gecko-dev/webtools/tinderbox/tbglobals.pl

# -*- Mode: perl; indent-tabs-mode: nil -*-
#
# The contents of this file are subject to the Netscape Public
# License Version 1.1 (the "License"); you may not use this file
# except in compliance with the License. You may obtain a copy of
# the License at http://www.mozilla.org/NPL/
#
# Software distributed under the License is distributed on an "AS
# IS" basis, WITHOUT WARRANTY OF ANY KIND, either express or
# implied. See the License for the specific language governing
# rights and limitations under the License.
#
# The Original Code is the Tinderbox build tool.
#
# The Initial Developer of the Original Code is Netscape Communications
# Corporation. Portions created by Netscape are
# Copyright (C) 1998 Netscape Communications Corporation. All
# Rights Reserved.
#
# Contributor(s): 

# Reading the log backwards saves time when we only want the tail.
use Backwards;

#
# Global variabls and functions for tinderbox
#

#
# Constants
#
$::CI_CHANGE=0;
$::CI_DATE=1;
$::CI_WHO=2;
$::CI_REPOSITORY=3;
$::CI_DIR=4;
$::CI_FILE=5;
$::CI_REV=6;
$::CI_STICKY=7;
$::CI_BRANCH=8;
$::CI_LINES_ADDED=9;
$::CI_LINES_REMOVED=10;
$::CI_LOG=11;

#
# Global variables
#

# From load_data()
$ignore_builds = {};
$scrape_builds = {};

# From get_build_name_index()
$build_name_index = {};     
$build_names = [];
$name_count = 0;

# Frome get_build_time_index()
$build_time_index = {};
$build_time_times = [];
$mindate_time_count = 0;  # time_count that corresponds to the mindate
$time_count = 0;

$build_table = [];
$who_list = [];
@note_array = ();

$gzip = '/usr/bin/gzip';
$bzip2 = '/usr/bin/bzip2';

$data_dir='data';

# Set this to show real end times for builds instead of just using
# the start of the next build as the end time.
$display_accurate_build_end_times = 1;

1;

sub trick_taint{
    my $in = shift;
    return undef if !defined($in);
    $in =~ /(.*)/;
    return $1;
}

sub lock{
}

sub unlock{
}

sub print_time {
  my ($t) = @_;
  my ($sec,$minute,$hour,$mday,$mon);
  ($sec,$minute,$hour,$mday,$mon,undef) = localtime($t);
  sprintf("%02d/%02d %02d:%02d:%02d",$mon+1,$mday,$hour,$minute,$sec);
}

#------------------
# is_today( $time )
#
# Takes a Unix time_t and returns true if it represents a time that has
# the same 'day of year' and 'year' as today.  Returns true if no arguments
# were given.
#
sub is_today {
  my ($tt) = @_;

  my $today = 1;

  if ( $tt ) {
    my $tt_year = (localtime($tt))[5];
    my $tt_yday = (localtime($tt))[7];

    my $now_year = (localtime())[5];
    my $now_yday = (localtime())[7];

    if ( ($tt_year ne $now_year) or ($tt_yday ne $now_yday) ) {
      $today = 0;
    }
  }

  return $today;
}

#---------------------------------
# both_are_today( $time1, $time2 )
#
# Tests both $time1 and $time2 to see if either are not today.  Returns true
# in that case and false otherwise.
#
sub both_are_today {
  my ($t11, $t22) = @_;

  return 0 if ! is_today($t11);
  return 0 if ! is_today($t22);
  return 1;
}

#---------------------------------
# get_local_hms( $time, $qualify )
#
# Converts a Unix time_t value into a date using format HH:MM.  If $qualify
# is true, it qualifies the date by prepending mm/DD.  Returns the resulting
# string.
#
sub get_local_hms {
  my ($t, $need_to_qualify) = @_;

  my (undef,$minute,$hour,$mday,$mon,undef) = localtime($t);
  my $formatted_date = sprintf("%02d:%02d",$hour,$minute);

  if ($need_to_qualify) {
    $formatted_date = sprintf("%02d/%02d",$mon+1,$mday) . " " . $formatted_date;
  }

  return $formatted_date;
}

#--------------------------------------
# get_time_difference( $time1, $time2 )
#
# Calculates a human-readable difference in time between two Unix time_t
# values.  Returns the resulting string.
#
sub get_time_difference {
  my ($t11, $t22) = @_;

  $t11 = 0 if !$t11;
  $t22 = 0 if !$t22;

  # Flip $t11 and $t22 if $t22 isn't later than $t11.
  if ( $t11 gt $t22 ) {
    my $temp = $t11;
    $t11 = $t22;
    $t22 = $temp;
  }

  my $time_diff = $t22 - $t11;

  # @time_slots is an array that we will keep our time difference in.
  #   0 => seconds
  #   1 => minutes
  #   2 => hours
  #   3 => days
  my @time_slots;

  # Zero out the array.
  for my $ii ( 0 .. 3 ) {
    $time_slots[$ii] = 0;
  }
  
  # What's in $time_diff is the difference in seconds.  Modulo 60 to get a
  # number of seconds within a minute.
  $time_slots[0] = $time_diff % 60;
  $time_diff = $time_diff / 60;

  # Now what's left in $time_diff is the difference in minutes.  Modulo 60 to
  # get a number of minutes within an hour.
  $time_slots[1] = $time_diff % 60;
  $time_diff = $time_diff / 60;

  # Now what's left in $time_diff is the difference in hours.  Modulo 24 to get
  # a number of hours within a day.
  $time_slots[2] = $time_diff % 24;
  $time_diff = $time_diff / 24;

  # Now what's left in $time_diff are the number of days.  Floor it to get rid
  # of any fractional left from previous maths.
  $time_slots[3] = floor($time_diff);

  format_time_difference(@time_slots);
}

#--------------------------------------
# format_time_difference( @time_slots )
#
# Called by get_time_difference(), takes its @time_slots arrary and parses it
# into human-readable text.  Returns the resulting string.
#
sub format_time_difference {
  my @time_slots = @_;

  # As above, @time_slots is an array that holds our time difference.
  #   0 => seconds
  #   1 => minutes
  #   2 => hours
  #   3 => days

  my @formatted;

  if ( $time_slots[3] eq 1 ) {
    push(@formatted, "1 day");
  } elsif ( $time_slots[3] gt 1 ) {
    push(@formatted, $time_slots[3] . " days");
  }

  if ( $time_slots[2] eq 1 ) {
    push(@formatted, "1 hour");
  } elsif ( $time_slots[2] gt 1 ) {
    push(@formatted, $time_slots[2] . " hours");
  }

  if ( $time_slots[1] eq 1 ) {
    push(@formatted, "1 minute");
  } elsif ( $time_slots[1] gt 1 ) {
    push(@formatted, $time_slots[1] . " minutes");
  }

  # I love that we can track seconds and let the world know, but the unit is
  # too small to be of utility right now.  Sticking this code in an always-
  # false conditional in case we want to turn this on in the future.
  if ( 0 ) {
    if ( $time_slots[0] eq 1 ) {
      push(@formatted, "1 second");
    } elsif ( $time_slots[0] gt 1 ) {
      push(@formatted, $time_slots[0] . " seconds");
    }
  }

  my $formatted_diff;
  if ( scalar(@formatted) gt 0 ) {
    $formatted_diff = join(', ', @formatted);
  } else {
    $formatted_diff = "no time";
  }

  return $formatted_diff;
}

sub url_encode {
  my ($s) = @_;

  $s =~ s/\%/\%25/g;
  $s =~ s/\=/\%3d/g;
  $s =~ s/\?/\%3f/g;
  $s =~ s/ /\%20/g;
  $s =~ s/\n/\%0a/g;
  $s =~ s/\r//g;
  $s =~ s/\"/\%22/g;
  $s =~ s/\'/\%27/g;
  $s =~ s/\|/\%7c/g;
  $s =~ s/\&/\%26/g;
  $s =~ s/\+/\%2b/g;
  return $s;
}

sub url_decode {
  my ($value) = @_;
  $value =~ tr/+/ /;
  $value =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg;
  return $value;
}


sub value_encode {
  my ($s) = @_;
  $s =~ s@&@&@g;
  $s =~ s@<@&lt;@g;
  $s =~ s@>@&gt;@g;
  $s =~ s@\"@&quot;@g;
  return $s;
}


BEGIN {
  my $have_loaded_treedata = 0;

  sub tb_load_treedata {
    my $tree = shift;

    return if $have_loaded_treedata;
    $have_loaded_treedata = 1;
    require "$tree/treedata.pl" if -r "$tree/treedata.pl";
  }
}

sub tb_load_data {
  $tree = $form{'tree'};  # Testing: $tree = "SeaMonkey";

  return undef unless $tree;
  
  tb_load_treedata($tree);
        
  $ignore_builds = {};
  $scrape_builds = {};

  require "$tree/ignorebuilds.pl" if -r "$tree/ignorebuilds.pl";
  require "$tree/scrapebuilds.pl" if -r "$tree/scrapebuilds.pl";
        
  $td = {};
  $td->{name} = $tree;
  $td->{num} = 0;
  $td->{cvs_module} = $cvs_module;
  $td->{cvs_branch} = $cvs_branch;
  $td->{ignore_builds} = $ignore_builds;
  $td->{scrape_builds} = $scrape_builds;
  $cvs_root = '/m/src' if $cvs_root eq '';
  $td->{cvs_root} = $cvs_root;

  $build_list = load_buildlog($td);
  
  get_build_name_index($build_list);
  get_build_time_index($build_list);
  
  load_who($td, $who_list);

  make_build_table($td, $build_list);

  $td->{scrape}     = load_scrape($td);
  $td->{warnings}   = load_warnings($td);

  return $td;
}

sub tb_loadquickparseinfo {
  my ($tree, $build, $times, $includeStatusOfBuilding) = (@_);
  local $_;

  return if (! -d "$tree" || ! -r "$tree/build.dat");
  $maxdate = time;
  require "$tree/ignorebuilds.pl" if -r "$tree/ignorebuilds.pl";
    
  my $bw = Backwards->new("$tree/build.dat") or die;
    
  my $latest_time = 0;
  my $tooearly = 0;
  while( $_ = $bw->readline ) {
    chop;
    my ($buildtime, $buildname, $buildstatus) = (split /\|/)[1,2,4];
    
    if ($includeStatusOfBuilding or
        $buildstatus =~ /^success|busted|testfailed$/) {

      # Ignore stuff in the future.
      next if $buildtime > $maxdate;

      $latest_time = $buildtime if $buildtime > $latest_time;

      # Ignore stuff more than 12 hours old
      if ($buildtime < $latest_time - 12*60*60) {
        # Hack: A build may give a bogus time. To compensate, we will
        # not stop until we hit 20 consecutive lines that are too early.
        # XXX bug 225735: This is the wrong way of doing things.  When a
        # flood of backed up mail from one machine comes in, it can easily
        # be more than 20.  We should be looking at the mail receipt time
        # to decide when we're done rather than the build time.

        last if $tooearly++ > 20;
        next;
      }
      $tooearly = 0;

      next if exists $ignore_builds->{$buildname};
      next if exists $build->{$buildname}
              and $times->{$buildname} >= $buildtime;
      
      $build->{$buildname} = $buildstatus;
      $times->{$buildname} = $buildtime;
    }
  }
}

sub tb_last_status {
  my ($build_index) = @_;

  for (my $tt=0; $tt < $time_count; $tt++) {
    my $br = $build_table->[$tt][$build_index];
    next unless defined $br and $br->{buildstatus};
    next unless $br->{buildstatus} =~ /^(success|busted|testfailed)$/;
    return $br->{buildstatus};
  }
  return 'building';
}

sub tb_check_password {
  if ($form{password} eq '' and defined $cookie_jar{tinderbox_password}) {
    $form{password} = $cookie_jar{tinderbox_password};
  }
  my $correct = '';
  if (open(REAL, '<data/passwd')) {
    $correct = <REAL>;
    close REAL;
    $correct =~ s/\s+$//;   # Strip trailing whitespace.
  }
  $form{password} =~ s/\s+$//;      # Strip trailing whitespace.
  if ($form{password} ne '') {
    open(TRAPDOOR, "../bonsai/data/trapdoor $form{'password'} |") 
      or die "Can't run trapdoor func!";
    my $encoded = <TRAPDOOR>;
    close TRAPDOOR;
    $encoded =~ s/\s+$//;   # Strip trailing whitespace.
    if ($encoded eq $correct) {
      if ($form{rememberpassword} ne '') {
        print "Set-Cookie: tinderbox_password=$form{'password'} ;"
             ." path=/ ; expires = Sun, 1-Mar-2020 00:00:00 GMT\n";
      }
      return;
    }
  }

  # Force a return here to test w/o a password.
  # return;

  require 'header.pl';

  print "Content-type: text/html\n";
  print "Set-Cookie: tinderbox_password= ; path=/ ; "
       ." Expires = Sun, 1-Mar-2020 00:00:00 GMT\n";
  print "\n";

  EmitHtmlHeader("What's the magic word?",
                 "You need to know the magic word to use this page.");

  if ($form{password} ne '') {
    print "<B>Invalid password; try again.<BR></B>";
  }
  print q(
    <FORM method=post>
    <B>Password:</B>
    <INPUT NAME=password TYPE=password><BR>
    <INPUT NAME=rememberpassword TYPE=checkbox>
    If correct, remember password as a cookie<BR>
  );
    
  while (my ($key,$value) = each %form) {
    next if $key eq "password" or $key eq "rememberpassword";

    my $enc = value_encode($value);
    print "<INPUT TYPE=HIDDEN NAME=\"$key\" VALUE=\"$enc\">\n";
  }
  print "<INPUT TYPE=SUBMIT value=Submit></FORM>\n";
  exit;
}

sub tb_find_build_record {
  my ($tree, $logfile) = @_;
  local $_;

  my $log_entry = '';
  my ($bw) = Backwards->new("$tree/build.dat") or die;
  while( $_ = $bw->readline ) {
    $log_entry = $_ if /$logfile/;
  }

  chomp($log_entry);
  # Skip the logfile in the parse since it is already known.
  my ($mailtime, $buildtime, $buildname, $errorparser,
      $buildstatus, $binaryurl) = (split /\|/, $log_entry)[0..4,6];

  $buildrec = {    
    mailtime    => $mailtime,
    buildtime   => $buildtime,
    buildname   => $buildname,
    errorparser => $errorparser,
    buildstatus => $buildstatus,
    logfile     => $logfile,
    binaryurl   => $binaryurl,
    td          => undef
  };
  return $buildrec;
}

# end of public functions
#============================================================

sub load_buildlog {
  my ($treedata) = $_[0];

  # In general you always want to make "$_" a local
  # if it is used. That way it is restored upon return.
  local $_;
  my $build_list = [];


  if (not defined $maxdate) {
    $maxdate = time();
  }
  if (not defined $mindate) {
    $mindate = $maxdate - 24*60*60;
  }
  
  my ($bw) = Backwards->new("$treedata->{name}/build.dat") or die;

  my $tooearly = 0;
  my $internal_build_list;
  LOOP: while( $_ = $bw->readline ) {
    chomp;
    my ($mailtime, $buildtime, $buildname,
     $errorparser, $buildstatus, $logfile, $binaryurl) = split /\|/;
    
    my ($endtime);

    # Ignore stuff in the future.
    next if $buildtime > $maxdate;
    
    # Ignore stuff in the past (but get a 2 hours of extra data)
    if ($buildtime < $mindate - 2*60*60) {
      # Occasionally, a build might show up with a bogus time.  So,
      # we won't judge ourselves as having hit the end until we
      # hit a full 20 lines in a row that are too early.
      # XXX bug 225735: This is the wrong way of doing things.  When a
      # flood of backed up mail from one machine comes in, it can easily
      # be more than 20.  We should be looking at the mail receipt time
      # to decide when we're done rather than the build time.
      last if $tooearly++ > 20;
      
      next;
    }
    $tooearly = 0;

    # We should get clients reporting their actual end time.  This would
    # alleviate an obvious problem where a build starts immediately on the
    # same system after another completed.  Now, since we infer the end time
    # based on the mail time, the reported end time of the first build can
    # be later than the reported start time of the second.

    # Completed builds are eligible for inference of their end time.  If the
    # build is on-going, it doesn't have an end time, so leave it undefined.
    $endtime = $mailtime if ($buildstatus ne 'building');

    if ($form{noignore} or not $treedata->{ignore_builds}->{$buildname}) {

      # Latest record in build.dat for this (buildtime, buildname) tuple wins.
      if ( $internal_build_list->{$buildtime}->{$buildname} ) {
        next LOOP;
      } else {
        $internal_build_list->{$buildtime}->{$buildname} = 1;
      }

      my $buildrec = {    
                      endtime     => $endtime,
                      mailtime    => $mailtime,
                      buildtime   => $buildtime,
                      buildname   => $buildname,
                      errorparser => $errorparser,
                      buildstatus => $buildstatus,
                      logfile     => $logfile,
                      binaryurl   => $binaryurl,
                      td          => $treedata
                     };
      push @{$build_list}, $buildrec;
    }
  }
  return $build_list;
}

# Load data about who checked in when
#   File format: <build_time>|<email_address>
#
sub load_who {
  my ($treedata, $who_list) = @_;
  local $_;
  
  open(WHOLOG, "<$treedata->{name}/who.dat");
  while (<WHOLOG>) {
    chomp;
    my ($checkin_time, $email) = split /\|/;

    # Find the time slice where this checkin belongs.
    for (my $ii = $time_count - 1; $ii >= 0; $ii--) {
      if ($checkin_time < $build_time_times->[$ii]) {
        $who_list->[$ii+1]->{$email} = 1;
        last;
      } elsif ($ii == 0) {
        $who_list->[0]->{$email} = 1;
      }
    }
  }

  # Ignore the last one
  #
  #if ($time_count > 0) {
  #  $who_list->[$time_count] = {};
  #}
}



# Load data about scrape data.
#   File format: <logfile>|<aaa>|<bbb>|...
#
sub load_scrape {
  my $treedata = $_[0];
  local $_;

  my $scrape = {};
  
  open(SCRAPELOG, "<$treedata->{name}/scrape.dat");
  while (<SCRAPELOG>) {
    chomp;
    my @list =  split /\|/;
    my $logfile = @list[0];
    shift(@list);

    $scrape->{$logfile} = [ @list ];
  }
  return $scrape;
}


# Load data about build warnings
#   File format: <logfile>|<warning_count>
#
sub load_warnings {
  my $treedata = $_[0];
  local $_;

  my $warnings = {};

  open(WARNINGLOG, "<$treedata->{name}/warnings.dat");
  while (<WARNINGLOG>) {
    chomp;
    my ($logfile, $warning_count) = split /\|/;
    $warnings->{$logfile} = $warning_count;
  }
  return $warnings;
}

sub get_build_name_index {
  my ($build_list) = @_;

  # Get all the unique build names.
  #
  foreach my $build_record (@{$build_list}) {
    $build_name_index->{$build_record->{buildname}} = 1;
  }
    
  my $ii = 0;
  foreach my $name (sort keys %{$build_name_index}) {
    $build_names->[$ii] = $name;
    $build_name_index->{$name} = $ii;
    $ii++;
  }
  $name_count = $#{$build_names} + 1;
}

sub get_build_time_index {
  my ($build_list) = @_;

  # Get all the unique build names.
  #
  foreach my $br (@{$build_list}) {
    $build_time_index->{$br->{buildtime}} = 1;
    if ( $br->{endtime} and $display_accurate_build_end_times ) {
      $build_time_index->{$br->{endtime}} = 1;
    }
  }

  my $ii = 0;
  foreach my $time (sort {$b <=> $a} keys %{$build_time_index}) {
    $build_time_times->[$ii] = $time;
    $build_time_index->{$time} = $ii;
    $mindate_time_count = $ii if $time >= $mindate;
    $ii++;
  }
  $time_count = $#{$build_time_times} + 1;
}

sub make_build_table {
  my ($treedata, $build_list) = @_;
  my ($ti, $bi, $ti1, $br);

  # Create the build table
  #
  for (my $ii=0; $ii < $time_count; $ii++){
    $build_table->[$ii] = [];
  }

  # Populate the build table with build data
  #
  foreach $br (reverse @{$build_list}) {
    $ti = $build_time_index->{$br->{buildtime}};
    $bi = $build_name_index->{$br->{buildname}};
    $build_table->[$ti][$bi] = $br;
  }

  &load_notes($treedata);

  for ($bi = $name_count - 1; $bi >= 0; $bi--) {
    for ($ti = $time_count - 1; $ti >= 0; $ti--) {
      if (defined($br = $build_table->[$ti][$bi])
          and not defined($br->{rowspan})) {

        # Find the next-defined cell after us.  We may run all the way to the
        # end of the page and not find a defined cell.  That's okay.
        $ti1 = $ti+1;
        while ( $ti1 < $time_count and not defined $build_table->[$ti1][$bi] ) {
          $ti1++;
        }
        if (defined($br1 = $build_table->[$ti1][$bi])) {
          $br->{previousbuildtime} = $br1->{buildtime};
        }

        $ti1 = $ti-1;

        if ( $br->{buildstatus} eq 'building' or not $display_accurate_build_end_times ) {
          # If the current record represents a system that's still building,
          # we'll use the old style and let the build window "slide" up to the
          # next defined build record.
          while ( $ti1 >= 0 and not defined $build_table->[$ti1][$bi] ) {
            $build_table->[$ti1][$bi] = -1;
            $ti1--;
          }
        } else {
          # If the current record has a non 'building' status, we stop the
          # build window at its "endtime".
          while ( $ti1 >= 0 and not defined $build_table->[$ti1][$bi]
                  and $build_time_times->[$ti1] < $br->{endtime} ) {
            $build_table->[$ti1][$bi] = -1;
            $ti1--;
          }
        }

        if ($ti1 > 0 and defined($br1 = $build_table->[$ti1][$bi])) {
          $br->{nextbuildtime} = $br1->{buildtime};
        }

        $br->{rowspan} = $ti - $ti1;
        unless ($br->{rowspan} == 1) {
          $build_table->[$ti1+1][$bi] = $br;
          $build_table->[$ti][$bi] = -1;
        }
      }
    }
  }

  # After we've filled out $build_table with all of the applicable build
  # records, we can do another pass over the table and fill in all the other
  # pieces with null records to pretty up our output.  Ain't that fanciful.
  #
  # When we're not using the optional "show real end times" feature, this
  # section of code is a no-op.  Every cell will be defined, either with a
  # real $br or being set to -1.

  for ($bi = $name_count - 1; $bi >= 0; $bi--) {
    for ($ti = $time_count - 1; $ti >= 0; $ti--) {
      if (not defined($build_table->[$ti][$bi])) {
        my $ti1 = $ti;
        while ( $ti1 >= 0 and not defined $build_table->[$ti1][$bi] ) {
          $build_table->[$ti1][$bi] = -1;
          $ti1--;
        }

        my $null_record_br = {};
        $null_record_br->{buildstatus} = "null";
        $null_record_br->{rowspan} = $ti - $ti1;
        $build_table->[$ti1+1][$bi] = $null_record_br;
      }
    }
  }
}

sub load_notes {
  my $treedata = $_[0];

  open(NOTES,"<$treedata->{name}/notes.txt") 
    or print "<h2>warning: Couldn't open $treedata->{name}/notes.txt </h2>\n";
  while (<NOTES>) {
    chop;
    my ($nbuildtime,$nbuildname,$nwho,$nnow,$nenc_note) = split /\|/;
    my $ti = $build_time_index->{$nbuildtime};
    my $bi = $build_name_index->{$nbuildname};

    if (defined $ti and defined $bi) {
      $build_table->[$ti][$bi]->{hasnote} = 1;
      unless (defined $build_table->[$ti][$bi]->{noteid}) {
        $build_table->[$ti][$bi]->{noteid} = $#note_array + 1;
      }
      $noteid = $build_table->[$ti][$bi]->{noteid};
      $now_str = &print_time($nnow);
      $note = &url_decode($nenc_note);
      
      $note_array[$noteid] = '' unless $note_array[$noteid];
      $note_array[$noteid] = "<pre>\n[<b><a href=mailto:$nwho>"
        ."$nwho</a> - $now_str</b>]\n$note\n</pre>"
        .$note_array[$noteid];
    }
  }
  close NOTES;
}

sub split_cgi_args {
    local (@args, $pair, $key, $value, $s);
    
    if ($ENV{"REQUEST_METHOD"} eq 'POST') {
        $s .= $_ while (<>);
    }
    else {
        $s = $ENV{"QUERY_STRING"};
    }
    
    $s =~ tr/+/ /;
    @args= split(/\&/, $s );
    
    for $pair (@args) {
        ($key, $value) = split(/=/, $pair);
        $key   =~ s/%([a-fA-F0-9]{2})/pack("C", hex($1))/eg;
        $value =~ s/%([a-fA-F0-9]{2})/pack("C", hex($1))/eg;
        $form{$key} = $value;
    }
    
    # extract the cookies from the HTTP_COOKIE environment 
    %cookie_jar = split('[;=] *',$ENV{'HTTP_COOKIE'});
}

sub make_cgi_args {
    local($k,$v,$ret);
    for $k (sort keys %form){
        $ret .= ($ret eq "" ? '?' : '&');
        $v = $form{$k};
        $ret .= &url_encode($k);
        $ret .= '=';
        $ret .= &url_encode($v);
    }
    return $ret;
}

@weekdays = ('Sun','Mon','Tue','Wed','Thu','Fri','Sat');
@months = ('Jan','Feb','Mar','Apr','May','Jun',
           'Jul','Aug','Sep','Oct','Nov','Dec');

sub toGMTString {
    local ($seconds) = $_[0];

    local ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst)
        = gmtime($seconds);
    $year += 1900;

    sprintf('%s, %02d-%s-%d %02d:%02d:%02d GMT',
            $weekdays[$wday],$mday,$months[$mon],$year,$hour,$min,$sec);
}