py-grpcio 1.5.4

gRPC with autogen by Pydantic models

Pydantic & gRPC

py-grpcio is a microframework and high-level wrapper of grpcio to simplify work with the original library using abstractions, useful python objects and pydantic models.

Examples of use are given below and also duplicated in the example directory.

---

Install latest

pip install py-grpcio

---

Example

Models

Pydantic models that describe messages for client-server interaction.

from uuid import UUID, uuid4
from datetime import datetime

from pydantic import Field

from py_grpcio import Message

from example.server.service.enums import Names


class PingRequest(Message):
    id: UUID = Field(default_factory=uuid4)


class PingResponse(Message):
    id: UUID
    timestamp: datetime = Field(default_factory=datetime.now)


class ComplexModel(Message):
    name: Names


class ComplexRequest(Message):
    id: UUID
    model: ComplexModel


class ComplexResponse(Message):
    id: UUID
    model: ComplexModel

---

Server

Basic implementation of gRPC services on the server side.

You need to describe the service abstractly and duplicate this service on the client side.

from abc import abstractmethod

from py_grpcio import BaseService

from example.server.service.models import PingRequest, PingResponse, ComplexRequest, ComplexResponse


class BaseExampleService(BaseService):
    @abstractmethod
    async def ping(self: 'BaseExampleService', request: PingRequest) -> PingResponse:
        ...

    @abstractmethod
    async def complex(self: 'BaseExampleService', request: ComplexRequest) -> ComplexResponse:
        ...

---

Full implementation of the gRPC service with methods.

from example.server.service.base import BaseExampleService
from example.server.service.models import PingRequest, PingResponse, ComplexRequest, ComplexResponse


class ExampleService(BaseExampleService):
    async def ping(self: 'ExampleService', request: PingRequest) -> PingResponse:
        return PingResponse(id=request.id)

    async def complex(self: 'BaseExampleService', request: ComplexRequest) -> ComplexResponse:
        return ComplexResponse(**request.model_dump())

---

Run the ExampleService on Server.

from py_grpcio import BaseServer

from example.server.service import ExampleService


if __name__ == '__main__':
    server: BaseServer = BaseServer()
    server.add_service(service=ExampleService)
    server.run()

---

Note that on the client side, this class must be named the same as it is named in the full server-side implementation.

That is, if on the server we call the base class as BaseExampleService and the class with the implementation of methods as ExampleService, then on the client side the abstract service should be called ExampleService.

Client

from abc import abstractmethod

from py_grpcio import BaseService

from example.server.service.models import PingRequest, PingResponse, ComplexRequest, ComplexResponse


class ExampleService(BaseService):
    @abstractmethod
    async def ping(self: 'ExampleService', request: PingRequest) -> PingResponse:
        ...

    @abstractmethod
    async def complex(self: 'ExampleService', request: ComplexRequest) -> ComplexResponse:
        ...

---

Calling the ExampleService endpoints by Client.

from uuid import uuid4
from asyncio import run

from loguru import logger

from example.client.services.example.enums import Names
from example.client.services.example import (
    ExampleService, PingRequest, PingResponse, ComplexModel, ComplexRequest, ComplexResponse
)

service: ExampleService = ExampleService(host='127.0.0.1')


async def main() -> None:
    response: PingResponse = await service.ping(request=PingRequest())
    logger.info(f'ping response: {response}')

    response: ComplexResponse = await service.complex(
        request=ComplexRequest(id=uuid4(), model=ComplexModel(name=Names.NAME_1))
    )
    logger.info(f'complex response: {response}')


if __name__ == '__main__':
    run(main())

---

Notes

• You can use the library on the client side even if the server is implemented differently by simply describing it as an abstract service

• The client can also be implemented using other libraries, the server that uses py-grpcio will still be able to accept such requests