public static String deRegularExpression(String content) {

content = deRegularScript(content); // 过滤script标签

String regEx_style = "

/* String regEx_inStyle = "style=\"([^\";]+;?)+\""; */

String regEx_html = "<[^>]+>"; // 定义HTML标签的正则表达式

Pattern p_style = Pattern.compile(regEx_style, Pattern.CASE_INSENSITIVE);

Matcher m_style = p_style.matcher(content);

content = m_style.replaceAll(""); // 过滤style标签

Pattern p_html = Pattern.compile(regEx_html, Pattern.CASE_INSENSITIVE);

Matcher m_html = p_html.matcher(content);

content = m_html.replaceAll(" "); // 过滤html标签

// 去除 

Pattern pattern = Pattern.compile(" ");

Matcher matcher = pattern.matcher(content);

content = matcher.replaceAll(" ");

return content.trim(); // 返回文本字符串

}

public static String deRegularScript(String content) {

String regEx_script = "

Pattern p_script = Pattern.compile(regEx_script, Pattern.CASE_INSENSITIVE);

Matcher m_script = p_script.matcher(content);

return m_script.replaceAll(""); // 过滤script标签

}