amqtt/docs/references/mqttclient.md

# MQTTClient API

The `amqtt.client.MQTTClient` class implements the client part of MQTT protocol. It can be used to publish and/or subscribe MQTT message on a broker accessible on the network through TCP or websocket protocol, both secured or unsecured.

## Usage examples

### Subscriber

The example below shows how to write a simple MQTT client which subscribes a topic and prints every messages received from the broker:

```python
import logging
import asyncio

from amqtt.client import MQTTClient, ClientException
from amqtt.mqtt.constants import QOS_1, QOS_2

logger = logging.getLogger(__name__)

async def uptime_coro():
    C = MQTTClient()
    await C.connect('mqtt://test.mosquitto.org/')
    # Subscribe to '$SYS/broker/uptime' with QOS=1
    # Subscribe to '$SYS/broker/load/#' with QOS=2
    await C.subscribe([
            ('$SYS/broker/uptime', QOS_1),
            ('$SYS/broker/load/#', QOS_2),
         ])
    try:
        for i in range(1, 100):
            message = await C.deliver_message()
            packet = message.publish_packet
            print("%d:  %s => %s" % (i, packet.variable_header.topic_name, str(packet.payload.data)))
        await C.unsubscribe(['$SYS/broker/uptime', '$SYS/broker/load/#'])
        await C.disconnect()
    except ClientException as ce:
        logger.error("Client exception: %s" % ce)

if __name__ == '__main__':
    formatter = "[%(asctime)s] %(name)s {%(filename)s:%(lineno)d} %(levelname)s - %(message)s"
    logging.basicConfig(level=logging.DEBUG, format=formatter)
    asyncio.get_event_loop().run_until_complete(uptime_coro())
```

When executed, this script gets the default event loop and asks it to run the `uptime_coro` until it completes.
`uptime_coro` starts by initializing a `MQTTClient` instance.
The coroutine then calls `connect()` to connect to the broker, here `test.mosquitto.org`.
Once connected, the coroutine subscribes to some topics, and then wait for 100 messages. Each message received is simply written to output.
Finally, the coroutine unsubscribes from topics and disconnects from the broker.

### Publisher

The example below uses the `MQTTClient` class to implement a publisher.
This test publish 3 messages asynchronously to the broker on a test topic.
For the purposes of the test, each message is published with a different Quality Of Service.
This example also shows two methods for publishing messages asynchronously.

```python
import logging
import asyncio

from amqtt.client import MQTTClient
from amqtt.mqtt.constants import QOS_0, QOS_1, QOS_2

logger = logging.getLogger(__name__)

async def test_coro():
    C = MQTTClient()
    await C.connect('mqtt://test.mosquitto.org/')
    tasks = [
        asyncio.ensure_future(C.publish('a/b', b'TEST MESSAGE WITH QOS_0')),
        asyncio.ensure_future(C.publish('a/b', b'TEST MESSAGE WITH QOS_1', qos=QOS_1)),
        asyncio.ensure_future(C.publish('a/b', b'TEST MESSAGE WITH QOS_2', qos=QOS_2)),
    ]
    await asyncio.wait(tasks)
    logger.info("messages published")
    await C.disconnect()


async def test_coro2():
    try:
        C = MQTTClient()
        ret = await C.connect('mqtt://test.mosquitto.org:1883/')
        message = await C.publish('a/b', b'TEST MESSAGE WITH QOS_0', qos=QOS_0)
        message = await C.publish('a/b', b'TEST MESSAGE WITH QOS_1', qos=QOS_1)
        message = await C.publish('a/b', b'TEST MESSAGE WITH QOS_2', qos=QOS_2)
        #print(message)
        logger.info("messages published")
        await C.disconnect()
    except ConnectException as ce:
        logger.error("Connection failed: %s" % ce)
        asyncio.get_event_loop().stop()


if __name__ == '__main__':
    formatter = "[%(asctime)s] %(name)s {%(filename)s:%(lineno)d} %(levelname)s - %(message)s"
    logging.basicConfig(level=logging.DEBUG, format=formatter)
    asyncio.get_event_loop().run_until_complete(test_coro())
    asyncio.get_event_loop().run_until_complete(test_coro2())
```

As usual, the script runs the publish code through the async loop. `test_coro()` and `test_coro2()` are ran in sequence.
Both do the same job. `test_coro()` publishes 3 messages in sequence. `test_coro2()` publishes the same message asynchronously.
The difference appears when looking at the sequence of MQTT messages sent.

`test_coro()` achieves:
```
amqtt/YDYY;NNRpYQSy3?o -out-> PublishPacket(ts=2015-11-11 21:54:48.843901, fixed=MQTTFixedHeader(length=28, flags=0x0), variable=PublishVariableHeader(topic=a/b, packet_id=None), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_0'"))
amqtt/YDYY;NNRpYQSy3?o -out-> PublishPacket(ts=2015-11-11 21:54:48.844152, fixed=MQTTFixedHeader(length=30, flags=0x2), variable=PublishVariableHeader(topic=a/b, packet_id=1), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_1'"))
amqtt/YDYY;NNRpYQSy3?o <-in-- PubackPacket(ts=2015-11-11 21:54:48.979665, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=1), payload=None)
amqtt/YDYY;NNRpYQSy3?o -out-> PublishPacket(ts=2015-11-11 21:54:48.980886, fixed=MQTTFixedHeader(length=30, flags=0x4), variable=PublishVariableHeader(topic=a/b, packet_id=2), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_2'"))
amqtt/YDYY;NNRpYQSy3?o <-in-- PubrecPacket(ts=2015-11-11 21:54:49.029691, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=2), payload=None)
amqtt/YDYY;NNRpYQSy3?o -out-> PubrelPacket(ts=2015-11-11 21:54:49.030823, fixed=MQTTFixedHeader(length=2, flags=0x2), variable=PacketIdVariableHeader(packet_id=2), payload=None)
amqtt/YDYY;NNRpYQSy3?o <-in-- PubcompPacket(ts=2015-11-11 21:54:49.092514, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=2), payload=None)
```

while `test_coro2()` runs:
```
amqtt/LYRf52W[56SOjW04 -out-> PublishPacket(ts=2015-11-11 21:54:48.466123, fixed=MQTTFixedHeader(length=28, flags=0x0), variable=PublishVariableHeader(topic=a/b, packet_id=None), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_0'"))
amqtt/LYRf52W[56SOjW04 -out-> PublishPacket(ts=2015-11-11 21:54:48.466432, fixed=MQTTFixedHeader(length=30, flags=0x2), variable=PublishVariableHeader(topic=a/b, packet_id=1), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_1'"))
amqtt/LYRf52W[56SOjW04 -out-> PublishPacket(ts=2015-11-11 21:54:48.466695, fixed=MQTTFixedHeader(length=30, flags=0x4), variable=PublishVariableHeader(topic=a/b, packet_id=2), payload=PublishPayload(data="b'TEST MESSAGE WITH QOS_2'"))
amqtt/LYRf52W[56SOjW04 <-in-- PubackPacket(ts=2015-11-11 21:54:48.613062, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=1), payload=None)
amqtt/LYRf52W[56SOjW04 <-in-- PubrecPacket(ts=2015-11-11 21:54:48.661073, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=2), payload=None)
amqtt/LYRf52W[56SOjW04 -out-> PubrelPacket(ts=2015-11-11 21:54:48.661925, fixed=MQTTFixedHeader(length=2, flags=0x2), variable=PacketIdVariableHeader(packet_id=2), payload=None)
amqtt/LYRf52W[56SOjW04 <-in-- PubcompPacket(ts=2015-11-11 21:54:48.713107, fixed=MQTTFixedHeader(length=2, flags=0x0), variable=PacketIdVariableHeader(packet_id=2), payload=None)
```

Both coroutines have the same results except that `test_coro2()` manages messages flow in parallel which may be more efficient.

## Reference

### MQTTClient API

The `amqtt.client` module provides the following methods in the `MQTTClient` class:

- `connect()`: Connect to a remote broker
- `disconnect()`: Disconnect from the broker
- `reconnect()`: Reconnect to the broker
- `ping()`: Send a PING request to the broker
- `publish()`: Publish a message to the broker
- `subscribe()`: Subscribe to one or more topics
- `unsubscribe()`: Unsubscribe from one or more topics
- `deliver_message()`: Receive the next message from the broker

### MQTTClient configuration

The `MQTTClient` `__init__` method accepts a `config` parameter which allows setup of behavior and default settings. This argument must be a Python dict object which may contain the following entries:

* `keep_alive`: keep alive (in seconds) to send when connecting to the broker (defaults to `10` seconds). `MQTTClient` will *auto-ping* the broker if no message is sent within the keep-alive interval. This avoids disconnection from the broker.
* `ping_delay`: *auto-ping* delay before keep-alive times out (defaults to `1` seconds).
* `default_qos`: Default QoS (`0`) used by `publish()` if `qos` argument is not given.
* `default_retain`: Default retain (`False`) used by `publish()` if `qos` argument is not given.
* `auto_reconnect`: enable or disable auto-reconnect feature (defaults to `True`).
* `reconnect_max_interval`: maximum interval (in seconds) to wait before two connection retries (defaults to `10`).
* `reconnect_retries`: maximum number of connect retries (defaults to `2`). Negative value will cause client to reconnect infinitely.

Default QoS and default retain can also be overridden by adding a `topics` with may contain QoS and retain values for specific topics. See the following example:

```python
config = {
    'keep_alive': 10,
    'ping_delay': 1,
    'default_qos': 0,
    'default_retain': False,
    'auto_reconnect': True,
    'reconnect_max_interval': 5,
    'reconnect_retries': 10,
    'topics': {
        'test': { 'qos': 1 },
        'some_topic': { 'qos': 2, 'retain': True }
    }
}
```

With this setting any message published will be set with QOS_0 and retain flag unset except for:

* messages sent to `test` topic: they will be sent with QOS_1
* messages sent to `some_topic` topic: they will be sent with QOS_2 and retain flag set

In any case, the `qos` and `retain` argument values passed to method `publish()` will override these settings.