* arm.md (cbranchsi4): Correct calculation of branch ranges.
[official-gcc.git] / contrib / texi2pod.pl
blobcb8ff466f196f0f22887105fff9db05aa93a0ad4
1 #! /usr/bin/perl -w
3 # This does trivial (and I mean _trivial_) conversion of Texinfo
4 # markup to Perl POD format. It's intended to be used to extract
5 # something suitable for a manpage from a Texinfo document.
7 use v5.6.0;
9 $output = 0;
10 $skipping = 0;
11 %sects = ();
12 $section = "";
13 @icstack = ();
14 @endwstack = ();
15 @skstack = ();
16 $shift = "";
17 %defs = ();
18 $fnno = 1;
20 while ($_ = shift) {
21 if (/^-D(.*)$/) {
22 if ($1 ne "") {
23 $flag = $1;
24 } else {
25 $flag = shift;
27 die "no flag specified for -D\n"
28 unless $flag ne "";
29 die "flags may only contain letters, digits, hyphens, and underscores\n"
30 unless $flag =~ /^[a-zA-Z0-9_-]+$/;
31 $defs{$flag} = "";
32 } elsif (/^-/) {
33 usage();
34 } else {
35 $in = $_, next unless defined $in;
36 $out = $_, next unless defined $out;
37 usage();
41 if (defined $in) {
42 open(STDIN, $in) or die "opening \"$in\": $!\n";
44 if (defined $out) {
45 open(STDOUT, ">$out") or die "opening \"$out\": $!\n";
48 while(<STDIN>)
50 # Certain commands are discarded without further processing.
51 /^\@(?:
52 [a-z]+index # @*index: useful only in complete manual
53 |need # @need: useful only in printed manual
54 |(?:end\s+)?group # @group .. @end group: ditto
55 |page # @page: ditto
56 |node # @node: useful only in .info file
57 )\b/x and next;
59 chomp;
61 # Look for filename and title markers.
62 /^\@setfilename\s+([^.]+)/ and $fn = $1, next;
63 /^\@settitle\s+([^.]+)/ and $tl = $1, next;
65 # Look for blocks surrounded by @c man begin SECTION ... @c man end.
66 # This really oughta be @ifman ... @end ifman and the like, but such
67 # would require rev'ing all other Texinfo translators.
68 /^\@c man begin ([A-Z]+)/ and $sect = $1, $output = 1, next;
69 /^\@c man end/ and do {
70 $sects{$sect} = "" unless exists $sects{$sect};
71 $sects{$sect} .= postprocess($section);
72 $section = "";
73 $output = 0;
74 next;
76 next unless $output;
78 # Discard comments. (Can't do it above, because then we'd never see
79 # @c man lines.)
80 /^\@c\b/ and next;
82 # End-block handler goes up here because it needs to operate even
83 # if we are skipping.
84 /^\@end\s+([a-z]+)/ and do {
85 # Ignore @end foo, where foo is not an operation which may
86 # cause us to skip, if we are presently skipping.
87 my $ended = $1;
88 next if $skipping && $ended !~ /^(?:ifset|ifclear|ignore|menu)$/;
90 die "\@end $ended without \@$ended at line $.\n" unless defined $endw;
91 die "\@$endw ended by \@end $ended at line $.\n" unless $ended eq $endw;
93 $endw = pop @endwstack;
95 if ($ended =~ /^(?:ifset|ifclear|ignore|menu)$/) {
96 $skipping = pop @skstack;
97 next;
98 } elsif ($ended =~ /^(?:example|smallexample)$/) {
99 $shift = "";
100 $_ = ""; # need a paragraph break
101 } elsif ($ended =~ /^(?:itemize|enumerate|table)$/) {
102 $_ = "\n=back\n";
103 $ic = pop @icstack;
104 } else {
105 die "unknown command \@end $ended at line $.\n";
109 # We must handle commands which can cause skipping even while we
110 # are skipping, otherwise we will not process nested conditionals
111 # correctly.
112 /^\@ifset\s+([a-zA-Z0-9_-]+)/ and do {
113 push @endwstack, $endw;
114 push @skstack, $skipping;
115 $endw = "ifset";
116 $skipping = 1 unless exists $defs{$1};
117 next;
120 /^\@ifclear\s+([a-zA-Z0-9_-]+)/ and do {
121 push @endwstack, $endw;
122 push @skstack, $skipping;
123 $endw = "ifclear";
124 $skipping = 1 if exists $defs{$1};
125 next;
128 /^\@(ignore|menu)\b/ and do {
129 push @endwstack, $endw;
130 push @skstack, $skipping;
131 $endw = $1;
132 $skipping = 1;
133 next;
136 next if $skipping;
138 # Character entities. First the ones that can be replaced by raw text
139 # or discarded outright:
140 s/\@copyright\{\}/(c)/g;
141 s/\@dots\{\}/.../g;
142 s/\@enddots\{\}/..../g;
143 s/\@([.!? ])/$1/g;
144 s/\@[:-]//g;
145 s/\@bullet(?:\{\})?/*/g;
146 s/\@TeX\{\}/TeX/g;
147 s/\@pounds\{\}/\#/g;
148 s/\@minus(?:\{\})?/-/g;
150 # Now the ones that have to be replaced by special escapes
151 # (which will be turned back into text by unmunge())
152 s/&/&amp;/g;
153 s/\@\{/&lbrace;/g;
154 s/\@\}/&rbrace;/g;
155 s/\@\@/&at;/g;
156 # POD doesn't interpret E<> inside a verbatim block.
157 if ($shift eq "") {
158 s/</&lt;/g;
159 s/>/&gt;/g;
160 } else {
161 s/</&LT;/g;
162 s/>/&GT;/g;
165 # Single line command handlers.
166 /^\@set\s+([a-zA-Z0-9_-]+)\s*(.*)$/ and $defs{$1} = $2, next;
167 /^\@clear\s+([a-zA-Z0-9_-]+)/ and delete $defs{$1}, next;
169 /^\@section\s+(.+)$/ and $_ = "\n=head2 $1\n";
170 /^\@subsection\s+(.+)$/ and $_ = "\n=head3 $1\n";
172 # Block command handlers:
173 /^\@itemize\s+(\@[a-z]+|\*|-)/ and do {
174 push @endwstack, $endw;
175 push @icstack, $ic;
176 $ic = $1;
177 $_ = "\n=over 4\n";
178 $endw = "itemize";
181 /^\@enumerate(?:\s+([A-Z0-9]+))?/ and do {
182 push @endwstack, $endw;
183 push @icstack, $ic;
184 if (defined $1) {
185 $ic = $1 . ".";
186 } else {
187 $ic = "1.";
189 $_ = "\n=over 4\n";
190 $endw = "enumerate";
193 /^\@table\s+(\@[a-z]+)/ and do {
194 push @endwstack, $endw;
195 push @icstack, $ic;
196 $ic = $1;
197 $ic =~ s/\@(?:samp|strong|key|gcctabopt|env)/B/;
198 $ic =~ s/\@(?:code|kbd)/C/;
199 $ic =~ s/\@(?:dfn|var|emph|cite|i)/I/;
200 $ic =~ s/\@(?:file)/F/;
201 $_ = "\n=over 4\n";
202 $endw = "table";
205 /^\@((?:small)?example)/ and do {
206 push @endwstack, $endw;
207 $endw = $1;
208 $shift = "\t";
209 $_ = ""; # need a paragraph break
212 /^\@itemx?\s*(.+)?$/ and do {
213 if (defined $1) {
214 # Entity escapes prevent munging by the <> processing below.
215 $_ = "\n=item $ic\&LT;$1\&GT;\n";
216 } else {
217 $_ = "\n=item $ic\n";
218 $ic =~ y/A-Ya-y1-8/B-Zb-z2-9/;
222 $section .= $shift.$_."\n";
225 die "No filename or title\n" unless defined $fn && defined $tl;
227 $sects{NAME} = "$fn \- $tl\n";
228 $sects{FOOTNOTES} .= "=back\n" if exists $sects{FOOTNOTES};
230 for $sect (qw(NAME SYNOPSIS DESCRIPTION OPTIONS ENVIRONMENT FILES
231 BUGS NOTES FOOTNOTES SEEALSO AUTHOR COPYRIGHT)) {
232 if(exists $sects{$sect}) {
233 $head = $sect;
234 $head =~ s/SEEALSO/SEE ALSO/;
235 print "=head1 $head\n\n";
236 print scalar unmunge ($sects{$sect});
237 print "\n";
241 sub usage
243 die "usage: $0 [-D toggle...] [infile [outfile]]\n";
246 sub postprocess
248 local $_ = $_[0];
250 # @value{foo} is replaced by whatever 'foo' is defined as.
251 s/\@value\{([a-zA-Z0-9_-]+)\}/$defs{$1}/g;
253 # Formatting commands.
254 # Temporary escape for @r.
255 s/\@r\{([^\}]*)\}/R<$1>/g;
256 s/\@(?:dfn|var|emph|cite|i)\{([^\}]*)\}/I<$1>/g;
257 s/\@(?:code|kbd)\{([^\}]*)\}/C<$1>/g;
258 s/\@(?:gccoptlist|samp|strong|key|option|env|command|b)\{([^\}]*)\}/B<$1>/g;
259 s/\@sc\{([^\}]*)\}/\U$1/g;
260 s/\@file\{([^\}]*)\}/F<$1>/g;
261 s/\@w\{([^\}]*)\}/S<$1>/g;
262 s/\@(?:dmn|math)\{([^\}]*)\}/$1/g;
264 # Handle @r inside bold.
265 1 while s/B<((?:[^<>]*|I<[^<>*]*>)*)R<([^>]*)>/B<$1>${2}B</g;
267 # Cross references are thrown away, as are @noindent and @refill.
268 # (@noindent is impossible in .pod, and @refill is unnecessary.)
269 # @* is also impossible in .pod; we discard it and any newline that
270 # follows it. Similarly, our macro @gol must be discarded.
272 s/\(?\@xref\{(?:[^\}]*)\}(?:[^.<]|(?:<[^<>]*>))*\.\)?//g;
273 s/\s+\(\@pxref\{(?:[^\}]*)\}\)//g;
274 s/;\s+\@pxref\{(?:[^\}]*)\}//g;
275 s/\@noindent\s*//g;
276 s/\@refill//g;
277 s/\@gol//g;
278 s/\@\*\s*\n?//g;
280 # @uref can take one, two, or three arguments, with different
281 # semantics each time. @url and @email are just like @uref with
282 # one argument, for our purposes.
283 s/\@(?:uref|url|email)\{([^\},]*)\}/&lt;B<$1>&gt;/g;
284 s/\@uref\{([^\},]*),([^\},]*)\}/$2 (C<$1>)/g;
285 s/\@uref\{([^\},]*),([^\},]*),([^\},]*)\}/$3/g;
287 # Turn B<blah I<blah> blah> into B<blah> I<blah> B<blah> to
288 # match Texinfo semantics of @emph inside @samp.
289 s/&LT;/</g;
290 s/&GT;/>/g;
291 1 while (s/B<([^<>]*)I<([^>]+)>/B<$1>I<$2>B</g);
292 1 while (s/I<([^<>]*)B<([^>]+)>/I<$1>B<$2>I</g);
293 s/[BI]<>//g;
294 s/([BI])<(\s+)([^>]+)>/$2$1<$3>/g;
295 s/([BI])<([^>]+?)(\s+)>/$1<$2>$3/g;
297 # Extract footnotes. This has to be done after all other
298 # processing because otherwise the regexp will choke on formatting
299 # inside @footnote.
300 while (/\@footnote/g) {
301 s/\@footnote\{([^\}]+)\}/[$fnno]/;
302 add_footnote($1, $fnno);
303 $fnno++;
306 return $_;
309 sub unmunge
311 # Replace escaped symbols with their equivalents.
312 local $_ = $_[0];
314 s/&lt;/E<lt>/g;
315 s/&gt;/E<gt>/g;
316 s/&lbrace;/\{/g;
317 s/&rbrace;/\}/g;
318 s/&at;/\@/g;
319 s/&amp;/&/g;
320 return $_;
323 sub add_footnote
325 unless (exists $sects{FOOTNOTES}) {
326 $sects{FOOTNOTES} = "\n=over 4\n\n";
329 $sects{FOOTNOTES} .= "=item $fnno.\n\n"; $fnno++;
330 $sects{FOOTNOTES} .= $_[0];
331 $sects{FOOTNOTES} .= "\n\n";