Skip to content

Latest commit

 

History

History
322 lines (271 loc) · 9.55 KB

README.md

File metadata and controls

322 lines (271 loc) · 9.55 KB
Raw

openapi-pydantic

PyPI PyPI - License

OpenAPI schema implemented in Pydantic. Both Pydantic 1.8+ and 2.x are supported.

The naming of the classes follows the schema in OpenAPI specification.

This library is forked from OpenAPI Schema Pydantic (at version 1.2.4) which is no longer actively maintained.

Installation

pip install openapi-pydantic

Try me

from openapi_pydantic import OpenAPI, Info, PathItem, Operation, Response

# Construct OpenAPI by pydantic objects
open_api = OpenAPI(
    info=Info(
        title="My own API",
        version="v0.0.1",
    ),
    paths={
        "/ping": PathItem(
            get=Operation(
                responses={
                    "200": Response(
                        description="pong"
                    )
                }
            )
        )
    },
)
# Note: for Pydantic 1.x, replace `model_dump_json` with `json`
print(open_api.model_dump_json(by_alias=True, exclude_none=True, indent=2))

Result:

{
  "openapi": "3.1.1",
  "info": {
    "title": "My own API",
    "version": "v0.0.1"
  },
  "servers": [
    {
      "url": "/"
    }
  ],
  "paths": {
    "/ping": {
      "get": {
        "responses": {
          "200": {
            "description": "pong"
          }
        },
        "deprecated": false
      }
    }
  }
}

Take advantage of Pydantic

Pydantic is a great tool. It allows you to use object / dict / mixed data for input.

The following examples give the same OpenAPI result as above:

from openapi_pydantic import parse_obj, OpenAPI, PathItem, Response

# Construct OpenAPI from dict, inferring the correct schema version
open_api = parse_obj({
    "openapi": "3.1.1",
    "info": {"title": "My own API", "version": "v0.0.1"},
    "paths": {
        "/ping": {
            "get": {"responses": {"200": {"description": "pong"}}}
        }
    },
})


# Construct OpenAPI v3.1 schema from dict
# Note: for Pydantic 1.x, replace `model_validate` with `parse_obj`
open_api = OpenAPI.model_validate({
    "info": {"title": "My own API", "version": "v0.0.1"},
    "paths": {
        "/ping": {
            "get": {"responses": {"200": {"description": "pong"}}}
        }
    },
})

# Construct OpenAPI with mix of dict/object
# Note: for Pydantic 1.x, replace `model_validate` with `parse_obj`
open_api = OpenAPI.model_validate({
    "info": {"title": "My own API", "version": "v0.0.1"},
    "paths": {
        "/ping": PathItem(
            get={"responses": {"200": Response(description="pong")}}
        )
    },
})

Use Pydantic classes as schema

  • The Schema Object in OpenAPI has definitions and tweaks in JSON Schema, which are hard to comprehend and define a good data class
  • Pydantic already has a good way to create JSON schema. Let's not reinvent the wheel.

The approach to deal with this:

  1. Use PydanticSchema objects to represent the Schema in OpenAPI object
  2. Invoke construct_open_api_with_schema_class to resolve the JSON schemas and references
from pydantic import BaseModel, Field

from openapi_pydantic import OpenAPI
from openapi_pydantic.util import PydanticSchema, construct_open_api_with_schema_class

def construct_base_open_api() -> OpenAPI:
    # Note: for Pydantic 1.x, replace `model_validate` with `parse_obj`
    return OpenAPI.model_validate({
        "info": {"title": "My own API", "version": "v0.0.1"},
        "paths": {
            "/ping": {
                "post": {
                    "requestBody": {"content": {"application/json": {
                        "schema": PydanticSchema(schema_class=PingRequest)
                    }}},
                    "responses": {"200": {
                        "description": "pong",
                        "content": {"application/json": {
                            "schema": PydanticSchema(schema_class=PingResponse)
                        }},
                    }},
                }
            }
        },
    })

class PingRequest(BaseModel):
    """Ping Request"""
    req_foo: str = Field(description="foo value of the request")
    req_bar: str = Field(description="bar value of the request")

class PingResponse(BaseModel):
    """Ping response"""
    resp_foo: str = Field(description="foo value of the response")
    resp_bar: str = Field(description="bar value of the response")

open_api = construct_base_open_api()
open_api = construct_open_api_with_schema_class(open_api)

# print the result openapi.json
# Note: for Pydantic 1.x, replace `model_dump_json` with `json`
print(open_api.model_dump_json(by_alias=True, exclude_none=True, indent=2))

Result:

{
  "openapi": "3.1.1",
  "info": {
    "title": "My own API",
    "version": "v0.0.1"
  },
  "servers": [
    {
      "url": "/"
    }
  ],
  "paths": {
    "/ping": {
      "post": {
        "requestBody": {
          "content": {
            "application/json": {
              "schema": {
                "$ref": "#/components/schemas/PingRequest"
              }
            }
          },
          "required": false
        },
        "responses": {
          "200": {
            "description": "pong",
            "content": {
              "application/json": {
                "schema": {
                  "$ref": "#/components/schemas/PingResponse"
                }
              }
            }
          }
        },
        "deprecated": false
      }
    }
  },
  "components": {
    "schemas": {
      "PingRequest": {
        "title": "PingRequest",
        "required": [
          "req_foo",
          "req_bar"
        ],
        "type": "object",
        "properties": {
          "req_foo": {
            "title": "Req Foo",
            "type": "string",
            "description": "foo value of the request"
          },
          "req_bar": {
            "title": "Req Bar",
            "type": "string",
            "description": "bar value of the request"
          }
        },
        "description": "Ping Request"
      },
      "PingResponse": {
        "title": "PingResponse",
        "required": [
          "resp_foo",
          "resp_bar"
        ],
        "type": "object",
        "properties": {
          "resp_foo": {
            "title": "Resp Foo",
            "type": "string",
            "description": "foo value of the response"
          },
          "resp_bar": {
            "title": "Resp Bar",
            "type": "string",
            "description": "bar value of the response"
          }
        },
        "description": "Ping response"
      }
    }
  }
}

Notes

Use of OpenAPI.model_dump() / OpenAPI.model_dump_json() / OpenAPI.json() / OpenAPI.dict()

When using OpenAPI.model_dump() / OpenAPI.model_dump_json() / OpenAPI.json() / OpenAPI.dict() functions, the arguments by_alias=True, exclude_none=True have to be in place. Otherwise the resulting json will not fit the OpenAPI standard.

# OK (Pydantic 2)
open_api.model_dump_json(by_alias=True, exclude_none=True, indent=2)
# OK (Pydantic 1)
open_api.json(by_alias=True, exclude_none=True, indent=2)

# Not good
open_api.model_dump_json(indent=2)
open_api.json(indent=2)

More info about field aliases:

OpenAPI version Field alias info
3.1 here
3.0 here

Non-pydantic schema types

Some schema types are not implemented as pydantic classes. Please refer to the following for more info:

OpenAPI version Non-pydantic schema type info
3.1 here
3.0 here

Use OpenAPI 3.0 instead of 3.1

Some UI renderings (e.g. Swagger) still do not support OpenAPI 3.1.x. The old 3.0.x version is available by importing from different paths:

from openapi_pydantic.v3.v3_0 import OpenAPI, ...
from openapi_pydantic.v3.v3_0.util import PydanticSchema, construct_open_api_with_schema_class

Pydantic version compatibility

Compatibility with both major versions of Pydantic (1.8+ and 2.*) is mostly achieved using a module called compat.py. It detects the installed version of Pydantic and exports version-specific symbols for use by the rest of the package. It also provides all symbols necessary for type checking. The compat.py module is not intended to be imported by other packages, but other packages may find it helpful as an example of how to span major versions of Pydantic.

Credits

This library is based from the original implementation by Kuimono of OpenAPI Schema Pydantic which is no longer actively maintained.

License

MIT License