/*
 * [FindRecentlyDied.java]
 *
 * Summary: Find authors who have recently died.
 *
 * Copyright: (c) 2012-2017 Roedy Green, Canadian Mind Products, http://mindprod.com
 *
 * Licence: This software may be copied and used freely for any purpose but military.
 *          http://mindprod.com/contact/nonmil.html
 *
 * Requires: JDK 1.8+
 *
 * Created with: JetBrains IntelliJ IDEA IDE http://www.jetbrains.com/idea/
 *
 * Version History:
 *  1.0 2012-12-23 initial release
 */
package com.mindprod.repair;

import com.mindprod.common18.EIO;
import com.mindprod.csv.CSVReader;
import com.mindprod.entities.DeEntifyStrings;
import com.mindprod.http.Get;

import java.io.BufferedReader;
import java.io.EOFException;
import java.io.File;
import java.io.IOException;
import java.util.concurrent.TimeUnit;

import static java.lang.System.*;

/**
 * Find authors who have recently died.
 * <p/>
 * This does not work. Google turns off service.
 *
 * @author Roedy Green, Canadian Mind Products
 * @version 1.0 2012-12-23 initial release
 * @since 2012-12-23
 */
public class FindRecentlyDied
    {
    public static void main( final String[] args ) throws IOException
        {
        final BufferedReader br = EIO.getBufferedReader( new File( "E:/com/mindprod/repair/authors.csv" ), 64 * 1204, EIO.UTF8 );
        final CSVReader r = new CSVReader( br, ',', '\"', "#", true /* hide comments */, true /* trim */,
                true /* trim */, true /* multiline */ );
        try
            {
            while ( true )
                {
                final String[] fields = r.getAllFieldsInLine();
                if ( fields.length != 2 )
                    {
                    continue;
                    }
                final String author = DeEntifyStrings.deEntifyHTML( fields[ 0 ], ' ' );
                final String born = fields[ 1 ];
                // do GET query like this to google.
                // GET /search?q=%22Mitch+Kapor%22+born+1950-11-01+died
                final Get get = new Get();
                get.setInstanceFollowRedirects( true );
                get.setConnectTimeout( ( int ) TimeUnit.SECONDS.toMillis( 40 ) );
                get.setParms( "q", "\"" + author + "\" born " + born );
                final String response = get.send( "www.google.ca", -1, "/search", Get.UTF8 );
                int responseCode = get.getResponseCode();
                if ( responseCode != 200 || response == null || response.length() == 0 )
                    {
                    err.println( responseCode + " " + get.getResponseMessage() + " no response for " + author );
                    continue;
                    }
                final String lc = response.toLowerCase()
                        .replace( '\n', ' ' )
                        .replace( '\t', ' ' )
                        .replace( '<', ' ' )
                        .replace( '>', ' ' )
                        .replace( '.', ' ' )
                        .replace( ',', ' ' );
                final int where = lc.indexOf( " died " );
                if ( where >= 0 )
                    {
                    out.println( "-------------------------------------" );
                    out.println( author + " " + born );
                    final int start = Math.max( 0, where - 100 );
                    final int end = Math.min( where + 100, response.length() );
                    out.println( response.substring( start, where ) );
                    out.println( response.substring( where, end ) );
                    out.println( "-------------------------------------" );
                    }
                else
                    {
                    out.println( author + " " + born + " is still alive." );
                    }
                }
            }
        catch ( EOFException e )
            {
            }
        }
    }