"<h1>\u3000\u3000abc</h1>" -> "# \u3000\u3000abc" -> "<h1>abc</h1>"
"<h1> abc</h1>" -> "# \xa0\xa0abc" -> "<h1>abc</h1>"
because str.strip remove not only ascii whitespace like \r\n\t\x20, but also unicode space like    and so on.
str.strip(string.whitespace) can solve it.