ASGI for asynchronous web applications. Designed for flexibility and simplicity, it enables high-concurrency web apps with built-in WebSockets, session management, middleware, and optional template rendering. Extensible for integration with ASGI-compatible tools like python-socketio and ServeStatic, it’s inspired by CherryPy and licensed under the BSD 3-Clause License.

Key Features

• 🔄 Routing: Automatic mapping of URLs to functions with support for dynamic and query parameters.
• 🔒 Sessions: Simple, plugable, session management using cookies.
• 🎨 Templates: Jinja2, if installed, for rendering dynamic HTML pages.
• ⚙️ Middleware: Support for custom request middleware enabling functions like rate limiting, authentication, logging, and more.
• 🌐 Real-Time Communication: Built-in WebSocket support for real-time, bidirectional communication.
• ✨ ASGI-Powered: Built w/ asynchronous support for modern web servers like Uvicorn, Hypercorn, and Daphne, enabling high concurrency.
• 🛠️ Lightweight Design: Only optional dependencies for flexibility and faster development/deployment.
• ⚡ Blazing Fast: Check out how MicroPie compares to other popular ASGI frameworks below!

Useful Links

• Homepage: patx.github.io/micropie
• API Reference: README.md#api-documentation
• PyPI Page: pypi.org/project/MicroPie
• GitHub Project: github.com/patx/micropie
• File Issue/Request: github.com/patx/micropie/issues
• Example Applications: github.com/patx/micropie/tree/main/examples
• Introduction Lightning Talk: Introduction to MicroPie on YouTube

Latest Release Notes

View the latest release notes here. It is useful to check release notes each time a new version of MicroPie is published. Any breaking changes (rare, but do happen) also appear here.

Installing MicroPie

Installation

Install MicroPie with standard optional dependencies via pip:

pip install micropie[standard]

This will install MicroPie along with jinja2 for template rendering, and multipart for parsing multipart form data.

If you would like to install all optional dependencies (everything from standard plus orjson and uvicorn) you can run:

pip install micropie[all]

Minimal Setup

You can also install MicroPie without ANY dependencies via pip:

pip install micropie

For an ultra-minimalistic approach, download the standalone script (development version):

micropie.py

Place it in your project directory, and you are good to go. Note that jinja2 must be installed separately to use the _render_template method and/or multipart for handling file data (the _parse_multipart method), but this is optional and you can use MicroPie without them. To install the optional dependencies use:

pip install jinja2 multipart

By default MicroPie will use the json library from Python's standard library. If you need faster performance you can use orjson. MicroPie will use orjson if installed by default. If it is not installed, MicroPie will fallback to json. This means with or without orjson installed MicroPie will still handle JSON requests/responses the same. To install orjson and take advantage of it's performance, use:

pip install orjson

Install an ASGI Web Server

In order to test and deploy your apps you will need a ASGI web server like Uvicorn, Hypercorn or Daphne.

If you installed micropie[all] Uvicorn should be ready to use. If you didn't install all of MicroPie's optional dependencies, use:

pip install uvicorn

Getting Started

Create Your First ASGI App

Save the following as app.py:

from micropie import App

class MyApp(App):
    async def index(self):
        return "Welcome to MicroPie ASGI."

app = MyApp()

Run the server with:

uvicorn app:app

Access your app at http://127.0.0.1:8000.

Core Features

Route Handlers

MicroPie's route handlers map URLs to methods in your App subclass, handling HTTP requests with flexible parameter mapping and response formats.

Key Points

• Automatic Mapping: URLs map to method names (e.g., /greet → greet, / → index).
• Private Methods: Methods starting with _ (e.g., _private_method) are private and inaccessible via URLs, returning 404. Security Note: Use _ for sensitive methods to prevent external access.
• Parameters: Automatically populated from:
  • Path segments (e.g., /greet/Alice → name="Alice").
  • Query strings (e.g., ?name=Alice).
  • Form data (POST/PUT/PATCH).
  • Session data (self.request.session).
  • File uploads (self.request.files).
  • Default values in method signatures.
• HTTP Methods: Handlers support all methods (GET, POST, etc.). Check self.request.method to handle specific methods.
• Responses:
  • String, bytes, or JSON-serializable object.
  • Tuple: (status_code, body) or (status_code, body, headers).
  • Sync/async generator for streaming.

Advanced Usage

• Custom Routing: Use middleware for explicit routing (see examples/middleware and examples/explicit_routing).
• Errors: Auto-handled 404/400; customize via middleware.
• Dynamic Params: Use *args for multiple path parameters.

Flexible HTTP Routing for GET Requests

MicroPie automatically maps URLs to methods within your App class. Routes can be defined as either synchronous or asynchronous functions, offering good flexibility.

For GET requests, pass data through query strings or URL path segments, automatically mapped to method arguments.

class MyApp(App):
    async def greet(self, name="Guest"):
        return f"Hello, {name}!"

    async def hello(self):
        name = self.request.query_params.get("name", [None])[0]
        return f"Hello {name}!"

Access:

• http://127.0.0.1:8000/greet?name=Alice returns Hello, Alice!, same as http://127.0.0.1:8000/greet/Alice returns Hello, Alice!.
• http://127.0.0.1:8000/hello/Alice returns a 500 Internal Server Error because it is expecting http://127.0.0.1:8000/hello?name=Alice, which returns Hello Alice!

Flexible HTTP POST Request Handling

MicroPie also supports handling form data submitted via HTTP POST requests. Form data is automatically mapped to method arguments. It is able to handle default values and raw/JSON POST data:

class MyApp(App):
    async def submit_default_values(self, username="Anonymous"):
        return f"Form submitted by: {username}"

    async def submit_catch_all(self):
        username = self.request.body_params.get("username", ["Anonymous"])[0]
        return f"Submitted by: {username}"

By default, MicroPie's route handlers can accept any request method, it's up to you how to handle any incoming requests! You can check the request method (and an number of other things specific to the current request state) in the handler withself.request.method. You can see how to handle POST JSON data at examples/api and examples/json.

Real-Time Communication with WebSockets and Socket.IO

MicroPie includes built-in support for WebSocket connections. WebSocket routes are defined in your App subclass using methods prefixed with ws_, mirroring the simplicity of MicroPie's HTTP routing. For example, a method named ws_chat handles WebSocket connections at ws://<host>/chat.

MicroPie’s WebSocket support allows you to:

• Define WebSocket handlers with the same intuitive automatic routing as HTTP (e.g., /chat maps to ws_chat method).
• Access query parameters, path parameters, and session data in WebSocket handlers, consistent with HTTP requests.
• Manage WebSocket connections using the WebSocket class, which provides methods like accept, receive_text, send_text, and close.

Check out a basic example:

from micropie import App

class Root(App):

    async def ws_echo(self, ws):
        await ws.accept()
        while True:
            msg = await ws.receive_text()
            await ws.send_text(f"Echo: {msg}")

app = Root()

Use Socket.IO for Advanced Real-Time Features

If you want more advanced real-time features like automatic reconnection, broadcasting, or fallbacks (e.g., polling), you can integrate Socket.IO with your MicroPie app using Uvicorn as the server. See examples/socketio for integration instructions and examples.

Jinja2 Template Rendering

Dynamic HTML generation is supported via Jinja2. This happens asynchronously using Pythons asyncio library, so make sure to use the async and await with this method.

app.py

class MyApp(App):
    async def index(self):
        return await self._render_template("index.html", title="Welcome", message="Hello from MicroPie!")

templates/index.html

<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <title>{{ title }}</title>
</head>
<body>
    <h1>{{ message }}</h1>
</body>
</html>

Static File Serving

Here again, like Websockets, MicroPie does not have a built in static file method. While MicroPie does not natively support static files, if you need them, you can easily implement it in your application code or integrate dedicated libraries like ServeStatic or Starlette’s StaticFiles alongside Uvicorn to handle async static file serving. Check out examples/static_content to see this in action.

Streaming Responses

Support for streaming responses makes it easy to send data in chunks.

class MyApp(App):
    async def stream(self):
        async def generator():
            for i in range(1, 6):
                yield f"Chunk {i}\n"
        return generator()

Sessions and Cookies

Built-in session handling simplifies state management:

class MyApp(App):
    async def index(self):
        if "visits" not in self.request.session:
            self.request.session["visits"] = 1
        else:
            self.request.session["visits"] += 1
        return f"You have visited {self.request.session['visits']} times."

You also can use the SessionBackend class to create your own session backend. You can see an example of this in examples/sessions.

Middleware

MicroPie allows you to create plug-able middleware to hook into the request life cycle. Take a look a trivial example using HttpMiddleware to send the console messages before and after the request is processed. Check out examples/middleware to see more.

from micropie import App, HttpMiddleware

class MiddlewareExample(HttpMiddleware):
    async def before_request(self, request):
        print("Hook before request")

    async def after_request(self, request, status_code, response_body, extra_headers):
        print("Hook after request")


class Root(App):
    async def index(self):
        print("Hello, World!")
        return "Hello, World!"

app = Root()
app.middlewares.append(MiddlewareExample())

Middleware provides an easy and reusable way to extend the MicroPie framework. We can do things such as rate limiting, checking for max upload size in multipart requests, explicit routing, CSRF protection, and more.

MicroPie apps can be deployed using any ASGI server. For example, using Uvicorn if our application is saved as app.py and our App subclass is assigned to the app variable we can run it with:

uvicorn app:app --workers 4 --port 8000

Learn by Examples

The best way to get an idea of how MicroPie works is to see it in action! Check out the examples folder for more advanced usage, including:

• Template rendering
• Custom HTTP request handling
• File uploads
• Serving static content
• Session usage
• JSON Requests and Responses
• Socket.io Integration
• Async Streaming
• Middleware including, explicit routing and more
• Form handling and POST requests
• Websockets
• And more

Comparisons

Features vs Other Popular Frameworks

Feature	MicroPie	Flask	CherryPy	Bottle	Django	FastAPI
Routing	Automatic	Manual	Automatic	Manual	Views	Manual
Template Engine	Jinja2 (Opt.)	Jinja2	Plugin	SimpleTpl	Django	Jinja2
Middleware	Yes	Yes	Yes	Yes	Yes	Yes
Session Handling	Plugin	Plugin	Built-in	Plugin	Built-in	Plugin
Async Support	Yes	No	No	No	Yes	Yes
Built-in Server	No	No	Yes	Yes	Yes	No

Benchmark Results

The table below summarizes the performance of various ASGI frameworks based on a 15-second wrk test with 4 threads and 64 connections, measuring a simple "hello world" JSON response. Learn More.

Framework	Total Requests	Req/Sec	Transfer/Sec (MB/s)	Avg Latency (ms)	Stdev Latency (ms)	Max Latency (ms)
Blacksheep	831,432	55,060.05	7.98	1.15	0.39	15.11
MicroPie	791,721	52,685.82	8.09	1.35	1.09	21.59
Starlette	779,092	51,930.45	7.03	1.22	0.39	17.42
Litestar	610,059	40,401.18	5.47	1.57	0.63	33.66
FastAPI	281,493	18,756.73	2.54	3.52	1.82	56.73

Suggestions or Feedback?

We welcome suggestions, bug reports, and pull requests!

• File issues or feature requests here.
• Security issues that should not be public, email harrisonerd [at] gmail.com.

API Documentation

Session Backend Abstraction

MicroPie provides an abstraction for session backends, allowing you to define custom session storage mechanisms.

SessionBackend Class

Methods

• load(session_id: str) -> Dict[str, Any]

  • Abstract method to load session data given a session ID.

• save(session_id: str, data: Dict[str, Any], timeout: int) -> None

  • Abstract method to save session data.

InMemorySessionBackend Class

An in-memory implementation of the SessionBackend.

Methods

• __init__()

  • Initializes the in-memory session backend.

• load(session_id: str) -> Dict[str, Any]

  • Loads session data for the given session ID.

• save(session_id: str, data: Dict[str, Any], timeout: int) -> None

  • Saves session data for the given session ID.

Middleware Abstraction

MicroPie allows you to create pluggable middleware to hook into the request lifecycle for both HTTP and WebSocket requests.

HttpMiddleware Class

Methods

• before_request(request: Request) -> Optional[Dict]

  • Abstract method called before the HTTP request is processed. Returns an optional dictionary with response details (status_code, body, headers) to short-circuit the request, or None to continue processing.

• after_request(request: Request, status_code: int, response_body: Any, extra_headers: List[Tuple[str, str]]) -> Optional[Dict]

  • Abstract method called after the HTTP request is processed but before the final response is sent. Returns an optional dictionary with updated response details (status_code, body, headers), or None to use defaults.

WebSocketMiddleware Class

Methods

• before_websocket(request: WebSocketRequest) -> Optional[Dict]

  • Abstract method called before the WebSocket handler is invoked. Returns an optional dictionary with close details (code, reason) to reject the connection, or None to continue processing.

• after_websocket(request: WebSocketRequest) -> None

  • Abstract method called after the WebSocket handler completes.

Request Objects

Request Class

Represents an HTTP request in the MicroPie framework.

Attributes

• scope: The ASGI scope dictionary for the request.
• method: The HTTP method of the request.
• path_params: List of path parameters.
• query_params: Dictionary of query parameters.
• body_params: Dictionary of body parameters.
• get_json: JSON request body object.
• session: Dictionary of session data.
• files: Dictionary of multipart data/streamed content.
• headers: Dictionary of headers.

WebSocketRequest Class

Represents a WebSocket request in the MicroPie framework, inheriting from Request.

Attributes

• Inherits all attributes from Request.

WebSocket Management

WebSocket Class

Manages WebSocket communication in the MicroPie framework.

Methods

• __init__(receive: Callable[[], Awaitable[Dict[str, Any]]], send: Callable[[Dict[str, Any]], Awaitable[None]]) -> None

  • Initializes a WebSocket instance with ASGI receive and send callables.

• accept(subprotocol: Optional[str] = None, session_id: Optional[str] = None) -> None

  • Accepts the WebSocket connection, optionally specifying a subprotocol and session ID for setting a cookie during the handshake.

• receive_text() -> str

  • Receives a text message from the WebSocket. Raises ConnectionClosed if the connection is closed or ValueError for unexpected message types.

• receive_bytes() -> bytes

  • Receives a binary message from the WebSocket. Raises ConnectionClosed if the connection is closed or ValueError for unexpected message types.

• send_text(data: str) -> None

  • Sends a text message over the WebSocket. Raises RuntimeError if the connection is not accepted.

• send_bytes(data: bytes) -> None

  • Sends a binary message over the WebSocket. Raises RuntimeError if the connection is not accepted.

• close(code: int = 1000, reason: Optional[str] = None) -> None

  • Closes the WebSocket connection with the specified code and optional reason.

Attributes

• accepted: Boolean indicating if the WebSocket connection is accepted.
• session_id: Optional string storing the session ID set during the handshake.

ConnectionClosed Class

An exception raised when a WebSocket connection is closed.

Application Base

App Class

The main ASGI application class for handling HTTP and WebSocket requests in MicroPie.

Attributes

• middlewares: List of HttpMiddleware instances for HTTP request processing.
• ws_middlewares: List of WebSocketMiddleware instances for WebSocket request processing.

Methods

• __init__(session_backend: Optional[SessionBackend] = None) -> None

  • Initializes the application with an optional session backend and empty middleware lists for HTTP and WebSocket requests.

• request -> Request

  • Retrieves the current request from the context variable.

• __call__(scope: Dict[str, Any], receive: Callable[[], Awaitable[Dict[str, Any]]], send: Callable[[Dict[str, Any]], Awaitable[None]]) -> None

  • ASGI callable interface for the server. Checks scope type and dispatches to HTTP or WebSocket handling.

• _asgi_app_http(scope: Dict[str, Any], receive: Callable[[], Awaitable[Dict[str, Any]]], send: Callable[[Dict[str, Any]], Awaitable[None]]) -> None

  • ASGI application entry point for handling HTTP requests.

• _asgi_app_websocket(scope: Dict[str, Any], receive: Callable[[], Awaitable[Dict[str, Any]]], send: Callable[[Dict[str, Any]], Awaitable[None]]) -> None

  • ASGI application entry point for handling WebSocket requests. Routes to methods prefixed with ws_ (e.g., ws_chat for /chat).

• request(self) -> Request

  • Accessor for the current request object. Returns the current request from the context variable.

• _parse_cookies(cookie_header: str) -> Dict[str, str]

  • Parses the Cookie header and returns a dictionary of cookie names and values.

• _parse_multipart(reader: asyncio.StreamReader, boundary: bytes) -> Tuple[Dict[str, List[str]], Dict[str, Dict[str, Any]]]

  • Asynchronously parses multipart/form-data from the given reader using the specified boundary. Returns a tuple of two dictionaries: form_data (text fields as key-value pairs) and files (file fields with metadata). Each file entry in files contains:
    • filename: The original filename of the uploaded file.
    • content_type: The MIME type of the file (defaults to application/octet-stream).
    • content: An asyncio.Queue containing chunks of file data as bytes, with a None sentinel signaling the end of the stream.
  • Handlers can consume the file data by iterating over the queue (e.g., using await queue.get()).
  • Requires: multipart

• _send_response(send: Callable[[Dict[str, Any]], Awaitable[None]], status_code: int, body: Any, extra_headers: Optional[List[Tuple[str, str]]] = None) -> None

  • Sends an HTTP response using the ASGI send callable.

• _send_websocket_close(send: Callable[[Dict[str, Any]], Awaitable[None]], code: int, reason: str) -> None

  • Sends a WebSocket close message with the specified code and reason.

• _redirect(location: str) -> Tuple[int, str]

  • Generates an HTTP redirect response.

• _render_template(name: str, **kwargs: Any) -> str

  • Renders a template asynchronously using Jinja2.
  • Requires: jinja2

The App class is the main entry point for creating MicroPie applications. It implements the ASGI interface and handles HTTP and WebSocket requests.

Response Formats

Handlers can return responses in the following formats:

1. String or bytes or JSON
2. Tuple of (status_code, body)
3. Tuple of (status_code, body, headers)
4. Async or sync generator for streaming responses

Error Handling

MicroPie provides built-in error handling for common HTTP status codes:

• 404 Not Found: Automatically returned for non-existent routes
• 400 Bad Request: Returned for missing required parameters
• 500 Internal Server Error: Returned for unhandled exceptions

Custom error handling can be implemented through middleware.

---

© 2025 Harrison Erd