/* * Created on Jul 6, 2006
*
* TODO To change the template for this generated file go to
* Window - Preferences - Java - Code Style - Code Templates
*/

/**
 * @author Betsy Rolland
 *
 * TODO To change the template for this generated type comment go to
 * Window - Preferences - Java - Code Style - Code Templates
 */

package org.biomediator.wrappers;

import java.io.BufferedReader;
import java.io.InputStreamReader;
import java.io.PrintWriter;
import java.io.*;
import java.net.URL;
import java.util.HashMap;
import java.util.Iterator;
import java.util.TreeSet;
import java.util.*;

import org.biomediator.util.AttribOption;
import org.biomediator.util.StringUtility;
import org.biomediator.xml.XMLParser;
import org.biomediator.xml.XMLAttributeParser;


import com.oroinc.text.regex.Perl5StreamInput;

import java.io.FileReader;

import org.xml.sax.XMLReader;
import org.xml.sax.Attributes;
import org.xml.sax.InputSource;
import org.xml.sax.helpers.XMLReaderFactory;
import org.xml.sax.helpers.DefaultHandler;

public class InterProWrapper extends AbstractWrapper {

	public static final String IP_HELP_TEXT =
		"        Description:\n" +
        "                Provides an interface for searching the InterPro database.\n" +
        "\n" +
        "        Usage:\n" +
        "        Servlet examples:\n" +
	   "				 ?term=term[tvalue=IPR010405,hexencoded=false,field=AccNumber]" +
	   "				 ?term=term[tvalue=ipr000004,hexencoded=false,field=SecAcc]" +
	   "				 ?term=term[tvalue=BRCA1*,hexencoded=false,field=AllText]" ;
	
	public static final String IP_SOURCE_TAG = "interpro_source";
	public static final String IP_CMD_TERM = "term";
	public static final String IP_OPTION_MAP_FIELD = "field";
	public static final String IP_OPTION_MAP_TVALUE = "tvalue";
    public static final String IP_OPTION_MAP_HEXENCODED = "hexencoded";
	public static final int MAX_DOCSIZE = 1024 * 10000;

	
	   
    public String proxyUrl = null;
	
    public static void main(String[] args) {

    	InterProWrapper ip = new InterProWrapper();    			//invoking the constructor
        ip.out = new PrintWriter(System.out, true);
        if (args.length > 0) {
            ip.processRequest(args[0]);
        }
        else {
            ip.processRequest("");
        }
    }
    
    public void processRequest(String query) {
        
        // Make sure the servlet is initialized.
        initializeServlet(IP_HELP_TEXT, IP_SOURCE_TAG, getWrapperProperties("interpro"));
        String data2 = "";
        		
        // Parse the query string and process commands.
        if (query == null || query.equals("")) {
            query = CMD_HELP;
        }
        AttribOption ao = new AttribOption(query);
        processCommands(ao, getWrapperProperties("interpro"));
        
        // Check to see if we should exit immeditely.
        if (isExitCmd) {
            return;
        }
        
        String proxy = wrapperProperties.getProperty(WrapperProperties.PROP_PROXY_URL);
        if (proxy == null) {
            proxyUrl = "";
        }
        else {
            proxyUrl = proxy + "?";
        }
    	
        String convertedQuery = null;
        TreeSet termTypeSet = ao.getAttribMapSet(IP_CMD_TERM);
        if (termTypeSet != null) {
        	convertedQuery = constructIPQuery(termTypeSet.iterator());
        }
        
        String urlBase = "http://srs.ebi.ac.uk/srsbin/cgi-bin/wgetz?[INTERPRO-";
        String view = "]+-view+InterProPrintXML+-mime+text/xml";
       
        
        try {
            // Download XML page.
        	
        	String queryURL = proxyUrl + urlBase + convertedQuery + view;
            //out.println("queryURL is " + queryURL);
        	InputStream inStream = new BufferedInputStream(new URL(queryURL).openStream());
        	BufferedReader br = new BufferedReader(new InputStreamReader(inStream));
        	String line = new String();
        	StringBuffer lineBf = new StringBuffer();
        	
        	while ((line = br.readLine()) != null)
        		lineBf.append(line + " ");

        
        	String newString = lineBf.toString();
        	
        	//out.println("newString is " + newString);
        	newString = newString.replaceAll("<p>", "");
        	newString = newString.replaceAll("\\.</p>", "");
        	newString = newString.replaceAll("<cite idref[^>]*>", "");
        	//out.println("newString is " + newString);
        	
//        	<cite idref="PUB00014805" /> 
        	
        	ByteArrayInputStream newStream = new ByteArrayInputStream(newString.getBytes());
        	
            //BufferedReader br = new BufferedReader(new InputStreamReader(new URL(queryURL).openStream()));
            StringBuffer lineBuf = new StringBuffer();
            lineBuf.append(HEADER_XML + "\n");
            

            XMLAttributeParser xap = new XMLAttributeParser(newStream, lineBuf);
            lineBuf = xap.parseData();
            
            
            //String line = null;
            
            /*while ((line = br.readLine()) != null) {
                lineBuf.append(line + "\n");
                if (lineBuf.length() > MAX_DOCSIZE) {
                    break;
                }
            }
            br.close();*/
       
            // Isolate and process the portion of the page that contains data. 
            String data = lineBuf.toString();

            // Convert retrieved document to valid XML (order of replacements is important).
            data = data.replaceAll("<\\?xml version=\"1\\.0\"[^>]*>", HEADER_XML);
            data = data.replaceAll("<!DOCTYPE[^>]*>", "");
            data = data.replaceAll("InterProEntrySet",IP_SOURCE_TAG);
            data = data.replaceAll("/InterProEntrySet",IP_SOURCE_TAG);
         
            
 
            // Print XML output to the client.
            out.print((data != null) ? data : "");
            out.flush();
    
            
        } //end of try
        	catch (Exception e) {
            System.err.println(e);
        }
    } //end of processRequest
    
    
    public String constructIPQuery(Iterator queryIterator) {

        StringBuffer termBuf = new StringBuffer("");
        HashMap groupMap = new HashMap();
        String ttype = null;
        String tvalue = null;
        String field = null;
        String hexencoded = null;
        String retQuery = null;
        
        if (queryIterator.hasNext()) {
            while (queryIterator.hasNext()) {
                // Retrieve the term values that were parsed from the query URL.
                HashMap queryMap = (HashMap) queryIterator.next();
                ttype = (String) queryMap.get(AttribOption.OPTION_PARAM_NAME);
                tvalue = (String) queryMap.get(IP_OPTION_MAP_TVALUE);
                field = (String) queryMap.get(IP_OPTION_MAP_FIELD);

                
                hexencoded = (String) queryMap.get(IP_OPTION_MAP_HEXENCODED);
                retQuery = field + ":" + tvalue;

                // Convert term values from hex to ascii if necessary.
                if (hexencoded != null && hexencoded.equals("true")) {
                	retQuery = StringUtility.convertHexToASCII(retQuery);
                }
           }
        }

        return retQuery;
    }
    
 
        
    public void terminate() {
        // TODO Auto-generated method stub

    }

}