transex3/scripts/keyidGen.pl

   1 :
   2 eval 'exec perl -S $0 ${1+"$@"}'
   3     if 0;
   4 #*************************************************************************
   5 #
   6 # DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   7 #
   8 # Copyright 2008 by Sun Microsystems, Inc.
   9 #
  10 # OpenOffice.org - a multi-platform office productivity suite
  11 #
  12 # $RCSfile: keyidGen.pl,v $
  13 #
  14 # $Revision: 1.3 $
  15 #
  16 # This file is part of OpenOffice.org.
  17 #
  18 # OpenOffice.org is free software: you can redistribute it and/or modify
  19 # it under the terms of the GNU Lesser General Public License version 3
  20 # only, as published by the Free Software Foundation.
  21 #
  22 # OpenOffice.org is distributed in the hope that it will be useful,
  23 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  24 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  25 # GNU Lesser General Public License version 3 for more details
  26 # (a copy is included in the LICENSE file that accompanied this code).
  27 #
  28 # You should have received a copy of the GNU Lesser General Public License
  29 # version 3 along with OpenOffice.org.  If not, see
  30 # <http://www.openoffice.org/license.html>
  31 # for a copy of the LGPLv3 License.
  32 #
  33 #*************************************************************************
  34 #
  35 # add keyids to sdf file
  36 #
  37
  38 use Compress::Zlib();
  39
  40 print "\nkeyidGen version 1.0 \n\n";
  41
  42 my ( $infile,$outfile,$dbimport );
  43 get_options();
  44
  45 print_help() if ( !defined $infile || $help );
  46 exit 1 if ( !defined $infile );
  47 if ( ! defined $outfile )
  48 {
  49     $outfile = $infile;
  50     $outfile =~ s/\.sdf$//i;
  51     $outfile .= "_KeyID.sdf";
  52 }
  53
  54 $collisions = 0;
  55 %hashcodes = ();
  56 $count = 0;
  57 print "writing to $outfile\n";
  58 open INFILE,"<$infile" || die "could not open $infile $! $^E\n";
  59 open OUTFILE,">$outfile" || die "could not open $outfile $! $^E\n";
  60
  61 while ( <INFILE> )
  62 {
  63     $line = $_;
  64     chomp $line;
  65     $hash = 0;
  66     if ( $line =~ /^([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)$/ )
  67     {
  68         $string="$1 $2 $4 $5 $6 $7 $8";
  69         $hashp = makeID( $string );
  70
  71         if ( defined ($hashcodes{ $hashp } ) )
  72         {
  73             $collisions ++ unless $hashcodes{ $hashp } eq $string;
  74         }
  75         $hashcodes{ $hashp } = $string;
  76         $count++;
  77         if ( $dbimport )
  78         {
  79             my ( $pre, $post, $old );
  80             $pre = "$1\t$2\t";
  81             $post = "\t$4\t$5\t$6\t$7\t$8\t$9\t$10\t$11\t$12\t$13\t$14\t$15\n";
  82             $old = $3;
  83             $old =~ s/;{0,1}keyid:......;{0,1}//;
  84             $old =~ s/^0$//;
  85             if ( $old ne "" ) { $old .= ";"; }
  86             print OUTFILE "$pre${old}keyid:$hashp$post";
  87         }
  88         else
  89         {
  90             print OUTFILE "$1\t$2\t$3\t$4\t$5\t$6\t$7\t$8\t$9\t$10\t".makekidstr($hashp,$11)."\t".makekidstr($hashp,$12)."\t$13\t".makekidstr($hashp,$14)."\t$15\n";
  91         }
  92     }
  93 }
  94 print "$count entries\n";
  95 print "$collisions collisions\n";
  96
  97 close INFILE;
  98 close OUTFILE;
  99
 100 sub makeID
 101 {
 102     my ( $String ) = shift;
 103     my ( $hash );
 104     # hardcoded to prevent windows installer to choke on bad directoryname :-((
 105     if ( $String eq "scp2 source\\ooo\\directory_ooo.ulf LngText STR_DIR_KAPITEL   " )
 106     {
 107         return "keyid1";
 108     }
 109
 110     $hash = Compress::Zlib::crc32( $String, undef );
 111     return makenumber( $hash );
 112 }
 113
 114 sub makenumber
 115 {
 116     $h = shift;
 117     #                  1         2         3         4
 118     #         1234567890123456789012345678901234567890
 119     $symbols="0123456789abcdefghijklmnopqrstuvwxyz+-<=>";
 120     $order = length($symbols);
 121     $result = "";
 122     while ( length( $result ) < 6 )
 123     {
 124         $result .= substr( $symbols, ($h % $order), 1 );
 125         $h = int( $h / $order );
 126     }
 127     die "makenumber failed because number is too big (this cannot be so this is a strange error)" if $h > 0;
 128
 129     return reverse $result;
 130 }
 131
 132
 133 sub makekidstr
 134 {
 135     $kid = shift;
 136     $str = shift;
 137
 138     if ( $str ne "" )
 139     {
 140         # special handling for strings starting with font descriptions like {&Tahoma8} (win system integration)
 141         if ( $str =~ s/^(\{\&[^\}]+\})// )
 142         {
 143             return "$1$kid‖$str";
 144         }
 145         else
 146         {
 147             return "$kid‖$str";
 148         }
 149     }
 150     else
 151     {
 152         return "";
 153     }
 154 #    return "default";
 155 }
 156
 157 sub print_help
 158 {
 159     print "\n\n";
 160     print "keyidGen 0.5 for sdf files\n";
 161     print "--------------------------\n";
 162     print "Usage:\n";
 163     print "keyidGen <infile> [<outfile>] [-dbimport]\n";
 164     print "                   add keyids to the entries and write them to a file with\n";
 165     print "                   _KeyID added to the name\n";
 166     print "   -dbimport       Add KeyID to a new column instead of to the strings.\n";
 167     print "                   This is needed to import the IDs into tha database.\n";
 168     print "\n\n";
 169 }
 170
 171
 172 sub get_options {
 173     my ($arg,$has_infile);
 174
 175     while ($arg = shift @ARGV) {
 176         $arg =~ /^-dbimport$/  and $dbimport = 1 and next;
 177         $arg =~ /^-help$/  and $help = 1 and next; #show help
 178
 179         if ( !$has_infile )
 180         {
 181             $infile = $arg;
 182             $has_infile = 1;
 183         }
 184         else
 185         {
 186             $outfile = $arg;
 187         }
 188     }
 189 }