#!/usr/bin/perl

# Fix up html files after they've been extracted from .lit books or
# .doc files.
# Strip out all the formatting so that what's left is plain-jane html.

$goout = "";
$lastline = "";
while ($line = <>) {
  chop $line;

  $line = $lastline . $line;
  # $a = chr(128);
  # $b = chr(255);
  # if ($line =~ /^.*([$a-$b]).*$/) {
  #  print STDERR "here: $1, " . ord($1) . ", $line\n";
  # }
  $line = replace_pattern ($line,"<[pP][^>]*>","<p>")
    if ($line =~ /<[pP]/);
  $line = replace_pattern ($line,"<style[^>]*>.*<\/style[^>]*>","")
    if ($line =~ /^.*<style.*$/);
  $line = replace_pattern ($line,"<span[^>]*>","")
    if ($line =~ /^.*<span.*$/);
  $line = replace_pattern ($line,"<\/span[^>]*>","")
    if ($line =~ /<\/span/);
  $line = replace_pattern ($line,"<\/div[^>]*>","")
    if ($line =~ /<\/div/);
  $line = replace_pattern ($line,"<div[^>]*>","")
    if ($line =~ /<div/);
  $line = replace_pattern ($line,"<![^>]*>","")
    if ($line =~ /<!/) ;
  $ch = chr(173);
  $line =~ s/$ch//g;
  $ch = chr(146);
  $line =~ s/$ch/'/g;
  $ch = chr(147);
  $line =~ s/$ch/"/g;
  $ch = chr(148);
  $line =~ s/$ch/"/g;
  $ch = chr(151);
  $line =~ s/$ch/ -- /g;
  $line =~ s/<o:p>//g;
  $line =~ s/<\/o:p>//g;
  $line =~ s/\&quot;/\"/g;
  $line =~ s/\&#822[012];/\"/g;
  $line =~ s/\&#821[67];/\'/g;
  $line =~ s/\&#8212;/ -- /g;
  $line =~ s/\&#8230;/.../g;
  $line =~ s/\&#8209;/-/g;
  #$line =~ s/\&#8482;/trademark/g;
  #$line =~ s/\&#174;/registered trademark/g;
  #$line =~ s/\&#169;/copyright/g;
  $line =~ s/\&#160;/\&nbsp;/g;
  $line =~ s/\&#173;//g;
  $line =~ s/\r//g;
  undef $lastline;
  if ($line =~ /^<[pP]>$/) {
    $lastline = $line;
    next;
  }
  print "$line\n";
}
print "$lastline\n" if $lastline;

exit 0;

sub replace_pattern {
  my ($line,$strip,$replace) = @_;

  my $fetch = 1;
  while (!($line =~ /^.*$strip.*$/) && $fetch) {
    $fetch = <>;
    $line .= " " . $fetch;
    chomp $line;
  }
  $line =~ s/$strip/$replace/g;
  return $line;
}