这对于示例是有效的。可能漏掉了一些边缘案例...

import re

s='"你好" means "Hello" and "再见" is "See you".hello_me bye-you what6 58then end_of_this'
print(re.findall(r'(?:\b|[^A-Za-z]|(?<=_))([a-z]+)(?:\b|[^A-Za-z]|(?=_))', s))

输出：

['means', 'and', 'is', 'you', 'hello', 'me', 'bye', 'you', 'what', 'then', 'end', 'of', 'this']

这似乎并不难。
之前已经描述了用(?<![a-zA-Z])表示的自定义单词边界
(?![a-zA-Z])之后
(?<![a-zA-Z])[a-z]+(?![a-zA-Z])
代码样本

import re

s='"你好" means "Hello" and "再见" is "See you".hello_me bye-you what6 58then end_of_this'
print(re.findall(r'(?<![a-zA-Z])[a-z]+(?![a-zA-Z])', s))

输出量

['means', 'and', 'is', 'you', 'hello', 'me', 'bye', 'you', 'what', 'then', 'end', 'of', 'this']