extract_mail

   1 #!/usr/bin/env perl
   2
   3 #=======================================================================
   4 # extract_mail
   5 # File ID: 953fa428-5d39-11df-8ca0-90e6ba3022ac
   6 # Extracts messages from a mailbox into separate files
   7 # ©opyleft 2002 Øyvind A. Holm <sunny@sunbase.org>
   8 # License: GNU General Public License version 2 or later.
   9 #=======================================================================
  10
  11 use strict;
  12 use warnings;
  13 use Time::Local;
  14
  15 my @Msg = ();
  16 my $Date = "";
  17 my $date_done = 0;
  18 my $dest_dir = "out";
  19 my $TZ_ext = "Z";
  20 my %mon_str = (
  21         "jan" => "01",
  22         "feb" => "02",
  23         "mar" => "03",
  24         "apr" => "04",
  25         "may" => "05",
  26         "jun" => "06",
  27         "jul" => "07",
  28         "aug" => "08",
  29         "sep" => "09",
  30         "oct" => "10",
  31         "nov" => "11",
  32         "dec" => "12"
  33 );
  34
  35 -d $dest_dir || mkdir($dest_dir, 0777) || die("mkdir($dest_dir): $!");
  36
  37 while (<>) {
  38         if (/^From /) {
  39                 D("========= $_");
  40                 if (scalar(@Msg)) {
  41                         if (length($Date)) {
  42                                 ($Date =~ /^\d{8}T\d{6}$/) || die("Line $.: Invalid date format: \"$Date\"");
  43                                 my $file_name = "$dest_dir/mail.${Date}$TZ_ext.txt";
  44                                 -e $file_name && (warn("$file_name: File already exists"), $Date .= "b");
  45                                 open(ToFP, ">$file_name") || die("$file_name: Opening for write: $!");
  46                                 for (@Msg) {
  47                                         print(ToFP $_);
  48                                 }
  49                                 close(ToFP);
  50                         } else {
  51                                 warn("Datoen var tom");
  52                         }
  53                 }
  54                 ($Date, $date_done, $TZ_ext) = ("", 0, "");
  55                 @Msg = ();
  56         }
  57         push(@Msg, $_);
  58         if (/^Date: / && !$date_done) {
  59                 if (
  60                                 /^\D* # Before the date
  61                                 (\d+) # Day
  62                                 \s+
  63                                 (\S+) # Month
  64                                 \s+
  65                                 (\d+) # Year
  66                                 \s+
  67                                 (\d+)   # Hours
  68                                 :
  69                                 (\d+)   # Minutes
  70                                 :
  71                                 (\d+)   # Seconds
  72                                 \s*
  73                                 (\S*)   # Time zone
  74                                 /x
  75                         ) {
  76                         my ($Year, $Mon, $Day, $Hour, $Min, $Sec, $Timezone) = ($3, $2, $1, $4, $5, $6, "");
  77                         (defined($7) && ($Timezone = $7)) || warn("Line $.: Date missing time zone");
  78                         $Mon = lc($Mon);
  79                         D("Day = \"$Day\", Mon = \"$Mon\"\n");
  80                         defined($mon_str{$Mon}) || (die("Unknown month: \"$Mon\""), ($Mon = "XX"));
  81                         $Mon = $mon_str{$Mon};
  82                         my $curr_date = timelocal($Sec, $Min, $Hour, $Day, $Mon-1, $Year);
  83                         my $zone_diff = 0;
  84                         if ($Timezone =~ /^(\+|-)(\d\d)(\d\d)$/) {
  85                                 $zone_diff = $2*3600 + $3*60;
  86                                 ($1 eq "+") && ($zone_diff = 0-$zone_diff);
  87                         } elsif ($Timezone =~ /^GMT$/i) {
  88                                 $zone_diff = 0;
  89                         } else {
  90                                 warn("Line $.: Unknown timezone format");
  91                                 $TZ_ext = "";
  92                         }
  93                         D("curr_date before tz: \"$curr_date\", Timezone = \"$Timezone\"\n");
  94                         $zone_diff = 0; # Temporary until the timezone problems are eliminated.
  95                         $curr_date += $zone_diff;
  96                         D("curr_date after tz : \"$curr_date\", Timezone = \"$Timezone\"\n");
  97                         D("zone_diff = \"$zone_diff\"\n");
  98                         ($Sec, $Min, $Hour, $Day, $Mon, $Year) = gmtime($curr_date);
  99                         $Year += 1900; # Urgh
 100                         $Mon   = sprintf("%02u", $Mon+1); # Urgh II
 101                         $Day   = sprintf("%02u", $Day);
 102                         $Hour  = sprintf("%02u", $Hour);
 103                         $Min   = sprintf("%02u", $Min);
 104                         $Sec   = sprintf("%02u", $Sec);
 105                         $Date = "$Year$Mon${Day}T$Hour$Min$Sec";
 106                         $date_done = 1;
 107                 } else {
 108                         warn("Line $.: Weird date: \"$Date\"");
 109                 }
 110         }
 111 }
 112
 113 sub D {
 114         print(STDERR @_);
 115 }
 116
 117 =pod
 118
 119 =head1 LICENCE
 120
 121 This program is free software; you can redistribute it and/or modify it
 122 under the terms of the GNU General Public License as published by the
 123 Free Software Foundation; either version 2 of the License, or (at your
 124 option) any later version.
 125
 126 This program is distributed in the hope that it will be useful, but
 127 WITHOUT ANY WARRANTY; without even the implied warranty of
 128 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 129 See the GNU General Public License for more details.
 130
 131 You should have received a copy of the GNU General Public License along
 132 with this program; if not, write to the Free Software Foundation, Inc.,
 133 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 134
 135 =cut
 136
 137 #### End of file extract_mail ####