fastapi-mqtt

MQTT is a lightweight publish/subscribe messaging protocol designed for M2M (machine to machine) telemetry in low bandwidth environments. Fastapi-mqtt is the client for working with MQTT.

For more information about MQTT, please refer to here: MQTT

Fastapi-mqtt wraps around gmqtt module. Gmqtt Python async client for MQTT client implementation. Module has support of MQTT version 5.0 protocol

Documentation: FastApi-MQTT

The key feature are:

MQTT specification avaliable with help decarator methods using callbacks:

on_connect()
on_disconnect()
on_subscribe()
on_message()
subscribe(topic)
MQTT Settings available with pydantic class
Authentication to broker with credentials
unsubscribe certain topics and publish to certain topics

🔨 Installation

pip install fastapi-mqtt

🕹 Guide

from typing import Any

from fastapi import FastAPI
from gmqtt import Client as MQTTClient

from fastapi_mqtt import FastMQTT, MQTTConfig

mqtt_config = MQTTConfig()

fast_mqtt = FastMQTT(config=mqtt_config)

app = FastAPI()
fast_mqtt.init_app(app)


@fast_mqtt.on_connect()
def connect(client: MQTTClient, flags: int, rc: int, properties: Any):
    client.subscribe("/mqtt")  # subscribing mqtt topic
    print("Connected: ", client, flags, rc, properties)

@fast_mqtt.subscribe("mqtt/+/temperature", "mqtt/+/humidity", qos=1)
async def home_message(client: MQTTClient, topic: str, payload: bytes, qos: int, properties: Any):
    print("temperature/humidity: ", topic, payload.decode(), qos, properties)

@fast_mqtt.on_message()
async def message(client: MQTTClient, topic: str, payload: bytes, qos: int, properties: Any):
    print("Received message: ", topic, payload.decode(), qos, properties)

@fast_mqtt.subscribe("my/mqtt/topic/#", qos=2)
async def message_to_topic_with_high_qos(
    client: MQTTClient, topic: str, payload: bytes, qos: int, properties: Any
):
    print(
        "Received message to specific topic and QoS=2: ", topic, payload.decode(), qos, properties
    )

@fast_mqtt.on_disconnect()
def disconnect(client: MQTTClient, packet, exc=None):
    print("Disconnected")

@fast_mqtt.on_subscribe()
def subscribe(client: MQTTClient, mid: int, qos: int, properties: Any):
    print("subscribed", client, mid, qos, properties)

@app.get("/test")
async def func():
    fast_mqtt.publish("/mqtt", "Hello from Fastapi")  # publishing mqtt topic
    return {"result": True, "message": "Published"}

Publish method:

async def func():
    fast_mqtt.publish("/mqtt", "Hello from Fastapi")  # publishing mqtt topic
    return {"result": True, "message": "Published"}

Subscribe method:

@fast_mqtt.on_connect()
def connect(client, flags, rc, properties):
    client.subscribe("/mqtt")  # subscribing mqtt topic
    print("Connected: ", client, flags, rc, properties)

Changing connection params

mqtt_config = MQTTConfig(
    host="mqtt.mosquito.org",
    port=1883,
    keepalive=60,
    username="username",
    password="strong_password",
)
fast_mqtt = FastMQTT(config=mqtt_config)

✅ Testing

Clone the repository and install it with poetry.
Run tests with pytest, using an external MQTT broker to connect (defaults to 'test.mosquitto.org').
Explore the fastapi app examples and run them with uvicorn

# (opc) Run a local mosquitto MQTT broker with docker
docker run -d --name mosquitto -p 9001:9001 -p 1883:1883 eclipse-mosquitto:1.6.15
# Set host for test broker when running pytest
TEST_BROKER_HOST=localhost pytest
# Run the example apps against local broker, with uvicorn
TEST_BROKER_HOST=localhost uvicorn examples.app:app --port 8000 --reload
TEST_BROKER_HOST=localhost uvicorn examples.ws_app.app:application --port 8000 --reload