grutatxt

   1 #!/usr/bin/perl
   2
   3 use lib '.';
   4
   5 #
   6 # grutatxt - Text to HTML (and other formats) converter
   7 # http://triptico.com/software/grutatxt.html
   8 #
   9 # Copyright (C) 2000/2011 Angel Ortega <angel@triptico.com>
  10 #
  11 # This program is free software; you can redistribute it and/or modify
  12 # it under the terms of the GNU General Public License as published by
  13 # the Free Software Foundation; either version 2 of the License, or
  14 # (at your option) any later version.
  15 #
  16 # This program is distributed in the hope that it will be useful,
  17 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  18 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  19 # GNU General Public License for more details.
  20 #
  21 # You should have received a copy of the GNU General Public License
  22 # along with this program; if not, write to the Free Software
  23 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  24 #
  25 # http://www.triptico.com
  26 #
  27
  28 use Getopt::Long;
  29 use locale;
  30 use Grutatxt;
  31
  32 $VERSION = $Grutatxt::VERSION . ':1';
  33
  34 # input file
  35 $input_file = '-';
  36
  37 # output file
  38 $output_file = '>-';
  39
  40 # CSS information
  41 $css = '';
  42 $embed_css = 0;
  43
  44 # page title
  45 $title = '';
  46
  47 # offset for the h? headers
  48 $header_offset = 0;
  49
  50 # default mode
  51 $mode = 'HTML';
  52
  53 # use real dl
  54 $dl_as_dl = 0;
  55
  56 # troff table type
  57 $table_type = 'allbox';
  58
  59 # abstract line number
  60 $abstract = 0;
  61
  62 # man page section
  63 $man_section = 1;
  64
  65 # default tab size in LaTeX mode
  66 $tabsize = 8;
  67
  68 # avoid time signature
  69 $no_time_sig = 0;
  70
  71 # disable pure verbatim mode
  72 $no_pure_verbatim = 0;
  73
  74 # enable TOC
  75 $toc = 0;
  76
  77 #####################################################################
  78
  79 # parse options
  80 if (!GetOptions('i|input=s'             =>      \$input_file,
  81                 'o|output=s'            =>      \$output_file,
  82                 'c|css=s'               =>      \$css,
  83                 'e|embed-css'           =>      \$embed_css,
  84                 't|title=s'             =>      \$title,
  85                 'f|header-offset=s'     =>      \$header_offset,
  86                 'b|table-headers'       =>      \$table_headers,
  87                 'ct|center-tables'      =>      \$center_tables,
  88                 'xt|expand-tables'      =>      \$expand_tables,
  89                 'sp|strip-parens'       =>      \$strip_parens,
  90                 'ts|tabsize=s'          =>      \$tabsize,
  91                 'nb|no-body'            =>      \$no_body,
  92                 'v|version'             =>      \$version,
  93                 'h|help'                =>      \$usage,
  94                 'm|mode=s'              =>      \$mode,
  95                 's|man-section=s'       =>      \$man_section,
  96                 'docclass=s'            =>      \$latex_docclass,
  97                 'papersize=s'           =>      \$papersize,
  98                 'encoding=s'            =>      \$encoding,
  99                 'dl'                    =>      \$dl_as_dl,
 100                 'no-time-sig'           =>      \$no_time_sig,
 101                 'no-pure-verbatim'      =>      \$no_pure_verbatim,
 102                 'toc'                   =>      \$toc
 103         ) or $usage) {
 104         usage();
 105 }
 106
 107 if ($version) {
 108         print "$VERSION\n"; exit(0);
 109 }
 110
 111 open I, $input_file or die "Can't open $input_file: $!";
 112 open O, ">$output_file" or die "Can't create $output_file: $!";
 113
 114 # if utf8 encoding is wanted, set the filehandles as utf8
 115 # so that regular expressions match all characters
 116 # (this is crap)
 117 if (defined($encoding) && $encoding =~ /^utf-?8/i) {
 118         binmode(I, ":utf8");
 119         binmode(O, ":utf8");
 120 }
 121
 122 $content = join('',<I>);
 123 close I;
 124
 125 $content_title = '';
 126
 127 # make tab to space conversion only in LaTeX mode
 128 $tabsize = 0 unless $mode =~ /^latex$/i;
 129
 130 $grutatxt = new Grutatxt(
 131                 'mode'          => $mode,
 132                 'header-offset' => $header_offset,
 133                 'table-headers' => $table_headers,
 134                 'center-tables' => $center_tables,
 135                 'expand-tables' => $expand_tables,
 136                 'strip-parens'  => $strip_parens,
 137                 'strip-dollars' => $strip_dollars,
 138                 'tabsize'       => $tabsize,
 139                 'dl-as-dl'      => $dl_as_dl,
 140                 'table-type'    => $table_type,
 141                 'title'         => \$content_title,
 142                 'abstract'      => \$abstract,
 143                 'page-name'     => $title,
 144                 'section'       => $man_section,
 145                 'docclass'      => $latex_docclass,
 146                 'papersize'     => $papersize,
 147                 'encoding'      => $encoding,
 148                 'no-pure-verbatim' => $no_pure_verbatim,
 149                 'toc'           => $toc
 150                 );
 151
 152 @result = $grutatxt->process($content);
 153
 154 $title = $content_title unless $title;
 155 $no_body = 1 unless $mode =~ /^html$/i;
 156
 157 unless ($no_body) {
 158         print O "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.0 Transitional//EN\"\n";
 159         print O "       \"http://www.w3.org/TR/REC-html40/loose.dtd\">\n";
 160         print O "<html><head>\n";
 161         print O "<meta http-equiv='Content-Type' content='text/html; charset=" .
 162                 ($encoding || 'iso-8859-1') . "'>\n";
 163         print O "<title>$title</title>\n";
 164
 165         printf O "<!-- converted from text by grutatxt $VERSION on %s -->\n", scalar(localtime)
 166                 unless $no_time_sig;
 167
 168         if ($css) {
 169                 if ($embed_css) {
 170                         if (open C, $css) {
 171                                 my ($c) = join('', <C>);
 172                                 close C;
 173
 174                                 print O "<style type='text/css'>\n";
 175                                 print O $c . "\n";
 176                                 print O "</style>\n";
 177                         }
 178                         else {
 179                                 die "Can't open '$css' CSS file.";
 180                         }
 181                 }
 182                 else {
 183                         print O "<link rel=StyleSheet href='$css' type='text/css'>";
 184                 }
 185         }
 186
 187         print O "</head><body>\n";
 188 }
 189
 190 foreach my $l (@result) {
 191         print O "$l\n";
 192 }
 193
 194 print O "</body></html>\n" unless $no_body;
 195
 196 close O;
 197
 198 exit(0);
 199
 200
 201 sub usage
 202 {
 203         print "grutatxt $VERSION - Grutatxt format processor\n";
 204         print "Copyright (C) 2000/2011 Angel Ortega <angel\@triptico.com>\n";
 205         print "This software is covered by the GPL license. NO WARRANTY.\n\n";
 206
 207         print "Usage:\n";
 208         print "\n";
 209         print "grutatxt [options] < input_text_file > output_html_file\n";
 210         print "\n";
 211         print "Global options:\n\n";
 212         print "    -i|--input=FILE            Input file (STDIN)\n";
 213         print "    -o|--output=FILE           Output file (STDOUT)\n";
 214         print "    -t|--title=TITLE           Document title (if unset,\n";
 215         print "                               level 1 heading is used)\n";
 216         print "    -sp|--strip-parens         Strip parentheses in function\n";
 217         print "                               names (shown monospaced anyway)\n";
 218         print "    -sd|--strip-dollars        Strip leading \$ in variable\n";
 219         print "                               names (shown monospaced anyway)\n";
 220         print "    -m|--mode=[HTML|troff|man|latex|rtf]\n";
 221         print "                               Output mode: HTML, troff, man, LaTEX or RTF\n";
 222         print "                               (default: HTML)\n";
 223         print "    --no-time-sig              Avoid time signature in HTML comment\n";
 224         print "    --no-pure-verbatim         Disable pure verbatim mode\n";
 225         print "    --toc                      Add a table of contents after abstract\n\n";
 226         print "HTML options:\n\n";
 227         print "    -c|--css=CSS_URL_OR_FILE   CSS URL (or file if using --embed-css)\n";
 228         print "    -e|--embed-css             Embed CSS instead of linking to it\n";
 229         print "    -f|--header-offset=NUMBER  Offset to add to <h1>,\n";
 230         print "                               <h2>... headers (default 0)\n";
 231         print "    -b|--table-headers         Use <th> instead of <td> in\n";
 232         print "                               the first row of each table\n";
 233         print "    -ct|--center-tables        Centers the tables\n";
 234         print "    -xt|--expand-tables        Expands the tables (width=100\%)\n";
 235         print "    -nb|-no-body               Don't generate <html><body>...\n";
 236         print "                               </body></html> enclosing.\n";
 237         print "    --encoding=ENCODING        Character encoding (default: iso-8859-1)\n";
 238         print "    -dl                        Use real <dl>, <dd> and <dt>\n";
 239         print "                               instead of tables in definition lists.\n";
 240         print "troff options:\n\n";
 241         print "    --table-type=TYPE          Table type. Possible values:\n";
 242         print "                               box, allbox, doublebox (default allbox)\n";
 243         print "man options:\n\n";
 244         print "    -s|--man-section=SECTION   Man page section (default: 1)\n\n";
 245         print "LaTeX options:\n\n";
 246         print "    --docclass=CLASS           Document class (default: report)\n";
 247         print "    --papersize=SIZE           Paper size (default: a4paper)\n";
 248         print "    --encoding=ENCODING        Character encoding (default: latin1)\n";
 249         print "    -ts|--tabsize=NUMBER       Tab size for tab to space conversions in\n";
 250         print "                               LaTeX verbatim environment (default: 8)\n";
 251
 252         exit(1);
 253 }