pydantic-to-typescript

A simple CLI tool for converting pydantic models into typescript interfaces. It supports all versions of pydantic, with polyfills for older versions to ensure that the resulting typescript definitions are stable and accurate.

Useful for any scenario in which python and javascript applications are interacting, since it allows you to have a single source of truth for type definitions.

This tool requires that you have the lovely json2ts CLI utility installed. Instructions can be found here: https://www.npmjs.com/package/json-schema-to-typescript

Installation

$ pip install pydantic-to-typescript

Pydantic V2 support

If you are encountering issues with pydantic>2, it is most likely because you're using an old version of pydantic-to-typescript. Run pip install 'pydantic-to-typescript>2' and/or add pydantic-to-typescript>=2 to your project requirements.

CI/CD

You can now use pydantic-to-typescript to automatically validate and/or update typescript definitions as part of your CI/CD pipeline.

The github action can be found here: https://github.com/marketplace/actions/pydantic-to-typescript. The available inputs are documented here: https://github.com/phillipdupuis/pydantic-to-typescript/blob/master/action.yml.

CLI

Prop	Description
‑‑module	name or filepath of the python module you would like to convert. All the pydantic models within it will be converted to typescript interfaces. Discoverable submodules will also be checked.
‑‑output	name of the file the typescript definitions should be written to. Ex: './frontend/apiTypes.ts'
‑‑exclude	name of a pydantic model which should be omitted from the resulting typescript definitions. This option can be defined multiple times, ex: --exclude Foo --exclude Bar to exclude both the Foo and Bar models from the output.
‑‑json2ts‑cmd	optional, the command used to invoke json2ts. The default is 'json2ts'. Specify this if you have it installed locally (ex: 'yarn json2ts') or if the exact path to the executable is required (ex: /myproject/node_modules/bin/json2ts)

---

Usage

Define your pydantic models (ex: /backend/api.py):

from fastapi import FastAPI
from pydantic import BaseModel
from typing import List, Optional

api = FastAPI()

class LoginCredentials(BaseModel):
    username: str
    password: str

class Profile(BaseModel):
    username: str
    age: Optional[int]
    hobbies: List[str]

class LoginResponseData(BaseModel):
    token: str
    profile: Profile

@api.post('/login/', response_model=LoginResponseData)
def login(body: LoginCredentials):
    profile = Profile(**body.dict(), age=72, hobbies=['cats'])
    return LoginResponseData(token='very-secure', profile=profile)

Execute the command for converting these models into typescript definitions, via:

$ pydantic2ts --module backend.api --output ./frontend/apiTypes.ts

or:

$ pydantic2ts --module ./backend/api.py --output ./frontend/apiTypes.ts

or:

from pydantic2ts import generate_typescript_defs

generate_typescript_defs("backend.api", "./frontend/apiTypes.ts")

The models are now defined in typescript...

/* tslint:disable */
/**
/* This file was automatically generated from pydantic models by running pydantic2ts.
/* Do not modify it by hand - just update the pydantic models and then re-run the script
*/

export interface LoginCredentials {
  username: string;
  password: string;
}
export interface LoginResponseData {
  token: string;
  profile: Profile;
}
export interface Profile {
  username: string;
  age?: number;
  hobbies: string[];
}

...and can be used in your typescript code with complete confidence.

import { LoginCredentials, LoginResponseData } from "./apiTypes.ts";

async function login(
  credentials: LoginCredentials,
  resolve: (data: LoginResponseData) => void,
  reject: (error: string) => void
) {
  try {
    const response: Response = await fetch("/login/", {
      method: "POST",
      headers: { "Content-Type": "application/json" },
      body: JSON.stringify(credentials),
    });
    const data: LoginResponseData = await response.json();
    resolve(data);
  } catch (error) {
    reject(error.message);
  }
}