package at.gv.egovernment.moa.id.util;

/**
 * HTTP Utilities  
 * 
 * @author Rudolf Schamberger
 * @version $Id$
 */

import java.io.BufferedInputStream;
import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.Reader;
import java.net.HttpURLConnection;
import java.net.URL;

import org.apache.regexp.RE;
import org.apache.regexp.RESyntaxException;

/**
 * 
 * @author Rudolf Schamberger
 *
 */
public class HTTPUtils {

  /**
   * Utility used to obtainin correct encoded HTTP content.
   * Reads a given Content adressed by HTTP-URL into String. 
   * Content encoding is considered by using the Content-Type HTTP header charset value.
   * @param URL HTTP URL to read from.
   * @return String representation of content
   * @throws IOException on data-reading problems
   */
  public static String readHttpURL(String URL) 
    throws IOException {
      
    URL url = new URL(URL);
    HttpURLConnection conn = (HttpURLConnection)url.openConnection();
    conn.setRequestMethod("GET");
    String contentType = conn.getContentType();
    RE regExp = null;
    try {
        regExp = new RE("(;.*charset=)(\"*)(.*[^\"])");
    } catch (RESyntaxException e) {
        //RESyntaxException is not possible = expr. is costant
    }
    boolean charsetSupplied = regExp.match(contentType);
    String encoding = "ISO-8859-1"; 		//default HTTP encoding 
    if (charsetSupplied) {
      encoding = regExp.getParen(3);
    }
    InputStream instream = new BufferedInputStream(conn.getInputStream());
    InputStreamReader isr = new InputStreamReader(instream, encoding);
    Reader in = new BufferedReader(isr);
    int ch;
  	StringBuffer buffer = new StringBuffer();
    while ((ch = in.read()) > -1) {
  		buffer.append((char)ch);
  	}     
    in.close();
    conn.disconnect();
    return buffer.toString();
  }

}