상세 컨텐츠

본문 제목

특정 url 소스 가져오기

Java

by 탑~! 2014. 9. 2. 17:58

본문

import java.io.*;

import java.net.URL;

import java.net.URLConnection;


import javax.swing.text.html.*;

import javax.swing.text.html.HTML.*;

import javax.swing.text.html.HTMLDocument;

import javax.swing.text.AttributeSet;

import javax.swing.text.Element;

import javax.swing.text.ElementIterator;

import javax.swing.text.StyleConstants;


import java.io.InputStreamReader;

import java.net.HttpURLConnection;


import javax.swing.text.BadLocationException;

import javax.swing.text.MutableAttributeSet;

import javax.swing.text.html.parser.ParserDelegator;


import java.util.regex.Matcher; 

import java.util.regex.Pattern; 


public class TestHtmlParse {


public static void main(String[] args) {

HttpURLConnection httpConn = null;  

String urlParameters=""; //파라메타값

String targetURL="http://www.nts.go.kr/cal/cal_check_02_01.asp?Source_Saup_No_2_2_9=2141154606&Target_Saup_No_2_2_9=1321765142";

try{

URL url = new URL(targetURL);

   httpConn = (HttpURLConnection)url.openConnection();

   

   //헤더 선언

   httpConn.setRequestMethod("POST");

   httpConn.setRequestProperty("Content-Type", "application/x-www-form-urlencoded");

   //httpConn.setRequestProperty("Content-Type", "application/x-www-form-urlencoded;charset=EUC-KR");

 //  httpConn.setRequestProperty("Cookie", "cookievalue="+ cookie);

   

   httpConn.setUseCaches (false);

   httpConn.setDoInput(true);

   httpConn.setDoOutput(true);    


   PrintWriter pw = new PrintWriter(new OutputStreamWriter(httpConn.getOutputStream()));

   //PrintWriter pw = new PrintWriter(new OutputStreamWriter(httpConn.getOutputStream(), "EUC-KR"));

   pw.write(urlParameters);

   pw.flush();

   pw.close();

   

//    OutputStream opstrm = httpConn.getOutputStream();

//    opstrm.write(param.getBytes());

//    opstrm.flush();

//    opstrm.close();

                      

   String line = null;

   BufferedReader in = new BufferedReader(new InputStreamReader(httpConn.getInputStream(), "euc-kr"));

   while ((line = in.readLine()) != null) {

    //<b style="color:blue">부가가치세 일반과세자</b>

Pattern pattern  =  Pattern.compile("<b[^*]style=\"color:blue\">*");

   Matcher m = pattern.matcher(line.trim());

   

   String special_str = "";

   while (m.find()) {

    special_str = m.group();

    System.out.println(line);

    System.out.println(special_str);

   }

   }

   in.close();

   

   


   

} catch (IOException ioe) {

ioe.printStackTrace();

}

}

}



'Java' 카테고리의 다른 글

Expression Language  (0) 2014.12.10
JSTL  (0) 2014.12.10
java document를 chm 파일로 다운로드  (0) 2014.06.10
java에서 sftp 연결하기  (0) 2014.06.10
MSSQL jdbc - sqljdbc.jar와 sqljdbc4.jar 차이  (0) 2013.03.19

관련글 더보기