我建议使用一个列表，弹出这个列表的第一个元素，如果这个元素是一个错误，那么把这个元素附加到同一个列表中。你只需要控制你允许的错误的数量。

import requests as r
import pandas as pd

max_error = 5

print("Start")

df_urls = pd.read_csv('./input/urls.csv')
print("DataFrame")
print(df_urls)

urls = list(df_urls.values)
print("Vector")
print(urls)

error_list = []
error_count = 0
while len(urls) >0 and error_count <= max_error:
    try:
        url = urls.pop(0)[0]
        x = r.get(url)
        print(x.status_code)
    except:
        urls.append(url)
        print("Error url:" + str(url))
        error_count += 1

print("Number of error: " + str(error_count))
print("Finish")

URL文件
网址
如何在Scrapy结束时运行代码或因出错而再次运行
如何在Scrapy结束时运行代码或因出错而再次运行
https://stackoverhflow.com/questions/9399900358/is-there-a-decent-alternative-to-
输出量：

Start
DataFrame
                                                urls
0  https://stackoverflow.com/questions/73003565/h...
1  https://stackoverflow.com/questions/73003565/h...
2  https://stackoverhflow.com/questions/939990035...
Vector
[array(['https://stackoverflow.com/questions/73003565/how-do-run-code-when-scrapy-ended-or-run-again-for-error'],
      dtype=object), array(['https://stackoverflow.com/questions/73003565/how-do-run-code-when-scrapy-ended-or-run-again-for-error'],
      dtype=object), array(['https://stackoverhflow.com/questions/9399900358/is-there-a-decent-alternative-to-dynamic-cast'],
      dtype=object)]
200
200
Error url:https://stackoverhflow.com/questions/9399900358/is-there-a-decent-alternative-to-dynamic-cast
Error url:h
Error url:h
Error url:h
Error url:h
Error url:h
Number of error: 6
Finish

Process finished with exit code 0