Join us and get access to hundreds of tutorials and a community of expert Pythonistas.

Unlock This Lesson

This lesson is for members only. Join us and get access to hundreds of tutorials and a community of expert Pythonistas.

Unlock This Lesson

Hint: You can adjust the default video playback speed in your account settings.
Hint: You can set the default subtitles language in your account settings.
Sorry! Looks like there’s an issue with video playback 🙁 This might be due to a temporary outage or because of a configuration issue with your browser. Please see our video player troubleshooting guide to resolve the issue.

Virtual Environment and FastAPI Setup

00:00 Before creating a REST API with FastAPI, it’s a good idea to create a new Python virtual environment. I’m in Visual Studio Code on Windows, but you can follow along on macOS or even Linux, or you can use your own editor, but Visual Studio Code has great Python support.

00:19 Press Control + ` to open the integrated terminal in Visual Studio Code, create a new directory for the demo project,

00:31 and then run Python -m venv .env to create a new virtual environment. Depending on how Python is installed on your machine, you may be able to invoke Python itself.

00:45 The -m option invokes a module named venv that will create a new virtual environment in a directory called .env, but you can name it whatever you like. After the virtual environment is finished, open the project directory in Visual Studio, and this can be done by executing this command.

01:07 In the project directory, Visual Studio will restart. Next press Control + Shift + P, or Command + Shift + P on macOS, to open the command pallet.

01:18 The Python extension provides support for virtual environments with the Python: Select Interpreter command. This will look for Python virtual environments—including those created by Conda, pipenv, and Poetry—and prompt you to select one.

01:33 Next, open a new integrated terminal, and VS Code will automatically activate the virtual environment for you. At the command line, install the FastAPI package using pip.

01:49 Create a new Python file called main.py in the project directory.

01:56 From the fastapi module, import the FastAPI class. Next, create an instance of the FastAPI class. This will be the FastAPI app, which you’ll run later.

02:10 Now you need to create a function to handle an HTTP request to the API.

02:17 Notice that you’re returning a dictionary, but recall in the last lesson that data returned from a REST API is usually JSON. If you’ve worked with the JSON module in the Python standard library, you know that marshaling data between Python dictionaries and JSON objects is a cinch.

02:35 The two are almost syntactically identical. FastAPI will translate this data from a dictionary to JSON for you and return it in an HTTP response with the appropriate status code, 200 for OK in this case.

02:52 To map the function to an endpoint, use the @get decorator from the FastAPI app, The @get decorator accepts a path. Now, any request using the .get() method or verb to the root of the API will be handled by the root() function.

03:10 One more thing: recall that FastAPI supports async. That’s easy to add. Simply use the async keyword before the def keyword, and your function is now async.

03:25 Because FastAPI supports async, the API needs to be run on a server that supports async using the ASGI protocol. You may have heard of WSGI, or Web Server Gateway Interface.

03:39 For example, the package for Flask implements a WSGI server you could use to run Flask applications. ASGI is Async Server Gateway Interface, and you can install the uvicorn package and use it to serve your API.

03:58 To use uvicorn to serve your API, you can import it into main.py and invoke the server in the entry point. This would be the only choice if you were using something like Replit to develop your API. However, it’s easier to run it from the command line.

04:15 This will tell uvicorn to run an ASGI app named app in the main module, listen on all IP addresses on port 8000, and to automatically reload the application if the code changes.

04:32 If you navigate to http://localhost:8000 in a browser, you’ll see the JSON returned by the root() function. Congratulations! You’ve created your first REST API using FastAPI. In the next lesson, you’ll learn how to use FastAPI’s interactive documentation.

sethcstenzel on April 12, 2022

Hmm, trying to follow this tutorial I am getting the following error when running uvicorn directly, and via import:

  File "C:\Users\s3711\Desktop\PYTHON\RealPython\basic\Python REST APIs With FastAPI\env\lib\site-packages\typing_extensions.py", line 146, in _collect_type_vars
    if _should_collect_from_parameters(t):
  File "C:\Users\s3711\Desktop\PYTHON\RealPython\basic\Python REST APIs With FastAPI\env\lib\site-packages\typing_extensions.py", line 120, in _should_collect_from_parameters
    t, (typing._GenericAlias, _types.GenericAlias, _types.UnionType)
AttributeError: module 'types' has no attribute 'UnionType'. Did you mean: 'FunctionType'?

Any ideas?

from fastapi import FastAPI
import uvicorn

app = FastAPI()

@app.get("/")
async def root():
    return {"message": "Hello world"}


if __name__ == "__main__":
    uvicorn.run("main:app", host="0.0.0.0", port=8080, log_level="info")

sethcstenzel on April 12, 2022

I found out this was a known issue with using beta version 3.10.0b3. After checking which version you were using I installed 3.10.0 and this resolved the issue. Seems it was a known issue: github.com/samuelcolvin/pydantic/issues/3300

Douglas Starnes RP Team on April 13, 2022

@sethcstenzel

You can use Python back to 3.6. The course uses 3.9 and I’ve been using 3.10 without any issues. If you are working from the Repl online, it uses 3.8 and I have not had any problems. I try to avoid using betas when possible.

Become a Member to join the conversation.