This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
autodoc.pl: Add N flag
[perl5.git] / autodoc.pl
CommitLineData
94bdecf9 1#!/usr/bin/perl -w
6294c161
DM
2#
3# Unconditionally regenerate:
4#
5# pod/perlintern.pod
6# pod/perlapi.pod
7#
8# from information stored in
9#
10# embed.fnc
11# plus all the .c and .h files listed in MANIFEST
12#
13# Has an optional arg, which is the directory to chdir to before reading
14# MANIFEST and *.[ch].
15#
52a9d53b 16# This script is invoked as part of 'make all'
151c3fe5 17#
f554dfc5
MH
18# '=head1' are the only headings looked for. If the first non-blank line after
19# the heading begins with a word character, it is considered to be the first
20# line of documentation that applies to the heading itself. That is, it is
21# output immediately after the heading, before the first function, and not
22# indented. The next input line that is a pod directive terminates this
23# heading-level documentation.
94bdecf9 24
1fcde0e9
KW
25# The meanings of the flags fields in embed.fnc and the source code is
26# documented at the top of embed.fnc.
27
56a0c332 28use strict;
a64c954a 29
7882b24a
NC
30if (@ARGV) {
31 my $workdir = shift;
32 chdir $workdir
33 or die "Couldn't chdir to '$workdir': $!";
34}
3d7c117d
MB
35require './regen/regen_lib.pl';
36require './regen/embed_lib.pl';
7882b24a 37
94bdecf9 38#
346f75ff 39# See database of global and static function prototypes in embed.fnc
94bdecf9
JH
40# This is used to generate prototype headers under various configurations,
41# export symbols lists for different platforms, and macros to provide an
42# implicit interpreter context argument.
43#
44
6a235718 45my %docs;
5ce57792
NC
46my %funcflags;
47my %macro = (
48 ax => 1,
49 items => 1,
50 ix => 1,
51 svtype => 1,
52 );
53my %missing;
94bdecf9
JH
54
55my $curheader = "Unknown section";
56
57sub autodoc ($$) { # parse a file and extract documentation info
58 my($fh,$file) = @_;
151c3fe5 59 my($in, $doc, $line, $header_doc);
f554dfc5
MH
60
61 # Count lines easier
62 my $get_next_line = sub { $line++; return <$fh> };
63
94bdecf9 64FUNC:
f554dfc5 65 while (defined($in = $get_next_line->())) {
5ce57792
NC
66 if ($in =~ /^#\s*define\s+([A-Za-z_][A-Za-z_0-9]+)\(/ &&
67 ($file ne 'embed.h' || $file ne 'proto.h')) {
68 $macro{$1} = $file;
69 next FUNC;
70 }
94bdecf9
JH
71 if ($in=~ /^=head1 (.*)/) {
72 $curheader = $1;
151c3fe5 73
f554dfc5
MH
74 # If the next non-space line begins with a word char, then it is
75 # the start of heading-ldevel documentation.
76 if (defined($doc = $get_next_line->())) {
77 # Skip over empty lines
78 while ($doc =~ /^\s+$/) {
79 if (! defined($doc = $get_next_line->())) {
80 next FUNC;
81 }
82 }
83
151c3fe5
KW
84 if ($doc !~ /^\w/) {
85 $in = $doc;
86 redo FUNC;
87 }
88 $header_doc = $doc;
151c3fe5
KW
89
90 # Continue getting the heading-level documentation until read
91 # in any pod directive (or as a fail-safe, find a closing
92 # comment to this pod in a C language file
93HDR_DOC:
f554dfc5 94 while (defined($doc = $get_next_line->())) {
151c3fe5
KW
95 if ($doc =~ /^=\w/) {
96 $in = $doc;
97 redo FUNC;
98 }
151c3fe5
KW
99
100 if ($doc =~ m:^\s*\*/$:) {
101 warn "=cut missing? $file:$line:$doc";;
102 last HDR_DOC;
103 }
104 $header_doc .= $doc;
105 }
106 }
94bdecf9
JH
107 next FUNC;
108 }
78c9d763 109 if ($in =~ /^=for\s+apidoc\s+(.*?)\s*\n/) {
0a60f600
KW
110 my $proto_in_file = $1;
111 my $proto = $proto_in_file;
94bdecf9 112 $proto = "||$proto" unless $proto =~ /\|/;
73060c4e 113 my($flags, $ret, $name, @args) = split /\s*\|\s*/, $proto;
0a60f600
KW
114 warn ("'$name' not \\w+ in '$proto_in_file' in $file")
115 if $flags !~ /N/ && $name !~ / ^ [_[:alpha:]] \w* $ /x;
94bdecf9
JH
116 my $docs = "";
117DOC:
f554dfc5 118 while (defined($doc = $get_next_line->())) {
72d4186d
KW
119
120 # Other pod commands are considered part of the current
121 # function's docs, so can have lists, etc.
122 last DOC if $doc =~ /^=(cut|for\s+apidoc|head)/;
94bdecf9
JH
123 if ($doc =~ m:^\*/$:) {
124 warn "=cut missing? $file:$line:$doc";;
125 last DOC;
72d4186d 126 }
94bdecf9
JH
127 $docs .= $doc;
128 }
129 $docs = "\n$docs" if $docs and $docs !~ /^\n/;
5ce57792
NC
130
131 # Check the consistency of the flags
132 my ($embed_where, $inline_where);
133 my ($embed_may_change, $inline_may_change);
134
d40e888a
KW
135 my $embed_docref = delete $funcflags{$name};
136 if ($embed_docref and %$embed_docref) {
137 $embed_where = $embed_docref->{flags} =~ /A/ ? 'api' : 'guts';
e4aeb12f 138 $embed_may_change = $embed_docref->{flags} =~ /x/;
d40e888a 139 $flags .= 'D' if $embed_docref->{flags} =~ /D/;
fb7c7bc0 140 $flags .= 'O' if $embed_docref->{flags} =~ /O/;
54c193ae
KW
141 $flags .= 'p' if $embed_docref->{flags} =~ /p/;
142 $flags .= 'M' if $embed_docref->{flags} =~ /M/;
d7cc3209 143 $flags .= 'T' if $embed_docref->{flags} =~ /T/;
5ce57792
NC
144 } else {
145 $missing{$name} = $file;
94bdecf9 146 }
5ce57792
NC
147 if ($flags =~ /m/) {
148 $inline_where = $flags =~ /A/ ? 'api' : 'guts';
149 $inline_may_change = $flags =~ /x/;
150
151 if (defined $embed_where && $inline_where ne $embed_where) {
152 warn "Function '$name' inconsistency: embed.fnc says $embed_where, Pod says $inline_where";
153 }
154
155 if (defined $embed_may_change
156 && $inline_may_change ne $embed_may_change) {
157 my $message = "Function '$name' inconsistency: ";
158 if ($embed_may_change) {
159 $message .= "embed.fnc says 'may change', Pod does not";
160 } else {
161 $message .= "Pod says 'may change', embed.fnc does not";
162 }
163 warn $message;
164 }
165 } elsif (!defined $embed_where) {
166 warn "Unable to place $name!\n";
167 next;
168 } else {
169 $inline_where = $embed_where;
170 $flags .= 'x' if $embed_may_change;
d40e888a
KW
171 @args = @{$embed_docref->{args}};
172 $ret = $embed_docref->{retval};
94bdecf9 173 }
5ce57792 174
7a6610ca
DM
175 if (exists $docs{$inline_where}{$curheader}{$name}) {
176 warn "$0: duplicate API entry for '$name' in $inline_where/$curheader\n";
177 next;
178 }
5ce57792
NC
179 $docs{$inline_where}{$curheader}{$name}
180 = [$flags, $docs, $ret, $file, @args];
181
151c3fe5
KW
182 # Create a special entry with an empty-string name for the
183 # heading-level documentation.
184 if (defined $header_doc) {
185 $docs{$inline_where}{$curheader}{""} = $header_doc;
186 undef $header_doc;
187 }
188
94bdecf9 189 if (defined $doc) {
e509e693 190 if ($doc =~ /^=(?:for|head)/) {
94bdecf9
JH
191 $in = $doc;
192 redo FUNC;
193 }
194 } else {
195 warn "$file:$line:$in";
196 }
197 }
198 }
199}
200
201sub docout ($$$) { # output the docs for one function
202 my($fh, $name, $docref) = @_;
203 my($flags, $docs, $ret, $file, @args) = @$docref;
d8c40edc 204 $name =~ s/\s*$//;
94bdecf9 205
d4e99c76
KW
206 if ($flags =~ /D/) {
207 $docs = "\n\nDEPRECATED! It is planned to remove this function from a
208future release of Perl. Do not use it for new code; remove it from
209existing code.\n\n$docs";
210 }
211 else {
58a428bb
KW
212 $docs = "\n\nNOTE: this function is experimental and may change or be
213removed without notice.\n\n$docs" if $flags =~ /x/;
d4e99c76 214 }
54c193ae
KW
215
216 # Is Perl_, but no #define foo # Perl_foo
217 my $p = $flags =~ /p/ && $flags =~ /o/ && $flags !~ /M/;
218
94bdecf9 219 $docs .= "NOTE: the perl_ form of this function is deprecated.\n\n"
fb7c7bc0 220 if $flags =~ /O/;
54c193ae
KW
221 if ($p) {
222 $docs .= "NOTE: this function must be explicitly called as Perl_$name";
d7cc3209 223 $docs .= " with an aTHX_ parameter" if $flags !~ /T/;
54c193ae
KW
224 $docs .= ".\n\n"
225 }
94bdecf9 226
d8c40edc 227 print $fh "=item $name\nX<$name>\n$docs";
94bdecf9
JH
228
229 if ($flags =~ /U/) { # no usage
8b5ff177 230 warn("U and s flags are incompatible") if $flags =~ /s/;
94bdecf9 231 # nothing
05ca4832 232 } else {
8b5ff177 233 if ($flags =~ /n/) { # no args
1fcde0e9
KW
234 warn("n flag without m") unless $flags =~ /m/;
235 warn("n flag but apparently has args") if @args;
1ded1f42
KW
236 print $fh "\t$ret\t$name";
237 } else { # full usage
1ded1f42
KW
238 my $n = "Perl_"x$p . $name;
239 my $large_ret = length $ret > 7;
240 my $indent_size = 7+8 # nroff: 7 under =head + 8 under =item
241 +8+($large_ret ? 1 + length $ret : 8)
242 +length($n) + 1;
243 my $indent;
244 print $fh "\t$ret" . ($large_ret ? ' ' : "\t") . "$n(";
245 my $long_args;
246 for (@args) {
247 if ($indent_size + 2 + length > 79) {
248 $long_args=1;
249 $indent_size -= length($n) - 3;
250 last;
251 }
252 }
253 my $args = '';
d7cc3209 254 if ($p && $flags !~ /T/) {
1ded1f42
KW
255 $args = @args ? "pTHX_ " : "pTHX";
256 if ($long_args) { print $fh $args; $args = '' }
257 }
258 $long_args and print $fh "\n";
259 my $first = !$long_args;
260 while () {
261 if (!@args or
262 length $args
263 && $indent_size + 3 + length($args[0]) + length $args > 79
264 ) {
265 print $fh
266 $first ? '' : (
267 $indent //=
268 "\t".($large_ret ? " " x (1+length $ret) : "\t")
269 ." "x($long_args ? 4 : 1 + length $n)
270 ),
271 $args, (","x($args ne 'pTHX_ ') . "\n")x!!@args;
272 $args = $first = '';
273 }
274 @args or last;
275 $args .= ", "x!!(length $args && $args ne 'pTHX_ ')
276 . shift @args;
277 }
278 if ($long_args) { print $fh "\n", substr $indent, 0, -4 }
279 print $fh ")";
280 }
8b5ff177 281 print $fh ";" if $flags =~ /s/; # semicolon "dTHR;"
1ded1f42 282 print $fh "\n\n";
94bdecf9
JH
283 }
284 print $fh "=for hackers\nFound in file $file\n\n";
285}
286
f83c6033
KW
287sub sort_helper {
288 # Do a case-insensitive dictionary sort, with only alphabetics
289 # significant, falling back to using everything for determinancy
1354d57e 290 return (uc($a =~ s/[[:^alpha:]]//r) cmp uc($b =~ s/[[:^alpha:]]//r))
f83c6033
KW
291 || uc($a) cmp uc($b)
292 || $a cmp $b;
293}
294
7b73ff98 295sub output {
5a0155e6 296 my ($podname, $header, $dochash, $missing, $footer) = @_;
6a4c4cd4
DM
297 #
298 # strip leading '|' from each line which had been used to hide
299 # pod from pod checkers.
300 s/^\|//gm for $header, $footer;
301
7882b24a
NC
302 my $fh = open_new("pod/$podname.pod", undef,
303 {by => "$0 extracting documentation",
f1f44974 304 from => 'the C source files'}, 1);
e0492643 305
7882b24a 306 print $fh $header;
e0492643 307
7b73ff98 308 my $key;
f83c6033 309 for $key (sort sort_helper keys %$dochash) {
7b73ff98 310 my $section = $dochash->{$key};
151c3fe5
KW
311 print $fh "\n=head1 $key\n\n";
312
313 # Output any heading-level documentation and delete so won't get in
314 # the way later
315 if (exists $section->{""}) {
316 print $fh $section->{""} . "\n";
317 delete $section->{""};
318 }
319 print $fh "=over 8\n\n";
320
f83c6033 321 for my $key (sort sort_helper keys %$section) {
7b73ff98
NC
322 docout($fh, $key, $section->{$key});
323 }
324 print $fh "\n=back\n";
325 }
326
5a0155e6 327 if (@$missing) {
a23e6e20 328 print $fh "\n=head1 Undocumented functions\n\n";
2616800a 329 print $fh $podname eq 'perlapi' ? <<'_EOB_' : <<'_EOB_';
474d0ac8 330The following functions have been flagged as part of the public API,
72d33970 331but are currently undocumented. Use them at your own risk, as the
ba4591a5
KW
332interfaces are subject to change. Functions that are not listed in this
333document are not intended for public use, and should NOT be used under any
334circumstances.
335
5a4fed09
KW
336If you feel you need to use one of these functions, first send email to
337L<perl5-porters@perl.org|mailto:perl5-porters@perl.org>. It may be
338that there is a good reason for the function not being documented, and it
339should be removed from this list; or it may just be that no one has gotten
340around to documenting it. In the latter case, you will be asked to submit a
341patch to document the function. Once your patch is accepted, it will indicate
342that the interface is stable (unless it is explicitly marked otherwise) and
343usable by you.
cf5f2f8f 344_EOB_
2616800a
FC
345The following functions are currently undocumented. If you use one of
346them, you may wish to consider creating and submitting documentation for
347it.
2616800a 348_EOB_
6a4c4cd4
DM
349 print $fh "\n=over\n\n";
350
cf5f2f8f
KW
351 for my $missing (sort @$missing) {
352 print $fh "=item $missing\nX<$missing>\n\n";
5a0155e6 353 }
cf5f2f8f
KW
354 print $fh "=back\n\n";
355}
7882b24a 356 print $fh $footer, "=cut\n";
5a0155e6 357
7882b24a 358 read_only_bottom_close_and_rename($fh);
cd093254
MM
359}
360
e8e591c9
NC
361foreach (@{(setup_embed())[0]}) {
362 next if @$_ < 2;
363 my ($flags, $retval, $func, @args) = @$_;
364 s/\b(?:NN|NULLOK)\b\s+//g for @args;
bc350081 365
5ce57792
NC
366 $funcflags{$func} = {
367 flags => $flags,
368 retval => $retval,
369 args => \@args,
370 };
371}
372
5ce57792
NC
373# glob() picks up docs from extra .c or .h files that may be in unclean
374# development trees.
741c0772
NC
375open my $fh, '<', 'MANIFEST'
376 or die "Can't open MANIFEST: $!";
377while (my $line = <$fh>) {
378 next unless my ($file) = $line =~ /^(\S+\.[ch])\t/;
5ce57792 379
1ae6ead9 380 open F, '<', $file or die "Cannot open $file for docs: $!\n";
5ce57792
NC
381 $curheader = "Functions in file $file\n";
382 autodoc(\*F,$file);
383 close F or die "Error closing $file: $!\n";
384}
741c0772 385close $fh or die "Error whilst reading MANIFEST: $!";
5ce57792
NC
386
387for (sort keys %funcflags) {
388 next unless $funcflags{$_}{flags} =~ /d/;
389 warn "no docs for $_\n"
bc350081 390}
94bdecf9 391
5ce57792
NC
392foreach (sort keys %missing) {
393 next if $macro{$_};
394 # Heuristics for known not-a-function macros:
395 next if /^[A-Z]/;
396 next if /^dj?[A-Z]/;
397
398 warn "Function '$_', documented in $missing{$_}, not listed in embed.fnc";
94bdecf9
JH
399}
400
5ce57792
NC
401# walk table providing an array of components in each line to
402# subroutine, printing the result
403
8c869419
KW
404# List of funcs in the public API that aren't also marked as experimental nor
405# deprecated.
e4aeb12f 406my @missing_api = grep $funcflags{$_}{flags} =~ /A/ && $funcflags{$_}{flags} !~ /[xD]/ && !$docs{api}{$_}, keys %funcflags;
5a0155e6 407output('perlapi', <<'_EOB_', $docs{api}, \@missing_api, <<'_EOE_');
6a4c4cd4
DM
408|=encoding UTF-8
409|
410|=head1 NAME
411|
412|perlapi - autogenerated documentation for the perl public API
413|
414|=head1 DESCRIPTION
415|X<Perl API> X<API> X<api>
416|
417|This file contains the documentation of the perl public API generated by
418|F<embed.pl>, specifically a listing of functions, macros, flags, and variables
419|that may be used by extension writers. L<At the end|/Undocumented functions>
420|is a list of functions which have yet to be documented. The interfaces of
421|those are subject to change without notice. Anything not listed here is
422|not part of the public API, and should not be used by extension writers at
423|all. For these reasons, blindly using functions listed in proto.h is to be
424|avoided when writing extensions.
425|
426|In Perl, unlike C, a string of characters may generally contain embedded
427|C<NUL> characters. Sometimes in the documentation a Perl string is referred
428|to as a "buffer" to distinguish it from a C string, but sometimes they are
429|both just referred to as strings.
430|
431|Note that all Perl API global variables must be referenced with the C<PL_>
432|prefix. Again, those not listed here are not to be used by extension writers,
433|and can be changed or removed without notice; same with macros.
434|Some macros are provided for compatibility with the older,
435|unadorned names, but this support may be disabled in a future release.
436|
437|Perl was originally written to handle US-ASCII only (that is characters
438|whose ordinal numbers are in the range 0 - 127).
439|And documentation and comments may still use the term ASCII, when
440|sometimes in fact the entire range from 0 - 255 is meant.
441|
442|The non-ASCII characters below 256 can have various meanings, depending on
443|various things. (See, most notably, L<perllocale>.) But usually the whole
444|range can be referred to as ISO-8859-1. Often, the term "Latin-1" (or
445|"Latin1") is used as an equivalent for ISO-8859-1. But some people treat
446|"Latin1" as referring just to the characters in the range 128 through 255, or
447|somethimes from 160 through 255.
448|This documentation uses "Latin1" and "Latin-1" to refer to all 256 characters.
449|
450|Note that Perl can be compiled and run under either ASCII or EBCDIC (See
451|L<perlebcdic>). Most of the documentation (and even comments in the code)
452|ignore the EBCDIC possibility.
453|For almost all purposes the differences are transparent.
454|As an example, under EBCDIC,
455|instead of UTF-8, UTF-EBCDIC is used to encode Unicode strings, and so
456|whenever this documentation refers to C<utf8>
457|(and variants of that name, including in function names),
458|it also (essentially transparently) means C<UTF-EBCDIC>.
459|But the ordinals of characters differ between ASCII, EBCDIC, and
460|the UTF- encodings, and a string encoded in UTF-EBCDIC may occupy a different
461|number of bytes than in UTF-8.
462|
463|The listing below is alphabetical, case insensitive.
464|
94bdecf9 465_EOB_
6a4c4cd4
DM
466|
467|=head1 AUTHORS
468|
469|Until May 1997, this document was maintained by Jeff Okamoto
470|<okamoto@corp.hp.com>. It is now maintained as part of Perl itself.
471|
472|With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
473|Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
474|Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
475|Stephen McCamant, and Gurusamy Sarathy.
476|
477|API Listing originally by Dean Roehrich <roehrich@cray.com>.
478|
479|Updated to be autogenerated from comments in the source by Benjamin Stuhl.
480|
481|=head1 SEE ALSO
482|
483|L<perlguts>, L<perlxs>, L<perlxstut>, L<perlintern>
484|
94bdecf9
JH
485_EOE_
486
79fc8511
FC
487# List of non-static internal functions
488my @missing_guts =
9f589e47 489 grep $funcflags{$_}{flags} !~ /[AS]/ && !$docs{guts}{$_}, keys %funcflags;
5a0155e6
TC
490
491output('perlintern', <<'END', $docs{guts}, \@missing_guts, <<'END');
6a4c4cd4
DM
492|=head1 NAME
493|
494|perlintern - autogenerated documentation of purely B<internal>
495| Perl functions
496|
497|=head1 DESCRIPTION
498|X<internal Perl functions> X<interpreter functions>
499|
500|This file is the autogenerated documentation of functions in the
501|Perl interpreter that are documented using Perl's internal documentation
502|format but are not marked as part of the Perl API. In other words,
503|B<they are not for use in extensions>!
504|
94bdecf9 505END
6a4c4cd4
DM
506|
507|=head1 AUTHORS
508|
509|The autodocumentation system was originally added to the Perl core by
510|Benjamin Stuhl. Documentation is by whoever was kind enough to
511|document their functions.
512|
513|=head1 SEE ALSO
514|
515|L<perlguts>, L<perlapi>
516|
94bdecf9 517END