/*
 * [AccountForQuotes.java]
 *
 * Summary: When a quotation file split, checks no quotes were lost. Works best with StripGenerated files.
 *
 * Copyright: (c) 2015-2017 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 2015-12-30 initial version
 */
package com.mindprod.htmlmacros;

import com.mindprod.common18.FNV1a64;
import com.mindprod.entities.DeEntifyStrings;
import com.mindprod.hunkio.HunkIO;

import java.io.File;
import java.io.IOException;
import java.util.HashSet;
import java.util.regex.Pattern;

import static java.lang.System.*;

/**
 * When a quotation file split, checks no quotes were lost. Works best with StripGenerated files.
 *
 * @author Roedy Green, Canadian Mind Products
 * @version 1.0 2015-12-30 initial version
 * @since 2015-12-30
 */
public class AccountForQuotes
    {
    /**
     * what we look for surrounding text of interest
     */
    private static final String CLOSE_TAG = "</blockquote>";

    /**
     * what we look for surrounding text of interest
     */
    private static final String OPEN_TAG = "<blockquote";

    /**
     * track which quotes appear in new files.
     */
    private static final HashSet<Long> extantQuotes = new HashSet<>( 10000 );

    private static final Pattern SPLIT_ON_SPACE = Pattern.compile( "\\s+" );

    /**
     * read all quotes in new split up files.
     */
    private static void readNewQuotes( final String[] args )
        {
        for ( int i = 1; i < args.length; i++ )
            {
            final File file = new File( args[ i ] );
            try
                {
                final String big = HunkIO.readEntireFile( file );
                int start0 = 0;
                int start1;
                while ( ( start1 = big.indexOf( OPEN_TAG, start0 ) ) >= 0 )
                    {
                    final int start2 = start1 + OPEN_TAG.length();
                    final int end1 = big.indexOf( CLOSE_TAG, start2 );
                    if ( end1 < 0 )
                        {
                        throw new IllegalArgumentException( ">>>missing </blockquote>" );
                        }
                    final int end2 = end1 + CLOSE_TAG.length();
                    final String rawBlockQuote = big.substring( start1, end2 );
                    final String plainBlockQuote = DeEntifyStrings.flattenHTML( rawBlockQuote, ' ' );
                    // process text <blockquote .. </blockquote>
                    // break into words
                    final String[] words = SPLIT_ON_SPACE.split( plainBlockQuote );
                    long hash = FNV1a64.computeHash( words );
                    extantQuotes.add( hash );
                    start0 = end2;
                    } // end while
                }
            catch ( IOException e )
                {
                e.printStackTrace( err );
                err.println( "Fatal error: unable to open " + file + " possibly locked in the editor." );
                System.exit( 2 );
                }
            }
        }

    /**
     * read original quotes, all in one file, first on command line
     */
    private static void readOldQuotes( final String[] args )
        {
        int errorCount = 0;
        final File file = new File( args[ 0 ] );
        try
            {
            final String big = HunkIO.readEntireFile( file );
            int start0 = 0;
            int start1;
            while ( ( start1 = big.indexOf( OPEN_TAG, start0 ) ) >= 0 )
                {
                final int start2 = start1 + OPEN_TAG.length();
                final int end1 = big.indexOf( CLOSE_TAG, start2 );
                if ( end1 < 0 )
                    {
                    throw new IllegalArgumentException( ">>>missing </blockquote>" );
                    }
                final int end2 = end1 + CLOSE_TAG.length();
                final String rawBlockQuote = big.substring( start1, end2 );
                final String plainBlockQuote = DeEntifyStrings.flattenHTML( rawBlockQuote, ' ' );
                // process text <blockquote .. </blockquote>
                // break into words
                final String[] words = SPLIT_ON_SPACE.split( plainBlockQuote );
                long hash = FNV1a64.computeHash( words );
                if ( !extantQuotes.contains( hash ) )
                    {
                    errorCount++;
                    err.println( "Error: quote in " + file + " is missing from the new files" );
                    err.println( rawBlockQuote );
                    err.println();
                    }
                start0 = end2;
                } // end while
            }
        catch ( IOException e )
            {
            e.printStackTrace( err );
            err.println( "Fatal error: unable to open " + file + " possibly locked in the editor." );
            System.exit( 2 );
            }
        out.println( errorCount + " errors" );
        }

    /**
     * read all quotes and store checksum of those in new1 new2 ... then process old, and see if its checksumes exist
     *
     * @param args old  new1 new2 ...
     *
     * @throws IOException if cannot read quotations
     */
    public static void main( String[] args ) throws IOException
        {
        readNewQuotes( args );
        readOldQuotes( args );
        }
    }