ascii2txt.pl

   1 #!/bin/perl
   2
   3 # Copyright (c) 2010,2011,2012 Todd T. Fries <todd@fries.net>
   4 #
   5 # Permission to use, copy, modify, and distribute this software for any
   6 # purpose with or without fee is hereby granted, provided that the above
   7 # copyright notice and this permission notice appear in all copies.
   8 #
   9 # THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  10 # WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  11 # MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
  12 # ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  13 # WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  14 # ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  15 # OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  16
  17 # read in 'mandoc -Tascii' formatted man pages, spit out txt useful for further
  18 # processing by other utilities
  19
  20 use strict;
  21 use warnings;
  22
  23 our $fileinfo = $ARGV[0];
  24
  25 our $verbose = 0;
  26
  27 my $line;
  28 my @lines;
  29 while(<STDIN>) {
  30         $line = $_;
  31         push @lines,$line;
  32 }
  33 my $oline = "";
  34 my $fmtline = "%s";
  35 foreach $line (@lines) {
  36
  37         my $newline = "";
  38         foreach my $seg (split(/(.\x08.)/,$line)) {
  39                 my $newseg = $seg;
  40                 $newseg =~ m/^(.)\x08(.)$/;
  41                 if (!defined($1) || !defined($2)) {
  42                         $newline .= $seg;
  43                         next;
  44                 }
  45                 if ($1 eq $2) {
  46                         $newline .= "${2}";
  47                         next;
  48                 }
  49                 if ($1 eq "_") {
  50                         $newline .= "${2}";
  51                         next;
  52                 }
  53                 $newline .= $seg;
  54                 next;
  55         }
  56         if ($verbose > 0) {
  57                 printf STDERR "==> text{bf,it}\n   line: <%s>\nnewline: <%s>\n",$line,$newline;
  58         }
  59         $line = $newline;
  60         $line =~ m/(.)\x08/;
  61         if (defined($1)) {
  62                 printf STDERR "Removing %s\\x08\n",$1;
  63         }
  64         $line =~ s/.\x08//g;
  65
  66         # combine adjacent entries
  67         foreach my $macro (("textbf", "textit")) {
  68                 $oline = "";
  69                 while ($oline ne $line) {
  70                         #printf STDERR "combine adjacent\n";
  71                         $oline = $line;
  72                         $line =~ s/\xab\\${macro}\{([^\}]*)\}\xbb\xab\\${macro}\{([^\}]*)\}\xbb/\xab\\${macro}\{$1$2\}\xbb/g;
  73                 }
  74         }
  75         # combine space separated
  76         foreach my $macro (("textbf")) {
  77                 #printf STDERR "combine space\n";
  78                 $oline = "";
  79                 while ($oline ne $line) {
  80                         $oline = $line;
  81                         $line =~ s/\xab\\${macro}\{([^\}]*)\}\xbb[ ]+\xab\\${macro}\{([^\}]*)\}\xbb/\xab\\${macro}\{$1 $2\}\xbb/g;
  82                 }
  83         }
  84
  85         # do the substitution one at a time to be sure to add all man pages, not just the last ones per line.
  86         # XXX provide an exceptions list, audio(9) has mono(1) and stereo(2)
  87         # XXX references, which are _not_ man pages
  88         $oline = "";
  89         while ($oline ne $line) {
  90                 $oline=$line;
  91                 $line =~ s/\{(http|ftp|https):\/\/(.*)\}/ $1:\/\/$2 /;
  92                 if (0) {
  93                 if ($line =~ m/ ([a-z][a-z0-9\.\-\_]*)\(([1-9])\)([,\.\) ])/) {
  94                         my $quote = texquote($1);
  95                         $line =~ s/ ([a-z][a-z0-9\.\-\_]*)\(([1-9])\)([,\.\) ])/ \xab\\man{$quote}{$2}\xbb$3/;
  96                 }
  97
  98                 if ($line =~ m/ ([a-z][a-z0-9\.\-\_]*)\(([1-9])\)$/) {
  99                         my $quote = texquote($1);
 100                         $line =~ s/ ([a-z][a-z0-9\.\-\_]*)\(([1-9])\)$/ \xab\\man{$quote}{$2}\xbb/;
 101                 }
 102                 }
 103         }
 104         my @macros = ("textbf","textit","man","href");
 105         # quote arguments for tex
 106         foreach my $macro (@macros) {
 107                 my $newline = "";
 108                 foreach my $seg (split(/(\xab\\${macro}\{[^\xbb]*\}\xbb)/,$line)) {
 109                         #printf STDERR "quote args\n";
 110                         my $newseg = $seg;
 111                         # check for nesting first; we only want to escape the
 112                         # inner most argument, process nested macro if it has a nested macro
 113                         # since the nested macro won't catch in the other regex cases
 114                         my $foundnest = 0;
 115                         foreach my $nest (@macros) {
 116                                 if ($macro eq $nest) {
 117                                         next;
 118                                 }
 119
 120                                 $newseg =~ m/^\xab\\${macro}\{[ ]*\\${nest}\{([^\xbb]*)\}\{([^\xbb]*)\}\}\xbb$/;
 121                                 if (defined($2)) {
 122                                         $foundnest = 1;
 123                                         $newline .= "\xab\\${macro}\{\\${nest}\{".texquote($1)."\}\{".texquote(${2})."\}\}\xbb";
 124                                         last;
 125                                 }
 126                                 $newseg =~ m/^\xab\\${macro}\{[ ]*\\${nest}\{([^\xbb]*)\}\}\xbb$/;
 127                                 if (defined($1)) {
 128                                         $foundnest = 1;
 129                                         $newline .= "\xab\\${macro}\{\\${nest}\{".texquote($1)."\}\}\xbb";
 130                                         last;
 131                                 }
 132                         }
 133                         if ($foundnest > 0) {
 134                                 next;
 135                         }
 136
 137                         # check for 2 args first
 138                         $newseg =~ m/^\xab\\${macro}\{([^\xbb]*)\}\{([^\xbb]*)\}\xbb$/;
 139                         if (defined($2)) {
 140                                 $newline .= "\xab\\${macro}\{".texquote($1)."\}\{".texquote(${2})."\}\xbb";
 141                                 next;
 142                         }
 143                         $newseg =~ m/^\xab\\${macro}\{([^\xbb]*)\}\xbb$/;
 144                         if (defined($1)) {
 145                                 $newline .= "\xab\\${macro}\{".texquote($1)."\}\xbb";
 146                                 next;
 147                         }
 148                         $newline .= $seg;
 149                 }
 150                 $line = $newline;
 151         }
 152         printf $fmtline,$line;
 153 }
 154
 155 1;
 156
 157 sub texquote {
 158         my ($text) = @_;
 159         my ($ret) = "";
 160         my ($esctest) = "";
 161         my ($escbase) = "BaCkSlAsH";
 162         my ($esccount) = 0;
 163
 164         #$verbose++;
 165         if ($verbose > 0) {
 166                 printf STDERR "\ntexquote: '%s' -> ",$text;
 167         }
 168
 169         if ($text =~ m/\\/) {
 170                 $esctest=sprintf "%s%d",$escbase,$esccount++;
 171                 while ($text =~ m/$esctest/) {
 172                         $esctest=sprintf "%s%d",$escbase,$esccount++;
 173                 }
 174                 $text =~ s/\\/$esctest/g;
 175                 if ($verbose > 0) {
 176                         printf STDERR "'%s' -> ",$text;
 177                 }
 178         }
 179
 180         $text =~ s/([%\{\}_#\&\$\^])/\\$1/g;
 181         $text =~ s/([<>\|\*~])/\{\$$1\$\}/g;
 182
 183         if ($esccount > 0) {
 184                 $text =~ s/$esctest/\$\\backslash\$/g;
 185         }
 186         if ($verbose > 0) {
 187                 printf STDERR "'%s'\n",$text;
 188         }
 189         #$verbose--;
 190
 191         return $text;
 192 }