Task arguments with Pydantic models and aliases

[mromanelli9]

First, I'd like to thank the author and all contributors for creating and maintaining this great library.

I encountered an issue while deserializing Pydantic models with aliased fields from task arguments.
As a temporary workaround, I created a custom wrapper around AsyncKicker class to call Pydantic model_dump with by_alias=True. However, I was wondering if there might be a better solution to allow developers to more easily customize Pydantic serialization/deserialization.

Example

import asyncio
from pydantic import BaseModel, Field
from taskiq_redis import ListQueueBroker, RedisAsyncResultBackend

REDIS_URL = "redis://localhost:6379"
broker = ListQueueBroker(REDIS_URL).with_result_backend(RedisAsyncResultBackend(REDIS_URL))

class ExampleModel(BaseModel):
    value_: str = Field(alias="value")

@broker.task
async def my_task(data: ExampleModel):
    print("Hello task")

async def main():
    data = ExampleModel(value="example")
    await my_task.kiq(data)

if __name__ == "__main__":
    asyncio.run(main())

The receiver raises an error when attempting to cast the task arguments.

[][taskiq.receiver.params_parser][WARNING][worker-0] Can't parse argument 0 for task my_task. Reason: 1 validation error for ExampleModel
value
  Field required [type=missing, input_value={'value_': 'example'}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.10/v/missing

[tuschla]

I second this. This is a regular use case especially when using FastAPI, so it should support aliases out of the box instead of throwing strange errors.