从 python re docs ， \w 匹配任何字母数字字符和下划线，相当于集合 [a-zA-Z0-9 _] 。因此 [\w\ .-] 将适当地匹配数字和字符。

From the python re docs, \w matches any alphanumeric character and underscores, equivalent to the set [a-zA-Z0-9_]. So [\w\.-] will appropriately match numbers as well as characters.

email = re.findall(r'[\w\.-]+@[\w\.-]+(\.[\w]+)+',line)

这篇文章更加广泛地讨论了匹配的电子邮件地址，您遇到了很多陷阱，导致您的代码无法捕获匹配的电子邮件地址。例如，电子邮件地址不能完全由标点符号组成（ ... @ .... ）。此外，地址通常有最大长度，具体取决于电子邮件服务器。此外，许多电子邮件服务器都匹配非英语字符。因此，根据您的需求，您可能需要一个更全面的模式。

This post discusses matching email addresses much more extensively, and there are a couple more pitfalls you run into matching email addresses that your code fails to catch. For example, email addresses cannot be made up entirely of punctuation (...@....). Additionally, there is often a maximum length on addresses, depending on the email server. Also, many email servers match non-english characters. So depending on your needs you may need a more comprehensive pattern.