HTML过滤

实用

public static String removeTag(String htmlStr) {
    try {
        String regEx_script = "<script[^>]*?>[\\s\\S]*?<\\/script>"; // script
        String regEx_style = "<style[^>]*?>[\\s\\S]*?<\\/style>"; // style
        String regEx_html = "<[^>]+>"; // HTML tag
        String regEx_space = "\\s+|\t|\r|\n|&nbsp;|\\[0-9+\\]";// other characters

        Pattern p_script = Pattern.compile(regEx_script, Pattern.CASE_INSENSITIVE);
        Matcher m_script = p_script.matcher(htmlStr);
        htmlStr = m_script.replaceAll("");
        Pattern p_style = Pattern.compile(regEx_style, Pattern.CASE_INSENSITIVE);
        Matcher m_style = p_style.matcher(htmlStr);
        htmlStr = m_style.replaceAll("");
        Pattern p_html = Pattern.compile(regEx_html, Pattern.CASE_INSENSITIVE);
        Matcher m_html = p_html.matcher(htmlStr);
        htmlStr = m_html.replaceAll("");
        Pattern p_space = Pattern.compile(regEx_space, Pattern.CASE_INSENSITIVE);
        Matcher m_space = p_space.matcher(htmlStr);
        htmlStr = m_space.replaceAll(" ");
        return htmlStr;
    }catch (NullPointerException nullException){
        return "";
    }
}

转载于:https://my.oschina.net/vintnee/blog/707721