今天來記錄一下 Python 平行程式的用法
Python 有三種方式可以平行執行程式
- multiprocess
- threading
- asyncio
以下分別是三種方式的範例程式碼
multiprocess
import multiprocessing
import os
import time
def compute_square(number):
process_name = multiprocessing.current_process().name
process_id = os.getpid()
print(
f"Process {process_name} (process id: {process_id}) computing square of {number}"
)
time.sleep(1)
print(f"Result: {number}^2 = {number * number}")
if __name__ == "__main__":
numbers = [2, 4, 6, 8]
processes = []
for num in numbers:
p = multiprocessing.Process(target=compute_square, args=(num,))
processes.append(p)
p.start()
for p in processes:
p.join()
print("All processes finished.")
threading
import os
import threading
import time
def compute_square(number):
thread_name = threading.current_thread().name
thread_id = threading.get_ident()
process_id = os.getpid()
print(
f"Thread {thread_name} (thread id: {thread_id}, process id: {process_id}) computing square of {number}"
)
time.sleep(1)
print(f"Result: {number}^2 = {number * number}")
numbers = [2, 4, 6, 8]
threads = []
for num in numbers:
t = threading.Thread(target=compute_square, args=(num,))
threads.append(t)
t.start()
for t in threads:
t.join()
print("All threads finished.")
asyncio
import asyncio
import os
import threading
async def compute_square(number):
task = asyncio.current_task()
task_id = id(task)
thread_id = threading.get_ident()
process_id = os.getpid()
print(
f"Asyncio Task (id: {task_id}) running in thread id: {thread_id}, process id: {process_id}, computing square of {number}"
)
await asyncio.sleep(1)
print(f"Result: {number}^2 = {number * number}")
async def main():
numbers = [2, 4, 6, 8]
tasks = [compute_square(num) for num in numbers]
await asyncio.gather(*tasks)
print("All asyncio tasks finished.")
if __name__ == "__main__":
asyncio.run(main())
從這三個範例可以看到執行的地方不同：
- multiprocessing 是在不同的 process 執行的。
- threading 是在同一個 process 的不同 thread 執行的。
- asyncio 是在同一個 process 的同一個 thread 執行的。
這三個範例都是簡單的執行一個單純的 task ，所以可以簡單的平行處理，如果 task 之間有順序性或是需要交換資料，都會讓平行處理的複雜度上升。