html.pl

   1 #!/usr/bin/perl -X
   2
   3 # (C) 2016-2017 Tobias Girstmair
   4 # Extracts hypertext formatted news from ORF Teletext
   5 # uses a modified version of vtx2ascii to decode pages
   6 # from dvbtext's spool directory.
   7
   8 # Usage: ./plain.pl <VTX-file>
   9 # Output: HTML
  10
  11 use strict;
  12 use warnings;
  13 use 5.010;
  14 # convert to utf-8 before outputting; WARN: still processes latin-1 (regexes too)!
  15 binmode STDOUT, ":encoding(utf8)";
  16
  17 my %meta;
  18 my $title;
  19 my $text = "";
  20 # TODO: run tzap/dvbtext in background if not already running
  21 my $page = shift;
  22 my ($subpage) = $page=~m/\d{3}_(\d{2})/; #requires ppp_ss file name scheme! @parens: https://stackoverflow.com/a/10034105
  23 $subpage += 0; # convert to number to remove leading zero
  24 # run through vtx2ascii (has been modified to output correct ISO 8859-1 without national replacements)
  25 open (VTX, "./vtx2ascii -a $page |") || die ("Can'r run vtx2ascii");
  26 my $last = "";
  27 my $is_10x = 0;
  28 do {
  29         # transliterate from ETSI EN 300 706 G0 German to latin-1 (ÄÖÜäöüß°§):
  30         tr/[\\]{|}~`@/\N{U+C4}\N{U+D6}\N{U+DC}\N{U+E4}\N{U+F6}\N{U+FC}\N{U+DF}\N{U+B0}\N{U+A7}/;
  31         my $line = $_;
  32         $line =~ s/^\s+|\s+$//g;
  33         chomp ($line);
  34
  35         given ($.) {
  36                 when (1) { %meta = parse_metadata ($line) ; $is_10x = ($meta{'page'}<110) }
  37                 when (2) { $meta{'subres'} = $line }
  38                 when (3) { $meta{'res'} = $line }
  39                 when (4) {}
  40                 when (5 + (1*$is_10x)) { $title = $line }
  41                 when (4 + (1*$is_10x)) {}
  42                 when (4 + (3*$is_10x)) { $title .=$line if ($title eq "")}
  43                 default { $text .= $last . "_EOL_" . ($last eq ""?"":($line eq ""?"<br>":" ")) }
  44         }
  45         $last = $line unless $. == (5+(2*$is_10x));
  46 } while (<VTX>);
  47 $text .= $last;
  48
  49 # substitute hyphenation:
  50 #  * replace with soft hyphen when it splits a word (between lowercase letters;
  51 #    still allows line break when necessary.
  52 #  * keep, when followed by uppercase letter (e.g. "03-Jährige", "PIN-Nummer")
  53 #  * remove in any other case (was: forced hyphenation due to space constraints)
  54 # ad _EOL_: linebreaks already stripped in loop above; wouldn't work either way
  55 # due to single line regex. INFO: Underscore is in DE-localized teletext charset.
  56 $text =~ s/([[:lower:]])-_EOL_ ([[:lower:]])/\1&shy;\2/g;
  57 $text =~ s/([[:alnum:]])-_EOL_ ([[:upper:]])/\1-\2/g;
  58 $text =~ s/_EOL_//g;
  59
  60 # remove ORFText idiosyncrasies
  61 $text =~ s/([[:alnum:]]),([[:alnum:]])/\1, \2/g; # no space after comma to save space
  62 $text =~ s/([[:alnum:]])\.([[:upper:][:digit:]])/\1. \2/g; # no space after period to save space (WARN: breaks URLS like tirol.ORF.at)
  63
  64 # adblocker: (keep it 7bit-ASCII; perl processes latin1, but output will be utf8)
  65 $text =~ s/Kalendarium - t.glich neu \. 734//g;
  66 $text =~ s/>>tirol\. ?ORF\.at//g;
  67
  68 my @tmp = split(' ',$meta{'date'});
  69 my $shortdate = substr $tmp[1], 0, 6;
  70 my $pagesubpage = $meta{'page'} . ($subpage > 0?".$subpage":"");
  71 my $moreinfo = "$meta{'res'} - $meta{'subres'}; $meta{'date'}";
  72 print "<p>$pagesubpage:</span> <b title='$moreinfo' onclick='javascript:alert(\"$moreinfo\")'>$title</b><br>$text</p>";
  73
  74 close (VTX);
  75
  76 sub parse_metadata {
  77         my @elems = split ' ', @_[0];
  78
  79         my %retval = (
  80                 'page' => shift @elems,
  81                 'channel' => shift @elems,
  82                 'date' => join (' ', @elems)
  83         );
  84
  85         return %retval;
  86 }