做个笔记免得每次都去找,重新写
public static void main(String[] args) throws IOException{
Pattern pt = Pattern.compile("<div id=\"article\">([\\s\\S]*?)<div class=\"clear\">");
String web = readFileByLines("C:\\Users\\tianjun\\Desktop\\search\\ch10\\heritrixproject\\jobs\\xiaoxue-20150314050416274\\mirror\\xiaoxue.hujiang.com\\yi\\shuxue\\p324532\\index.html");
web = new String(web.getBytes("gbk"),"utf-8");
Matcher mc =pt.matcher(web);
if(mc.find()){
String content = mc.group(1).trim();
System.out.println(content);
}
}
public static String readFileByLines(String fileName) {
File file = new File(fileName);
StringBuffer sb=new StringBuffer();
BufferedReader reader = null;
try {
System.out.println("==========================:");
reader = new BufferedReader(new FileReader(file));
String tempString = null;
while ((tempString = reader.readLine()) != null) {
sb.append(tempString);
sb.append("\n");
}
reader.close();
} catch (IOException e) {
e.printStackTrace();
} finally {
if (reader != null) {
try {
reader.close();
} catch (IOException e1) {
}
}
}
return sb.toString();
}