为什么我的线程不能让我的脚本运行时间更短？

Question

import re
import concurrent.futures
import time

setting1 = 101
setting2 = 321
setting3 = None

threads = 1000

start = time.perf_counter()

def load_keywords(keyword_file):
    with open(keyword_file, 'r') as keywordhandle:
        keywords = [keyword.strip() for keyword in keywordhandle.readlines()]
    return keywords

def gencombo(user, keywords):
    og_user = re.sub(r'\d', '', user)
    user_lower = og_user.lower()
    combos = []
    n_keywords = []

    for keyword in keywords:
        if keyword.lower() in user_lower:
            n_keywords.append(keyword)

    for k1 in n_keywords:
        k1 = str(k1)
        for k2 in n_keywords:
            k2 = str(k2)
            if k2 in k1 and k2 is not k1:
                n_keywords.remove(k2)

    n_keywords.reverse()

    for k1 in n_keywords:
        k1 = str(k1)
        for k2 in n_keywords:
            k2 = str(k2)
            if k2 in k1 and k2 is not k1:
                n_keywords.remove(k2)

    for keyword in n_keywords:
        if keyword.lower() in user_lower:
            print(f"Combos added for: {user}")
            if setting1 is not None:
                combos.append(f"{user}:{keyword}{str(setting1)}")
            if setting2 is not None:
                combos.append(f"{user}:{keyword}{str(setting2)}")
            if setting3 is not None:
                combos.append(f"{user}:{keyword}{str(setting3)}")


    return '\n'.join(combos) if combos else None

def gencombos(input_file, output_file, keyword_file):
    keywords = load_keywords(keyword_file)
    with open(input_file, 'r') as combohandle:
        with open(output_file, 'w') as handlenone:
            def threadrun():
                for user in combohandle:
                    combo = gencombo(user.strip(), keywords)
                    if combo:
                        handlenone.write(f"{combo}\n")
            with concurrent.futures.ThreadPoolExecutor(max_workers=threads) as executor:
                executor.submit(threadrun)

gencombos("users.txt", "output.txt", "keywords.txt")

finish = time.perf_counter()

input(f"Finished in {round(finish - start, 2)} seconds.")

我首先尝试单独使用线程，但后来我发现并发.futures 有助于运行某些函数。因此，我转向了这一点。然而，即使在切换之后，我发现配置的线程数并不重要。它仍然大约在同一时间运行。我希望你们中的一位能帮助我，因为我不明白。

Answer 1

我建议您在尝试学习该语言时采用更复杂的路线，但是 Python 中的线程，尤其是对于如此简单的东西，几乎永远不会看到性能的飞跃。事实上，有时当您使用正则表达式和循环列表等受 CPU 限制的任务时，由于解释器的工作方式，线程实际上会减慢程序速度。

1000个线程开销也不少了

从纯粹效率的角度来看，您可能会发现在这样的脚本中使用多处理而不是多线程更成功。