🌐 Async Programming (asyncio)

asyncio provides concurrency via cooperative multitasking:

• code runs in a single OS thread
• tasks voluntarily yield control at await
• great for "lots of I/O" workloads (many sockets, many requests)

Async is not automatically faster—it’s a tool for scaling I/O concurrency while keeping code readable.

---

✅ Coroutines and await

import asyncio

async def say_after(delay: float, msg: str) -> None:
    await asyncio.sleep(delay)
    print(msg)

asyncio.run(say_after(1.0, "hello"))

---

✅ Tasks and gather()

• asyncio.create_task(...) schedules a coroutine.
• asyncio.gather(...) waits for many tasks and returns results.

---

✅ Limiting Concurrency

Even in async code you sometimes need to limit parallel work (e.g. 1000 requests at once is still too many). Use an asyncio.Semaphore.

---

🔍 Key Takeaways

• async/await is for high-concurrency I/O.
• await is where other tasks can run.
• Use tasks + gather() to run many coroutines.

---

Back: Module 12 README