doc/texi2pod.pl

   1 #! /usr/bin/perl -w
   2
   3 #   Copyright (C) 1999, 2000, 2001 Free Software Foundation, Inc.
   4
   5 # This file is part of GNU CC.
   6
   7 # GNU CC is free software; you can redistribute it and/or modify
   8 # it under the terms of the GNU General Public License as published by
   9 # the Free Software Foundation; either version 2, or (at your option)
  10 # any later version.
  11
  12 # GNU CC is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 # GNU General Public License for more details.
  16
  17 # You should have received a copy of the GNU General Public License
  18 # along with GNU CC; see the file COPYING.  If not, write to
  19 # the Free Software Foundation, 51 Franklin Street, Fifth Floor,
  20 # Boston, MA 02110-1301 USA
  21
  22 # This does trivial (and I mean _trivial_) conversion of Texinfo
  23 # markup to Perl POD format.  It's intended to be used to extract
  24 # something suitable for a manpage from a Texinfo document.
  25
  26 $output = 0;
  27 $skipping = 0;
  28 %sects = ();
  29 @sects_sequence = ();
  30 $section = "";
  31 @icstack = ();
  32 @endwstack = ();
  33 @skstack = ();
  34 @instack = ();
  35 $shift = "";
  36 %defs = ();
  37 $fnno = 1;
  38 $inf = "";
  39 @ibase = ();
  40
  41 while ($_ = shift) {
  42     if (/^-D(.*)$/) {
  43         if ($1 ne "") {
  44             $flag = $1;
  45         } else {
  46             $flag = shift;
  47         }
  48         $value = "";
  49         ($flag, $value) = ($flag =~ /^([^=]+)(?:=(.+))?/);
  50         die "no flag specified for -D\n"
  51             unless $flag ne "";
  52         die "flags may only contain letters, digits, hyphens, dashes and underscores\n"
  53             unless $flag =~ /^[a-zA-Z0-9_-]+$/;
  54         $defs{$flag} = $value;
  55     } elsif (/^-I(.*)$/) {
  56         push @ibase, $1 ne "" ? $1 : shift;
  57     } elsif (/^-/) {
  58         usage();
  59     } else {
  60         $in = $_, next unless defined $in;
  61         $out = $_, next unless defined $out;
  62         usage();
  63     }
  64 }
  65
  66 push @ibase, ".";
  67
  68 if (defined $in) {
  69     $inf = gensym();
  70     open($inf, "<$in") or die "opening \"$in\": $!\n";
  71     push @ibase, $1 if $in =~ m|^(.+)/[^/]+$|;
  72 } else {
  73     $inf = \*STDIN;
  74 }
  75
  76 if (defined $out) {
  77     open(STDOUT, ">$out") or die "opening \"$out\": $!\n";
  78 }
  79
  80 while(defined $inf) {
  81 INF: while(<$inf>) {
  82     # Certain commands are discarded without further processing.
  83     /^\@(?:
  84          [a-z]+index            # @*index: useful only in complete manual
  85          |need                  # @need: useful only in printed manual
  86          |(?:end\s+)?group      # @group .. @end group: ditto
  87          |page                  # @page: ditto
  88          |node                  # @node: useful only in .info file
  89          |(?:end\s+)?ifnottex   # @ifnottex .. @end ifnottex: use contents
  90         )\b/x and next;
  91
  92     chomp;
  93
  94     # Look for filename and title markers.
  95     /^\@setfilename\s+([^.]+)/ and $fn = $1, next;
  96     /^\@settitle\s+([^.]+)/ and $tl = postprocess($1), next;
  97
  98     # Identify a man title but keep only the one we are interested in.
  99     /^\@c\s+man\s+title\s+([A-Za-z0-9-]+)\s+(.+)/ and do {
 100         if (exists $defs{$1}) {
 101             $fn = $1;
 102             $tl = postprocess($2);
 103         }
 104         next;
 105     };
 106
 107     /^\@include\s+(.+)$/ and do {
 108         push @instack, $inf;
 109         $inf = gensym();
 110
 111         for (@ibase) {
 112             open($inf, "<" . $_ . "/" . $1) and next INF;
 113         }
 114         die "cannot open $1: $!\n";
 115     };
 116
 117     # Look for blocks surrounded by @c man begin SECTION ... @c man end.
 118     # This really oughta be @ifman ... @end ifman and the like, but such
 119     # would require rev'ing all other Texinfo translators.
 120     /^\@c\s+man\s+begin\s+([A-Za-z ]+)/ and $sect = $1, push (@sects_sequence, $sect), $output = 1, next;
 121     /^\@c\s+man\s+end/ and do {
 122         $sects{$sect} = "" unless exists $sects{$sect};
 123         $sects{$sect} .= postprocess($section);
 124         $section = "";
 125         $output = 0;
 126         next;
 127     };
 128
 129     # handle variables
 130     /^\@set\s+([a-zA-Z0-9_-]+)\s*(.*)$/ and do {
 131         $defs{$1} = $2;
 132         next;
 133     };
 134     /^\@clear\s+([a-zA-Z0-9_-]+)/ and do {
 135         delete $defs{$1};
 136         next;
 137     };
 138
 139     next unless $output;
 140
 141     # Discard comments.  (Can't do it above, because then we'd never see
 142     # @c man lines.)
 143     /^\@c\b/ and next;
 144
 145     # End-block handler goes up here because it needs to operate even
 146     # if we are skipping.
 147     /^\@end\s+([a-z]+)/ and do {
 148         # Ignore @end foo, where foo is not an operation which may
 149         # cause us to skip, if we are presently skipping.
 150         my $ended = $1;
 151         next if $skipping && $ended !~ /^(?:ifset|ifclear|ignore|menu|iftex)$/;
 152
 153         die "\@end $ended without \@$ended at line $.\n" unless defined $endw;
 154         die "\@$endw ended by \@end $ended at line $.\n" unless $ended eq $endw;
 155
 156         $endw = pop @endwstack;
 157
 158         if ($ended =~ /^(?:ifset|ifclear|ignore|menu|iftex)$/) {
 159             $skipping = pop @skstack;
 160             next;
 161         } elsif ($ended =~ /^(?:example|smallexample|display)$/) {
 162             $shift = "";
 163             $_ = "";        # need a paragraph break
 164         } elsif ($ended =~ /^(?:itemize|enumerate|(?:multi|[fv])?table)$/) {
 165             $_ = "\n=back\n";
 166             $ic = pop @icstack;
 167         } else {
 168             die "unknown command \@end $ended at line $.\n";
 169         }
 170     };
 171
 172     # We must handle commands which can cause skipping even while we
 173     # are skipping, otherwise we will not process nested conditionals
 174     # correctly.
 175     /^\@ifset\s+([a-zA-Z0-9_-]+)/ and do {
 176         push @endwstack, $endw;
 177         push @skstack, $skipping;
 178         $endw = "ifset";
 179         $skipping = 1 unless exists $defs{$1};
 180         next;
 181     };
 182
 183     /^\@ifclear\s+([a-zA-Z0-9_-]+)/ and do {
 184         push @endwstack, $endw;
 185         push @skstack, $skipping;
 186         $endw = "ifclear";
 187         $skipping = 1 if exists $defs{$1};
 188         next;
 189     };
 190
 191     /^\@(ignore|menu|iftex)\b/ and do {
 192         push @endwstack, $endw;
 193         push @skstack, $skipping;
 194         $endw = $1;
 195         $skipping = 1;
 196         next;
 197     };
 198
 199     next if $skipping;
 200
 201     # Character entities.  First the ones that can be replaced by raw text
 202     # or discarded outright:
 203     s/\@copyright\{\}/(c)/g;
 204     s/\@dots\{\}/.../g;
 205     s/\@enddots\{\}/..../g;
 206     s/\@([.!? ])/$1/g;
 207     s/\@[:-]//g;
 208     s/\@bullet(?:\{\})?/*/g;
 209     s/\@TeX\{\}/TeX/g;
 210     s/\@pounds\{\}/\#/g;
 211     s/\@minus(?:\{\})?/-/g;
 212     s/\\,/,/g;
 213
 214     # Now the ones that have to be replaced by special escapes
 215     # (which will be turned back into text by unmunge())
 216     s/&/&amp;/g;
 217     s/\@\{/&lbrace;/g;
 218     s/\@\}/&rbrace;/g;
 219     s/\@\@/&at;/g;
 220
 221     # Inside a verbatim block, handle @var specially.
 222     if ($shift ne "") {
 223         s/\@var\{([^\}]*)\}/<$1>/g;
 224     }
 225
 226     # POD doesn't interpret E<> inside a verbatim block.
 227     if ($shift eq "") {
 228         s/</&lt;/g;
 229         s/>/&gt;/g;
 230     } else {
 231         s/</&LT;/g;
 232         s/>/&GT;/g;
 233     }
 234
 235     # Single line command handlers.
 236
 237     /^\@(?:section|unnumbered|unnumberedsec|center|heading)\s+(.+)$/
 238         and $_ = "\n=head2 $1\n";
 239     /^\@(?:subsection|subheading)\s+(.+)$/
 240         and $_ = "\n=head3 $1\n";
 241     /^\@(?:subsubsection|subsubheading)\s+(.+)$/
 242         and $_ = "\n=head4 $1\n";
 243
 244     # Block command handlers:
 245     /^\@itemize\s*(\@[a-z]+|\*|-)?/ and do {
 246         push @endwstack, $endw;
 247         push @icstack, $ic;
 248         $ic = $1 ? $1 : "*";
 249         $_ = "\n=over 4\n";
 250         $endw = "itemize";
 251     };
 252
 253     /^\@enumerate(?:\s+([a-zA-Z0-9]+))?/ and do {
 254         push @endwstack, $endw;
 255         push @icstack, $ic;
 256         if (defined $1) {
 257             $ic = $1 . ".";
 258         } else {
 259             $ic = "1.";
 260         }
 261         $_ = "\n=over 4\n";
 262         $endw = "enumerate";
 263     };
 264
 265     /^\@((?:multi|[fv])?table)\s+(\@[a-z]+)/ and do {
 266         push @endwstack, $endw;
 267         push @icstack, $ic;
 268         $endw = $1;
 269         $ic = $2;
 270         $ic =~ s/\@(?:samp|strong|key|gcctabopt|option|env)/B/;
 271         $ic =~ s/\@(?:code|kbd)/C/;
 272         $ic =~ s/\@(?:dfn|var|emph|cite|i)/I/;
 273         $ic =~ s/\@(?:file)/F/;
 274         $ic =~ s/\@(?:columnfractions)//;
 275         $_ = "\n=over 4\n";
 276     };
 277
 278     /^\@((?:small)?example|display)/ and do {
 279         push @endwstack, $endw;
 280         $endw = $1;
 281         $shift = "\t";
 282         $_ = "";        # need a paragraph break
 283     };
 284
 285     /^\@item\s+(.*\S)\s*$/ and $endw eq "multitable" and do {
 286         my $columns = $1;
 287         $columns =~ s/\@tab/ : /;
 288
 289         $_ = "\n=item B&LT;". $columns ."&GT;\n";
 290     };
 291
 292     /^\@tab\s+(.*\S)\s*$/ and $endw eq "multitable" and do {
 293         my $columns = $1;
 294         $columns =~ s/\@tab/ : /;
 295
 296         $_ = " : ". $columns;
 297         $section =~ s/\n+\s+$//;
 298     };
 299
 300     /^\@itemx?\s*(.+)?$/ and do {
 301         if (defined $1) {
 302             # Entity escapes prevent munging by the <> processing below.
 303             $_ = "\n=item $ic\&LT;$1\&GT;\n";
 304         } else {
 305             $_ = "\n=item $ic\n";
 306             $ic =~ y/A-Ya-y/B-Zb-z/;
 307             $ic =~ s/(\d+)/$1 + 1/eg;
 308         }
 309     };
 310
 311     $section .= $shift.$_."\n";
 312 }
 313 # End of current file.
 314 close($inf);
 315 $inf = pop @instack;
 316 }
 317
 318 die "No filename or title\n" unless defined $fn && defined $tl;
 319
 320 $sects{NAME} = "$fn \- $tl\n";
 321 $sects{FOOTNOTES} .= "=back\n" if exists $sects{FOOTNOTES};
 322
 323 unshift @sects_sequence, "NAME";
 324 for $sect (@sects_sequence) {
 325     if(exists $sects{$sect}) {
 326         $head = $sect;
 327         $head =~ s/SEEALSO/SEE ALSO/;
 328         print "=head1 $head\n\n";
 329         print scalar unmunge ($sects{$sect});
 330         print "\n";
 331     }
 332 }
 333
 334 sub usage
 335 {
 336     die "usage: $0 [-D toggle...] [infile [outfile]]\n";
 337 }
 338
 339 sub postprocess
 340 {
 341     local $_ = $_[0];
 342
 343     # @value{foo} is replaced by whatever 'foo' is defined as.
 344     while (m/(\@value\{([a-zA-Z0-9_-]+)\})/g) {
 345         if (! exists $defs{$2}) {
 346             print STDERR "Option $2 not defined\n";
 347             s/\Q$1\E//;
 348         } else {
 349             $value = $defs{$2};
 350             s/\Q$1\E/$value/;
 351         }
 352     }
 353
 354     # Formatting commands.
 355     # Temporary escape for @r.
 356     s/\@r\{([^\}]*)\}/R<$1>/g;
 357     s/\@(?:dfn|var|emph|cite|i)\{([^\}]*)\}/I<$1>/g;
 358     s/\@(?:code|kbd)\{([^\}]*)\}/C<$1>/g;
 359     s/\@(?:gccoptlist|samp|strong|key|option|env|command|b)\{([^\}]*)\}/B<$1>/g;
 360     s/\@sc\{([^\}]*)\}/\U$1/g;
 361     s/\@file\{([^\}]*)\}/F<$1>/g;
 362     s/\@w\{([^\}]*)\}/S<$1>/g;
 363     s/\@(?:dmn|math)\{([^\}]*)\}/$1/g;
 364
 365     # Cross references are thrown away, as are @noindent and @refill.
 366     # (@noindent is impossible in .pod, and @refill is unnecessary.)
 367     # @* is also impossible in .pod; we discard it and any newline that
 368     # follows it.  Similarly, our macro @gol must be discarded.
 369
 370     s/\@anchor{(?:[^\}]*)\}//g;
 371     s/\(?\@xref\{(?:[^\}]*)\}(?:[^.<]|(?:<[^<>]*>))*\.\)?//g;
 372     s/\s+\(\@pxref\{(?:[^\}]*)\}\)//g;
 373     s/;\s+\@pxref\{(?:[^\}]*)\}//g;
 374     s/\@ref\{([^\}]*)\}/$1/g;
 375     s/\@noindent\s*//g;
 376     s/\@refill//g;
 377     s/\@gol//g;
 378     s/\@\*\s*\n?//g;
 379
 380     # @uref can take one, two, or three arguments, with different
 381     # semantics each time.  @url and @email are just like @uref with
 382     # one argument, for our purposes.
 383     s/\@(?:uref|url|email)\{([^\},]*)\}/&lt;B<$1>&gt;/g;
 384     s/\@uref\{([^\},]*),([^\},]*)\}/$2 (C<$1>)/g;
 385     s/\@uref\{([^\},]*),([^\},]*),([^\},]*)\}/$3/g;
 386
 387     # Turn B<blah I<blah> blah> into B<blah> I<blah> B<blah> to
 388     # match Texinfo semantics of @emph inside @samp.  Also handle @r
 389     # inside bold.
 390     s/&LT;/</g;
 391     s/&GT;/>/g;
 392     1 while s/B<((?:[^<>]|I<[^<>]*>)*)R<([^>]*)>/B<$1>${2}B</g;
 393     1 while (s/B<([^<>]*)I<([^>]+)>/B<$1>I<$2>B</g);
 394     1 while (s/I<([^<>]*)B<([^>]+)>/I<$1>B<$2>I</g);
 395     s/[BI]<>//g;
 396     s/([BI])<(\s+)([^>]+)>/$2$1<$3>/g;
 397     s/([BI])<([^>]+?)(\s+)>/$1<$2>$3/g;
 398
 399     # Extract footnotes.  This has to be done after all other
 400     # processing because otherwise the regexp will choke on formatting
 401     # inside @footnote.
 402     while (/\@footnote/g) {
 403         s/\@footnote\{([^\}]+)\}/[$fnno]/;
 404         add_footnote($1, $fnno);
 405         $fnno++;
 406     }
 407
 408     return $_;
 409 }
 410
 411 sub unmunge
 412 {
 413     # Replace escaped symbols with their equivalents.
 414     local $_ = $_[0];
 415
 416     s/&lt;/E<lt>/g;
 417     s/&gt;/E<gt>/g;
 418     s/&lbrace;/\{/g;
 419     s/&rbrace;/\}/g;
 420     s/&at;/\@/g;
 421     s/&amp;/&/g;
 422     return $_;
 423 }
 424
 425 sub add_footnote
 426 {
 427     unless (exists $sects{FOOTNOTES}) {
 428         $sects{FOOTNOTES} = "\n=over 4\n\n";
 429     }
 430
 431     $sects{FOOTNOTES} .= "=item $fnno.\n\n"; $fnno++;
 432     $sects{FOOTNOTES} .= $_[0];
 433     $sects{FOOTNOTES} .= "\n\n";
 434 }
 435
 436 # stolen from Symbol.pm
 437 {
 438     my $genseq = 0;
 439     sub gensym
 440     {
 441         my $name = "GEN" . $genseq++;
 442         my $ref = \*{$name};
 443         delete $::{$name};
 444         return $ref;
 445     }
 446 }