#!/usr/bin/perl -w

######################################################################
###
###   IMPORTANT!
###   Please read the warranty and legal notice 
###   at the end of this file!
###
######################################################################

require 5.000;
use lib '/usr/local/bin/',"$ENV{HOME}/bin/",'/usr/stud/loescher/bin/';
use slutil;
use English;
use File::Basename;


######################################################################
### Voreinstellungen
######################################################################

$version = '1.0';

######################################################################
### Hauptprogramm
######################################################################

&Hilfe if ($#ARGV<0);

$doDelete = $FALSE;
if ($ARGV[0] =~ /-del/)
{
  $doDelete = $TRUE;
  shift;
}

open(QUELLE, "$ARGV[0]") || die "Kann Quelle nicht lesen!\n";
$i=0;
# Einlesen in Pärchen
while(<QUELLE>)
{
  /(.+?)\s+(\.*\/.+?)\s+\(identical\).*/;
  if (defined $2)
  {
   $quelle[$i] = [ $1 , $2 ];
   $i++;
  }
}
close QUELLE;

# Initialisieren mit dem ersten Wert
$ident[0] = [ $quelle[0][0], $quelle[0][1] ];

# Alle Quell-Zeilen durchlaufen
for($i=1; $i<=$#quelle; $i++)
{

  # Suchen nach bereits vorhandenen Einträgen in der Liste der Doppelten
  foreach $elem (@ident)
  {
    $such = quotemeta($quelle[$i][0]);
    $erstes  = grep(/$such/, @$elem);
    $such = quotemeta($quelle[$i][1]);
    $zweites = grep(/$such/, @$elem);
    # Es existieren schon beide in der Liste
    if ( $erstes && $zweites )
    {
      goto weiter;
    }
    # Wenn nur erstes existiert, dann zweites in die Liste aufnehmen
    if (   $erstes && ! $zweites )
    {
      push @$elem, $quelle[$i][1];
      goto weiter;
    }
    # Wenn nur zweites existiert, dann erstes in die Liste aufnehmen
    if ( ! $erstes &&   $zweites )
    {
      push @$elem, $quelle[$i][0];
      goto weiter;
    }
  }
  # Keines der beiden existiert in der Liste => neue Zeile anhängen
  push @ident, [ $quelle[$i][0], $quelle[$i][1] ];
 weiter:
}

foreach $ref (@ident)
{
  @$ref = &ElementeMitMeistenSonderzeichen(@$ref);
  unlink @$ref if ($doDelete);
  print "@$ref\n";
}


######################################################################
### Unterprogramme
######################################################################

sub ElementeMitMeistenSonderzeichen
{
  my $eins = shift;
  my $zwei;
  my @del  = ();
  while (@_)
  {
    $zwei = shift;
    if (&Sonder($eins) >= &Sonder($zwei))
    { 
      push @del,$eins;
      $eins = $zwei;
      next;
    }
    if (&Sonder($eins) < &Sonder($zwei))
    { 
      push @del,$zwei;
      next;
    }
  }
  return @del;
}


sub Sonder
{
  my $input = shift;
  my ($name,$path,$suffix);
  ($name,$path,$suffix) = fileparse($input,'');
  my $count = ($name =~ tr/A-Za-z0-9._-//cs)*10 
              + ($name =~ tr/0-9//s)*2 - length($name);
#  print "NAME: $name   COUNT: $count\n";
  return $count;
}


sub Hilfe
{
  die "
DelDoppelteFinddupe
~~~~~~~~~~~~~~~~~~~

Syntax: deldoppeltefinddupe [-del] finddupe-file
(Wenn -del angegeben ist wird tatsaechlich gelöscht, sonst wird nur eine
Liste ausgegeben.)

Löschen von doppelten Files, die Finddupe gefunden hat

Eingabe ist eine Datei der Form:
/tmp/ejccfnal.bat           /tmp/ejccaabl.bat           (identical) (25)
/tmp/ejccfnal.bat           /tmp/ejccjabl.bat           (identical) (25)
/tmp/ejccfnal.bat           /tmp/ejccgbbl.bat           (identical) (25)
/tmp/ejccloal.bat           /tmp/ejccaabl.bat           (identical) (25)
(wie von finddupe -vv erzeugt)

Voraussetzung: 
Dateiname dürfen auch Leerzeichen enthalten, aber nicht damit beginnen
oder enden.

";

}


######################################################################
#
# Warranty and legal notice
# ~~~~~~~~~~~~~~~~~~~~~~~~~
#
# Copyright (c) 1997 by Stephan Löscher  -  all rights reserved
# My Address: Stephan Löscher, Dr.Troll-str. 3, 82194 Gröbenzell, Germany
# Email: loescher@gmx.de
# WWW: http://www.loescher-online.de/
#
# This program is freeware.
# It is NOT Public-Domain-Software!
# The author (Stephan Löscher) does NOT give up his copyright, but he 
# reserves his copyright. Usage and copying is free of charge for private
# use, but NOT for commercial use!
# 
# You may and should copy this program free of charge, use it,
# give it to your friends, upload it to a BBS or something similar, under
# the following conditions:
# * Don't charge any money for it. If you upload it to a BBS, make sure that
#    it can be downloaded free (without paying for downloading it, except
#    for usage fees that have to be paid anyway). Small copying fees (up to
#    5 DM or 3 $US) may be charged.
#  * Only distribute the whole original package, with all the files included.
#  * This program may not be part of any commercial product or service without
#    the written permission by the author.
#  * If you want to include this program on a CD-ROM and/or book, please send
#    me a free copy of the CD/book (this is not a must, but I would appreciate
#    it very much).
# 
# Distribution of the program is explicitly desired, provided that the above
# conditions are accepted.
# 
# YOU ARE USING THIS PROGRAM AT YOUR OWN RISK! THE AUTHOR (STEPHAN LÖSCHER)
# IS NOT LIABLE FOR ANY DAMAGE OR DATA-LOSS CAUSED BY THE USE OF THIS PROGRAM
# OR BY THE INABILITY TO USE THIS PROGRAM. IF YOU ARE NOT SURE ABOUT THIS, OR
# IF YOU DON'T ACCEPT THIS, THEN DO NOT USE THIS PROGRAM!
# BECAUSE OF THE VARIOUS HARDWARE AND SOFTWARE ENVIRONMENTS INTO WHICH THIS
# PROGRAM MAY BE PUT, NO WARRANTY OF FITNESS FOR A PARTICULAR PURPOSE IS
# OFFERED.
# GOOD DATA PROCESSING PROCEDURE DICTATES THAT ANY PROGRAM BE THOROUGHLY
# TESTED WITH NON-CRITICAL DATA BEFORE RELYING ON IT.
# 
# No part of the documentation may be reproduced, transmitted, transcribed,
# stored in any retrieval system, or translated into any other language in
# whole or in part, in any form or by any means, whether it be electronic,
# mechanical, magnetic, optical, manual or otherwise, without prior written
# consent of the author, Stephan Löscher.
# 
# You may not make any changes or modifications to this software or this
# manual. You may not decompile, disassemble, or otherwise reverse-engineer
# the software in any way.
# If you got the source, then you are permitted to modify it if you
# contact me and tell me your enhancements.
# You also may include the source as a whole or parts of it into other
# programs, as long as you don't make profit directly out of selling
# the result. If you re-use code of this program then do not remove my name!
# If you include this source-code in your projects, mark it clearly as such
# "... derived from code XXX by Stephan Löscher".
# But don't distribute modified code!
# 
# If you believe your copy of this software has been tampered or altered in
# anyway, shape or form, please contact me immediately! Do not hesitate a
# moment to inform me. Remember, this software should be available to all, in
# the original form, so please do not accept modified or damaged versions of
# my software.
# 
# The author reserves his right for taking legal steps if the copyright or the
# license agreement is violated.
# 
# All product names mentioned in this software are trademarks or registered
# trademarks of their respective owners.
# 
# If you have any questions, ideas, suggestions for improvements or if you find
# bugs (I don't hope so.) then feel free to contact me. (Email is appreciated.)
# 
# I'm not a native english speaker. If you are one and discover some strange
# sounding parts in this documentation or in the program, please, feel free
# to point it out to me and give me suggestions for alteration!
# 
# If the program works for you, and you want to honour my efforts, you are
# invited to donate as much as you want... :)
#
# In any case, if you don't like the restrictions in this license, contact
# me, and we can work something out.
#
######################################################################