add ap for espnow, using mqtt_as

[Fisherman48]

Using espnow decentralized - with esp32 ap ssid/mac address to connect any client/ sensor esp.
My question : I just put in activation of ap in the up(client) function on the user level.
After a week without any issues, I'm curious, are ther other better/safer ways to activate ap (under the hood) ?
/Stefan

...
def runAp(): # start ap
    ap = network.WLAN(network.AP_IF)
    ap.active(True)
    # Print the wifi channel used AFTER finished connecting to access point
    # print("Proxy running on channel:", sta.config("channel"))
    # print("mac:", sta.config("mac"))
    print(ap.config('ssid'), ap.config('channel'), ap.config('mac') )

async def up(client): # Deal with WiFi outages
    while True:
        await client.up.wait()
        client.up.clear()
        await client.subscribe('ftx', 1)    # t,h,p from/to heat exchanger
        await client.subscribe('status', 1) # check/ cmds
        runAp() # start ap
        
async def main(client):  # Main program
    try:
        await client.connect()
    except OSError:
        print('Connection failed.')
    else:
        asyncio.create_task(up(client))
        asyncio.create_task(messages(client))
        asyncio.create_task(blink13())
        asyncio.create_task(api.sign())
        asyncio.create_task(asyncio.run(recv(e)))
        await x_handler()

# Define configuration
config['keepalive'] = 120
config["queue_len"] = 1  # Use event interface with default queue
client = MQTTClient(config)  # Set up client.
#client.DEBUG=True
e = aioespnow.AIOESPNow()  # Returns AIOESPNow enhanced with async support
e.active(True)

try:
    asyncio.run(main(client))
finally:
    print('finally new_event_loop ')
    client.close()
    asyncio.new_event_loop()

[peterhinch]

Have you tested the case where the client's WiFi link suffers an outage? The reason I ask is that up() will run runAp() again. I'm not sure whether this is OK when the AP is already running. An option might be

async def up(client): # Deal with WiFi outages
    while True:
        await client.up.wait()
        client.up.clear()
        # Down the AP and pause until complete
        await client.subscribe('ftx', 1)    # t,h,p from/to heat exchanger
        await client.subscribe('status', 1) # check/ cmds
        runAp() # start ap

It may be that there is no problem with the existing code, in which case it looks fine.

[Fisherman48]

OK - I do this for fun enyoing learning and tend to avoid destructive testing/ defensive coding.

After turned off the router/ wifi STA connection - there was no messages recieved/ printed from the espnow devices.

async def recv(e):
    async for peer, msg in e: # type(msg) bytestring
        # Receive espnow messages and forward them to MQTT broker over wifi
        print(peer, msg.decode('utf-8'))
        m=msg.decode('utf-8')
        await client.publish ('sensor espnow: ', m, 1) # for this test

Router on - first a print of latest subscribed msg (as usual when started) and then a normal print of espnow received msg.

So it looks like the ap is working without clear/reset (at least one outage).