去除html标记与查询html中字条串函数

去除html标记

pattern pattern = pattern.compile("<.+?>", pattern.dotall);
matcher matcher = pattern.matcher("<a href="index.html">主页</a>");
string string = matcher.replaceall("");
system.out.println(string);

查找html中对应条件字符串

pattern pattern = pattern.compile("href="(.+?)"");
matcher matcher = pattern.matcher("<a href="index.html">主页</a>");
if(matcher.find())
system.out.println(matcher.group(1));
}