/*
 * [FindMissingRFCs.java]
 *
 * Summary: Find which RFCs do not exist on Randy's site.
 *
 * Copyright: (c) 2013-2017 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 2013-04-06 initial release
 */
package com.mindprod.repair;

import com.mindprod.common18.EIO;
import com.mindprod.csv.CSVWriter;
import com.mindprod.hunkio.HunkIO;

import java.io.File;
import java.io.IOException;
import java.io.PrintWriter;
import java.util.BitSet;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

/**
 * Find which RFCs do not exist on Randy's site.
 * <p/>
 * generates E:\mindprod\embellishment\missingrfcs.csv
 *
 * @author Roedy Green, Canadian Mind Products
 * @version 1.0 2013-04-06 initial release
 * @since 2013-04-06
 */
public class FindMissingRFCs
    {
    private static final int MAX_RFC = com.mindprod.htmlmacros.macro.RFC.MAX_RFC;

    private static final Pattern pattern = Pattern.compile(
            ">RFC (\\d+):<", Pattern.CASE_INSENSITIVE );

    /**
     * We screen scrape the website to get a list of RFC numbers supported.
     * Then we emit just a list of missing RFCs.
     *
     * @param args not used
     *
     * @throws java.io.IOException
     */
    public static void main( final String[] args ) throws IOException
        {
        // rem we have download http://openrfc.org/rfc-index.pl C:\temp\rfcindex.html
        final String contents = HunkIO.readEntireFile( new File( "C:/temp/rfcindex.html" ) );
        BitSet present = new BitSet( MAX_RFC + 1 );
        final Matcher m = pattern.matcher( contents );  // Matchers are used both for matching and finding.
        while ( m.find() )
            {
            present.set( Integer.parseInt( m.group( 1 ) ) );
            }
        // find all present. Presume everything else is missing.
        final File outFile = new File( "E:/com/mindprod/repair/missingrfcs.csv" );
        final PrintWriter prw = EIO.getPrintWriter( outFile, 24 * 1024, EIO.UTF8 );
        final CSVWriter w = new CSVWriter( prw, -1 /* quote level */, ',', '\"', '#', true );
        for ( int i = 0; i < MAX_RFC + 1; i++ )
            {
            if ( !present.get( i ) )
                {
                w.put( i );
                w.nl();
                }
            }
        w.close();
        }
    }