Ceph » mgr » Dashboard

Python (as well as Typescript) lacks runtime typechecking, this may lead to a lot runtime issues. Part of this is alleviated by QA/Back-end API tests, but they cannot cover all the possible real-life situations.

Additionally, there's a lot of (potential) duplicate type definitions at both Front-end (Typescript) and Back-end (Python).

One of the most mature solutions for tackling these concerns is pydantic. For our purposes, it fulfills the following requirements:

• Runtime type-checking: integrated with mypy.
  

  >>> class RBD(pydantic.BaseModel):
  ...   id: int
  ...   name: str
  ... 
  
  >>> RBD(id=1, name='test_rbd')
  RBD(id=1, name='test_rbd')
  
  >>> RBD(id='1', name='test_rbd')
  RBD(id=1, name='test_rbd')
  
  >>> RBD(id='asd', name='test_rbd')
  Traceback (most recent call last):
    File "<stdin>", line 1, in <module>
    File "pydantic/main.py", line 338, in pydantic.main.BaseModel.__init__
  pydantic.error_wrappers.ValidationError: 1 validation error for RBD
  id
    value is not a valid integer (type=type_error.integer)
  
  

• Export Python types/schemas to JSON Schema/OpenAPI: this could be later imported to Angular/Typescript via 3rd party tools
  

  # OpenAPI compliant
  >>> RBD.schema_json()
  '{"title": "RBD", "type": "object", "properties": {"id": {"title": "Id", "type": "integer"}, "name": {"title": "Name", "type": "string"}}, "required": ["id", "name"]}'
  

As a con, current package support is limited to latest upstream distros *