Support `BaseModel` serialization to `Protobuf Message`

[so1n]

The current step of converting BaseModel to Protobuf Message is tedious and has poor performance. e.g:

from google.protobuf.json_format import ParseDict
from pydantic import BaseModel
from demo_pb2 import DemoMessage

class Demo(BaseModel):
    pass


ParseDict(Demo().dict(), DemoMessage())

Serialization can be optimized in the following ways:

• 1.Convert directly to Message:

  from protobuf_to_pydantic.p2p_model import P2PBaseModel
  
  class Demo(P2PBaseModel):
      pass
  
  Demo().to_message()

• 2.Serialized to dict, then converted to Message by the developer (JSON-like serialization on .dict() pydantic/pydantic#1409 (comment))

  from google.protobuf.json_format import ParseDict
  from pydantic import BaseModel
  from demo_pb2 import DemoMessage
  
  class Demo(BaseModel):
      pass
  
  
  DemoMessage(**Demo().dict())

[mikey0000]

So yeah came here to use this to convert from protobuf to model and am stumped as there isn't really strong documentation for more complex multi-level models or submessage types. Love that this generates all the models so easily so big thank you!

[so1n]

@mikey0000 Protobuf's JsonFormat implementation is more complex there are many cases that can not be dealt with, you need to modify his code to achieve. So I will delay the implementation of this feature.