source: DRBLTranslator/src/DRBLTranslator.java @ 79

import java.io.BufferedReader;
import java.io.BufferedWriter;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.apache.commons.httpclient.HttpClient;
import org.apache.commons.httpclient.HttpException;
import org.apache.commons.httpclient.methods.PostMethod;


public class DRBLTranslator {
  
  ArrayList<String> lines;
  ArrayList<Boolean> IsTranslated;
  ArrayList<Boolean> HasBeginQuote;
  ArrayList<Boolean> HasEndQuote;
  ArrayList<String> lstrs;
  ArrayList<String> rstrs;
  ArrayList<String> quotes;
  ArrayList<String> tstrs;
  ArrayList<String> ends;
  String output;
  
  public DRBLTranslator() {
    lines = new ArrayList<String>();
    IsTranslated = new ArrayList<Boolean>();
    HasBeginQuote = new ArrayList<Boolean>();
    HasEndQuote = new ArrayList<Boolean>();
    lstrs = new ArrayList<String>();
    rstrs = new ArrayList<String>();
    quotes = new ArrayList<String>();
    tstrs = new ArrayList<String>();
    ends = new ArrayList<String>();
    output = "";
  }

  public void reset() {
    lines.clear();
    IsTranslated.clear();
    HasBeginQuote.clear();
    HasEndQuote.clear();
    lstrs.clear();
    rstrs.clear();
    quotes.clear();
    tstrs.clear();
    ends.clear();
    output = "";
  }
  
  private static String googleTranslate(String langpair, String text) {

    /***********************************************************************
     * commons-logging commons-codec
     */
    HttpClient client = new HttpClient();

    /***********************************************************
     */
    client.getHostConfiguration().setHost("www.google.com", 80, "http");

    /***********************************************************
     */
    PostMethod post = new PostMethod("/translate_t");
    // post.addParameter(new NameValuePair("url", url));
    post.addParameter("langpair", langpair);
    post.addParameter("text", text);

    String s = null;
    try {
      /***********************************************************
       */
      client.executeMethod(post);

      /***********************************************************
       */
      InputStream in = post.getResponseBodyAsStream();
      BufferedReader buf = new BufferedReader(new InputStreamReader(in));
      Pattern p = Pattern
          .compile(".*id=\"*result_box\"*\\ *dir=\"*ltr\"*\\>");
      Matcher m = null;

      String line;
      while ((line = buf.readLine()) != null) {
        m = p.matcher(line);
        // System.out.println(line);
        if (m.find()) {
          s = line.substring(m.end());
          line = s;
          p = Pattern.compile("\\<\\ */\\ *div\\ *\\>");
          m = p.matcher(line);
          if (m.find()) {
            s = line.substring(0, m.start());
          }
          break;
        }
      }

      // s = GetTranslation.instostr(in);
    } catch (HttpException e) {
      e.printStackTrace();
    } catch (IOException e) {
      e.printStackTrace();
    }
    return s;
  }

  public static String instostr(InputStream in) throws IOException {
    StringBuffer out = new StringBuffer();
    byte[] b = new byte[4096];
    for (int n; (n = in.read(b)) != -1;) {
      out.append(new String(b, 0, n));
    }
    return out.toString();
  }
        
  public void dump() {
    for (int i=0; i<lines.size(); i++) {
      System.out.println("======== Line "+i+" ========");
      System.out.println(lines.get(i));
      System.out.println(IsTranslated.get(i));
      System.out.println(HasBeginQuote.get(i));
      System.out.println(HasEndQuote.get(i));
      System.out.println(lstrs.get(i));
      System.out.println(rstrs.get(i));
      System.out.println(tstrs.get(i));
      System.out.println(quotes.get(i));
      System.out.println(ends.get(i));
      System.out.println("======== Line "+i+" ========");
      System.out.println();
    }
  }
  
  public String toConfiguration() {
    output="";
    for (int i=0; i<lines.size(); i++) {
      
      if (IsTranslated.get(i)) {
        //strlines += lstrs.get(i) + "=" + quotes.get(i) + tstrs.get(i) + quotes.get(i);
        output += lstrs.get(i);
        if (HasBeginQuote.get(i)) {
          output += "=" + quotes.get(i);
        }
        output += tstrs.get(i);
        if (HasEndQuote.get(i)) {
          output += quotes.get(i);
        }
        output += ends.get(i);
        output += "\n";
      } else {
        output += lines.get(i) + "\n";
      }           
    }     
    return output;
  }

  public String getConfiguration() {
    return output;
  }
  public void loadFile(String filepath) {
    reset();
    BufferedReader reader = null;
    try {
      reader = new BufferedReader(new FileReader(filepath));      
    } catch (FileNotFoundException e) {
      e.printStackTrace();
    } catch (IOException e) {
      e.printStackTrace();
    }   
    
    int count = 0, index = 0;
    String line = null, lstr = null, rstr = null;
    Boolean hasEndQuote;
    try {     
      while ((line = reader.readLine()) != null) {
        line = line.trim();
        lines.add(line);
        if (line.contains("=")) {                    
          count++;
          index = line.indexOf("=");
          lstr = line.substring(0, index);
          rstr = line.substring(index + 2, line.length()-1);
          HasBeginQuote.add(true);
          if (line.charAt(index+1)!=line.charAt(line.length()-1)) {
            
            if (line.charAt(index+1)==line.charAt(line.length()-2)) {
              ends.add(String.valueOf(line.charAt(line.length()-1)));
              rstr = line.substring(index + 2, line.length()-2);
              HasEndQuote.add(true);
            } else {
              ends.add("");
              HasEndQuote.add(false);
            }
          } else {
            HasEndQuote.add(true);
            ends.add("");
          }
          lstrs.add(lstr);
          rstrs.add(rstr);
          IsTranslated.add(true);
          quotes.add(String.valueOf(line.charAt(index+1)));
        } else if (line.startsWith("#") || line.equals("")){
          IsTranslated.add(false);
          lstrs.add("");
          rstrs.add("");
          quotes.add("");
          HasBeginQuote.add(false);
          HasEndQuote.add(false);
          ends.add("");
        } else {
          hasEndQuote=false;
          IsTranslated.add(true);
          lstrs.add("");
          int lidx=0, ridx=0;
          char quote=' ';
          if (line.startsWith("'") || line.startsWith("\"")) {
            HasBeginQuote.add(true);
            lidx=1;
          } else {
            HasBeginQuote.add(false);
            lidx=0;
          }
          if (line.endsWith("'") || line.endsWith("\"")) {
            HasEndQuote.add(true);
            hasEndQuote=true;
            ridx=line.length()-1;
          } else if (line.charAt(line.length()-2)=='\'' || line.charAt(line.length()-2)=='\"'){
            HasEndQuote.add(true);
            hasEndQuote=true;
            ridx=line.length()-2;
          } else {
            HasEndQuote.add(false);
            ridx=line.length();
          }
          if (hasEndQuote) {
            if (line.endsWith("'") || line.endsWith("\"")) {
                quote = line.charAt(line.length()-1);
                quotes.add(String.valueOf(quote));
            } else {
                quote = line.charAt(line.length()-2);
                quotes.add(String.valueOf(quote));              
            }    
          } else {
            quote = line.charAt(0);
            quotes.add(String.valueOf(quote));
          }
          
          if (hasEndQuote) { 
              if (quote!=line.charAt(line.length()-1)) {
                ends.add(String.valueOf(line.charAt(line.length()-1)));
              } else
                ends.add("");
          } else
            ends.add("");
          rstrs.add(line.substring(lidx,ridx));         
        }
      }
      reader.close();
    } catch (IOException e) {
      e.printStackTrace();
    }   
  }
  
  public void translate(String langpair) {                
    String str = "", tstr="";       
    for (int i=0; i<rstrs.size(); i++){
      str += rstrs.get(i) + "\n";     
      
      if (i==0 && rstrs.size()>1)
        tstrs.add("");
      if ((i!=0 || (i==0 && rstrs.size()==1)) && (((i%300)==0) || i==(rstrs.size()-1)) ) {        
        //System.out.println("i="+i);
        //System.out.println("lstr="+lstrs.get(i));
          tstr = DRBLTranslator.googleTranslate(langpair, str).trim()
                                     .replaceAll("\\s*&quot;\\s*", "\"")
                                     .replaceAll("\\s*&gt;\\s*", ">")
                                     .replaceAll("\\s*&#39;\\s*", "'");

          str = "";         
          tstrs.addAll(Arrays.asList(tstr.split("\\s*<br>\\s*")));
          //System.out.println("tstr="+tstrs.get(i));         
          try {
          Thread.sleep(3000);
        } catch (InterruptedException e) {          
          e.printStackTrace();
        }
      }
    }
  }
  
  public void toFile(String out) {
    
    try {     
      BufferedWriter writer = new BufferedWriter(new FileWriter(out));                
      writer.write(getConfiguration());
      writer.flush();
      writer.close();
    } catch (FileNotFoundException e) {
      e.printStackTrace();
    } catch (IOException e) {
      e.printStackTrace();
    }   
  
  }

  public void srcToFile(String langpair, String in, String out) {
    loadFile(in);
    translate(langpair);
    toConfiguration();
    toFile(out);
  }
  public static void main(String[] args) throws HttpException, IOException {

    if (null == args || args.length < 3) {
      System.out
          .println("Usage: java DRBLTranslator (language_pair) (source_file) (target_file)");
      System.out
          .println("Example: java DRBLTranslator \"en|zh-TW\" \"en_US\" \"zh_TW\"");
      System.exit(1);
    }
        
    DRBLTranslator trans = new DRBLTranslator();    
    trans.srcToFile(args[0], args[1], args[2]);
    //trans.srcToFile("en|zh-TW", "en_US", "zh_TW");
    System.out.println("ok");
    //System.out.println(trans.toConfiguration());
    //trans.dump();

  }
}