Feature: Marshmallow support along with pydantic? about propan

[Lancetnik]

@shouldsee, hi there! Thanks for you kind feedback
Unfotunatelly, I can't support all serialization libraries, but you have a way to integrate your custom one
Just use the decode_message argument

from marshmallow import Schema, fields

from propan import RabbitBroker
from propan.brokers.rabbit import RabbitMessage
from propan.annotations import Logger


broker = RabbitBroker()
app = PropanApp(broker)

class User(Schema):
    name = fields.Str()

async def decode(msg: RabbitMessage, original):
    return User().loads(msg.body)

@broker.handle("test", decode_message=decode_marshmallow)
async def handle(msg, logger: Logger):
    logger.info(msg)

Or you can scale this code to any Schema via a closure

from marshmallow import Schema, fields

from propan import PropanApp, RabbitBroker
from propan.brokers.rabbit import RabbitMessage
from propan.annotations import Logger


broker = RabbitBroker()
app = PropanApp(broker)

class User(Schema):
    name = fields.Str()

def decode_marshmallow(schema: Schema):
    async def decode(msg: RabbitMessage, original):
        return schema.loads(msg.body)
    return decode

@broker.handle("test", decode_message=decode_marshmallow(User()))
async def handle(msg, logger: Logger):
    logger.info(msg)

@app.after_startup
async def test():
    await broker.publish({
        "name": "john"
    }, "test")

[Lancetnik]

I am using pydantic '.json_schema()' to get information about payloads
Other staff is generated by Propan: servers information, channels, bindings, etc
BTW, I will add an ability to set schema manually in custom serialization case in the PropanV2

[shouldsee]

Many thanks! BTW may I ask how the api documentation is generated from pydantic models? My main objective is to auto-generate api doc....