Building a Simple API

You can build APIs with plain Flask routes.

diagram the shape of one REST resource mermaid

  flowchart LR
  subgraph "/api/books"
    A["GET -> 200, a list"]
    B["POST -> 201 + Location, the created item"]
  end
  subgraph "/api/books/<id>"
    C["GET -> 200, or 404"]
    D["PUT/PATCH -> 200, the updated item"]
    E["DELETE -> 204, no body"]
  end
  F["a method the rule does not list"] --> G["405 Method Not Allowed"]
  H["a body that fails validation"] --> I["400 with the errors, not 500"]

A collection URL and an item URL, with the method deciding the operation. The pairing matters: POST goes to the collection because the server assigns the id, while PUT and DELETE need an item that already exists. Returning 201 with a Location header on create is what tells the client where the new thing lives.

Example: in-memory todo API (learning)

python

from flask import Flask, request
 
app = Flask(__name__)
 
todos = [
    {"id": 1, "title": "Learn Flask", "done": False},
]
 
 
@app.get("/api/todos")
def list_todos():
    return {"items": todos}
 
 
@app.post("/api/todos")
def create_todo():
    payload = request.get_json(silent=True) or {}
    title = (payload.get("title") or "").strip()
 
    if not title:
        return {"error": "validation_error", "message": "title is required"}, 400
 
    new_id = max([t["id"] for t in todos], default=0) + 1
    todo = {"id": new_id, "title": title, "done": False}
    todos.append(todo)
 
    return todo, 201

Notes

.get_json(silent=True) avoids raising exceptions on bad JSON
Always validate input
Use status codes consistently

Moving from in-memory to database

In real apps:

replace the list with SQLAlchemy models
implement to_dict() methods
add pagination and authentication

🧪 Try It Yourself

Exercise 1 – Create a Flask App

Exercise 2 – Dynamic Route

Exercise 3 – Return JSON

pch.coffeeTagline

pch.coffeeCta

pch.feedbackHeading

pch.feedbackSubheading