tools/convert_mbox.pl

   1 #!/usr/bin/perl
   2 # convert_mbox.pl
   3 # perl script to convert mbox file to files in a new MH directory
   4 # aka another mbox -> MH conversion tool
   5 # 29 April 2003
   6 # Fred Marton <Fred.Marton@uni-bayreuth.de>
   7 #
   8 # Fixed (hopefully) to account for From lines
   9 # that are of various length and that might have
  10 # time zone info at the end
  11 # 20 January 2004
  12 #
  13 # Note: Running this with the -w flag generates the following warnings:
  14 # Scalar value @word[1] better written as $word[1] at /path/to/convert_mbox.pl line 54
  15 # Scalar value @word[0] better written as $word[1] at /path/to/convert_mbox.pl line 56
  16 # Making these changes requires further changes in the script
  17 # that results in much longer run-times.
  18 #
  19 # Copyright (c) 2003 Fred Marton
  20 # Copyright (c) 2009 Ricardo Mones, based on the bash script
  21 #                    by Daniel Dickinson [1]
  22 #
  23 # This program is free software; you can redistribute it and/or
  24 # modify it under the terms of the GNU General Public License
  25 # as published by the Free Software Foundation; either version 3
  26 # of the License, or (at your option) any later version.
  27 #
  28 # This program is distributed in the hope that it will be useful,
  29 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  30 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  31 # GNU General Public License for more details.
  32 #
  33 # You should have received a copy of the GNU General Public License
  34 # along with this program; if not, write to the Free Software
  35 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
  36 # 02111-1307, USA.
  37 #
  38 # [1] http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=15;filename=convert_mbox.sh;att=1;bug=461435
  39
  40 use File::Path;
  41 use File::Basename;
  42
  43 # check for arguments
  44 &usage if ($#ARGV < 1);
  45 if ($ARGV[0] eq "-R") {
  46   # recursive, need more args
  47   &usage if ($#ARGV < 2);
  48   &usage unless (-d $ARGV[1]);
  49   &usage unless (-d $ARGV[2]);
  50   &convert_all ($ARGV[1], $ARGV[2]);
  51 }
  52 else {
  53   # default behaviour
  54   &convert_mbox ($ARGV[0], $$ARGV[1]);
  55 }
  56
  57 sub strip_sbd {
  58   # raw translation of bash, probably doable in real Perl better
  59   $dirname = shift;
  60   $nosbddir = "";
  61   $dirleft = $dirname;
  62   while ($dirleft ne "." and $dirleft ne "") {
  63     $enddir = &basename ($dirleft, ['.sbd']);
  64     if ($nosbddir ne "") {
  65       $nosbddir = "$enddir/$nosbddir";
  66     }
  67     else {
  68       $nosbddir = $enddir;
  69     }
  70     $dirleft = &dirname ($dirleft);
  71   }
  72   print "$nosbddir\n";
  73 }
  74
  75 sub convert_all {
  76   ($mboxdir, $targetdir) = @_;
  77   $tmpbase = '/tmp/frommbox';
  78   $tmplist = '/tmp/.convert_mbox.filelist';
  79   $curdir = qx/pwd/;
  80   chdir ($mboxdir);
  81   qx/find * -type -d -a ! -name '*.*'i | sort > $tmplist/;
  82   open (FLH, "<$tmplist") or die "cannot open $tmplist: $!\n";
  83   while (<FLH>) {
  84     chomp;
  85     &convert_one ($_, $targetdir, $tmpbase);
  86   }
  87   close (FLH);
  88   unlink ($tmplist);
  89   chdir ($curdir);
  90 }
  91
  92 sub convert_one {
  93   # targetdir isn't used in the original bash script, maybe a bug?
  94   ($file, $targetdir, $tmpbase) = @_;
  95   $dirname = &dirname ($file);
  96   $filename = &basename ($file);
  97   if ($dirname eq $filename) {
  98     $dirname = "";
  99   }
 100   $dirname = &strip_sbd ($dirname);
 101   if ($dirname ne "") {
 102     $dirname = "/$dirname";
 103   }
 104   &mkpath ($tmpbase . $dirname);
 105   $basefile = &basename ($file);
 106   $base1 = &basename ($file, ['.cmeta']);
 107   $base2 = &basename ($base1, ['.ibex.index']);
 108   $base3 = &basename ($base2, ['.ibex.index.data']);
 109   $base5 = &basename ($base3, ['.ev-summary']);
 110   $base4 = &basename ($base5, ['.ev-summary-meta']);
 111   $subdir = &basename ($dirname);
 112   $basedir = "/" . &dirname ($dirname);
 113   if ( $basedir eq '/.' or $basedir eq '/' ) {
 114     $basedir = $dirname;
 115     $subdir = "";
 116   }
 117   $basedir = $tmpbase . $basedir;
 118   &mkpath ($basedir);
 119   $dirisfile = 0;
 120   if ( -f "$basedir/$subdir" ) {
 121     $dirisfile = 1;
 122     qx/mv \"$basedir\/$subdir\" \"$basedir\/$subdir.tmp\"/;
 123     &mkpath ("$basedir/$subdir");
 124   }
 125   if ( ! -d "$basedir/$subdir/$base4" ) {
 126     print "$basedir/$subdir/$base4\n";
 127     &convert_mbox ($file, "$basedir/$subdir/$base4", 'quiet');
 128     if ($dirisfile == 1) {
 129       qx/mv \"$basedir\/$subdir.tmp\/*\" \"$basedir\/$subdir\/\"/;
 130     }
 131   }
 132 }
 133
 134
 135 sub convert_mbox {
 136   ($mbox, $mh, $quiet) = @_;
 137   # check to make sure there isn't something named MH already
 138   if (-e $mh) {
 139      die (" The directory \"$mh\" already exists.  Exiting.\n");
 140   }
 141   else {
 142      mkdir $mh;
 143   }
 144   # start numbering
 145   $i = 0;
 146   # open the mbox file
 147   open (IN, $mbox);
 148   while ($line = <IN>) {
 149     # check for the beginning of an e-mail
 150     @word = split(/ +/m,$line);
 151     # some lines might start with "From ", so check
 152     # to see if the [second-to-]last word is a year
 153     @word2 = split(/:/,$line);
 154     chomp($word2[$#word2]);
 155     @word3 = split(/ /,$word2[2]);
 156     $year = @word3[1];
 157     # ignore the MAILER-DAEMON message from pine
 158     if (@word[1] ne "MAILER-DAEMON") {
 159       # start a new file, assuming $year is > 1970
 160       if (@word[0] eq "From" && $year > 1970) {
 161         $i++;
 162         close (OUT);
 163         open (OUT, ">$mh/$i");
 164         print OUT $line;
 165       }
 166       else {
 167         # continue the file
 168         print OUT $line;
 169       }
 170     }
 171   }
 172   close (OUT);
 173   close (IN);
 174   # and we're done
 175   if (! defined($quiet)) {
 176     print "\n If it isn't there already, please move the directory \"$mh\"\n"
 177           . " into your MH directory and rebuild your folder tree.\n\n";
 178   }
 179 }
 180
 181 sub usage {
 182    die ("Usage: convert_mbox.pl MBOX MH_DIR\n"
 183         . "       convert_mbox.pl -R MBOXDIR MH_DIR\n"
 184         . "Where: \n"
 185         . "  -R  Converts recursively a directory of mboxes\n");
 186 }
 187
 188