grutatxt

   1 #!/usr/bin/perl
   2
   3 use lib '.';
   4
   5 #
   6 # grutatxt - Text to HTML (and other formats) converter
   7 # http://www.triptico.com/software/grutatxt.html
   8 #
   9 # Copyright (C) 2000/2007 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 #####################################################################
  72
  73 # parse options
  74 if (!GetOptions('i|input=s'             =>      \$input_file,
  75                 'o|output=s'            =>      \$output_file,
  76                 'c|css=s'               =>      \$css,
  77                 'e|embed-css'           =>      \$embed_css,
  78                 't|title=s'             =>      \$title,
  79                 'f|header-offset=s'     =>      \$header_offset,
  80                 'b|table-headers'       =>      \$table_headers,
  81                 'ct|center-tables'      =>      \$center_tables,
  82                 'xt|expand-tables'      =>      \$expand_tables,
  83                 'sp|strip-parens'       =>      \$strip_parens,
  84                 'ts|tabsize=s'          =>      \$tabsize,
  85                 'nb|no-body'            =>      \$no_body,
  86                 'v|version'             =>      \$version,
  87                 'h|help'                =>      \$usage,
  88                 'm|mode=s'              =>      \$mode,
  89                 's|man-section=s'       =>      \$man_section,
  90                 'docclass=s'            =>      \$latex_docclass,
  91                 'papersize=s'           =>      \$papersize,
  92                 'encoding=s'            =>      \$encoding,
  93                 'dl'                    =>      \$dl_as_dl,
  94                 'no-time-sig'           =>      \$no_time_sig)
  95               or $usage) {
  96         usage();
  97 }
  98
  99 if ($version) {
 100         print "$VERSION\n"; exit(0);
 101 }
 102
 103 open I, $input_file or die "Can't open $input_file: $!";
 104 open O, ">$output_file" or die "Can't create $output_file: $!";
 105
 106 # if utf8 encoding is wanted, set the filehandles as utf8
 107 # so that regular expressions match all characters
 108 # (this is crap)
 109 if (defined($encoding) && $encoding =~ /^utf-?8/i) {
 110         binmode(I, ":utf8");
 111         binmode(O, ":utf8");
 112 }
 113
 114 $content = join('',<I>);
 115 close I;
 116
 117 $content_title = '';
 118
 119 # make tab to space conversion only in LaTeX mode
 120 $tabsize = 0 unless $mode =~ /^latex$/i;
 121
 122 $grutatxt = new Grutatxt(
 123                 'mode'          => $mode,
 124                 'header-offset' => $header_offset,
 125                 'table-headers' => $table_headers,
 126                 'center-tables' => $center_tables,
 127                 'expand-tables' => $expand_tables,
 128                 'strip-parens'  => $strip_parens,
 129                 'strip-dollars' => $strip_dollars,
 130                 'tabsize'       => $tabsize,
 131                 'dl-as-dl'      => $dl_as_dl,
 132                 'table-type'    => $table_type,
 133                 'title'         => \$content_title,
 134                 'abstract'      => \$abstract,
 135                 'page-name'     => $title,
 136                 'section'       => $man_section,
 137                 'docclass'      => $latex_docclass,
 138                 'papersize'     => $papersize,
 139                 'encoding'      => $encoding
 140                 );
 141
 142 @result = $grutatxt->process($content);
 143
 144 $title = $content_title unless $title;
 145 $no_body = 1 unless $mode =~ /^html$/i;
 146
 147 unless ($no_body) {
 148         print O "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.0 Transitional//EN\"\n";
 149         print O "       \"http://www.w3.org/TR/REC-html40/loose.dtd\">\n";
 150         print O "<html><head>\n";
 151         print O "<meta http-equiv='Content-Type' content='text/html; charset=" .
 152                 ($encoding || 'iso-8859-1') . "'>\n";
 153         print O "<title>$title</title>\n";
 154
 155         printf O "<!-- converted from text by grutatxt $VERSION on %s -->\n", scalar(localtime)
 156                 unless $no_time_sig;
 157
 158         if ($css) {
 159                 if ($embed_css) {
 160                         if (open C, $css) {
 161                                 my ($c) = join('', <C>);
 162                                 close C;
 163
 164                                 print O "<style type='text/css'>\n";
 165                                 print O $c . "\n";
 166                                 print O "</style>\n";
 167                         }
 168                         else {
 169                                 die "Can't open '$css' CSS file.";
 170                         }
 171                 }
 172                 else {
 173                         print O "<link rel=StyleSheet href='$css' type='text/css'>";
 174                 }
 175         }
 176
 177         print O "</head><body>\n";
 178 }
 179
 180 foreach my $l (@result) {
 181         print O "$l\n";
 182 }
 183
 184 print O "</body></html>\n" unless $no_body;
 185
 186 close O;
 187
 188 exit(0);
 189
 190
 191 sub usage
 192 {
 193         print "grutatxt $VERSION - Grutatxt format processor\n";
 194         print "Copyright (C) 2000/2007 Angel Ortega <angel\@triptico.com>\n";
 195         print "This software is covered by the GPL license. NO WARRANTY.\n\n";
 196
 197         print "Usage:\n";
 198         print "\n";
 199         print "grutatxt [options] < input_text_file > output_html_file\n";
 200         print "\n";
 201         print "Global options:\n\n";
 202         print "    -i|--input=FILE            Input file (STDIN)\n";
 203         print "    -o|--output=FILE           Output file (STDOUT)\n";
 204         print "    -t|--title=TITLE           Document title (if unset,\n";
 205         print "                               level 1 heading is used)\n";
 206         print "    -sp|--strip-parens         Strip parentheses in function\n";
 207         print "                               names (shown monospaced anyway)\n";
 208         print "    -sd|--strip-dollars        Strip leading \$ in variable\n";
 209         print "                               names (shown monospaced anyway)\n";
 210         print "    -m|--mode=[HTML|troff|man|latex]\n";
 211         print "                               Output mode: HTML, troff, man or latex\n";
 212         print "                               (default: HTML)\n";
 213         print "    --no-time-sig              Avoid time signature in HTML comment\n";
 214         print "HTML options:\n\n";
 215         print "    -c|--css=CSS_URL_OR_FILE   CSS URL (or file if using --embed-css)\n";
 216         print "    -e|--embed-css             Embed CSS instead of linking to it\n";
 217         print "    -f|--header-offset=NUMBER  Offset to add to <h1>,\n";
 218         print "                               <h2>... headers (default 0)\n";
 219         print "    -b|--table-headers         Use <th> instead of <td> in\n";
 220         print "                               the first row of each table\n";
 221         print "    -ct|--center-tables        Centers the tables\n";
 222         print "    -xt|--expand-tables        Expands the tables (width=100\%)\n";
 223         print "    -nb|-no-body               Don't generate <html><body>...\n";
 224         print "                               </body></html> enclosing.\n";
 225         print "    --encoding=ENCODING        Character encoding (default: iso-8859-1)\n";
 226         print "    -dl                        Use real <dl>, <dd> and <dt>\n";
 227         print "                               instead of tables in definition lists.\n";
 228         print "troff options:\n\n";
 229         print "    --table-type=TYPE          Table type. Possible values:\n";
 230         print "                               box, allbox, doublebox (default allbox)\n";
 231         print "man options:\n\n";
 232         print "    -s|--man-section=SECTION   Man page section (default: 1)\n\n";
 233         print "LaTeX options:\n\n";
 234         print "    --docclass=CLASS           Document class (default: report)\n";
 235         print "    --papersize=SIZE           Paper size (default: a4paper)\n";
 236         print "    --encoding=ENCODING        Character encoding (default: latin1)\n";
 237         print "    -ts|--tabsize=NUMBER       Tab size for tab to space conversions in\n";
 238         print "                               LaTeX verbatim environment (default: 8)\n";
 239
 240         exit(0);
 241 }