' # That part of the <title>

#!/usr/bin/perl

$infile = 'refered_to.dat'; # this was the captured output of the java program
$outfile = 'apache_rewrites.conf'; # here's where the rewrites are going
$archivedir = 'writeon/archives'; # this is the directory where the copies
                                  # of the original individual archive html is

# my MT constructs titles as 'AkuAku SF: <title>'
# That part of the <title></title> data needs to be stripped
$blogname = 'AkuAku SF';

open( IN, $infile ) || die "cannot open '$infile'\n";
open( OUT, ">$outfile") || die "cannot open '$outfile'";

while (<IN>) {
  ($count, $nada, $post) = split;
  #print "$count $post\n";
  if ($post =~ /([\d]+\.shtml)/) {
     $entry = $1;
     #print "$count $entry\n";

     if (open( HTML, "$archivedir/$entry")) {
        ($title, $date) = parseHTML( HTML );
        if (defined $title && defined $date) {
          #print "title is $title\n";
          $newurl = getNewUrl($title, $date);
          #print "newurl is $newurl\n";

          print "RewriteRule ^$post\$ http://AkuAku.org/archives/$newurl\n";
        } else {
          #print "date or title undefined\n";
        }
        close HTML;
     } else {
        #print "could not open '$archivedir/$entry'\n";
     } 
  }
}

close OUT;
close IN;
exit;

sub parseHTML( HTML ) {
  my $title, $date;
  while (<HTML>) {
    if ($_ =~ /<title>(.*)<\/title>/) {
      $title = $1;
      $title =~ s/$blogname: //;
    }
    if ($_ =~ /dc:date="(\d{4})-(\d{2})-\d{2}/) {
      $date = "$1/$2";
    }
  }
  ($title, $date);
}


# the following is a lot of educated guesswork
sub getNewUrl( $title, $date ) {
  $newurl = $title;
  $newurl =~ s/[:\.\@=\-,'";\?\(\)]//g;
  $newurl =~ s/ {2,}/ /g;
  $newurl =~ s/ /_/g;
  $newurl =~ tr/[A-Z]/[a-z]/;
  $newurl = substr( $newurl, 0, 15);
  $newurl =~ s/^_*//;
  $newurl =~ s/_*$//;
  $newurl = $date . '/' . $newurl . '.shtml';
}