This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Note that mro[gs]et_private_data are documented
[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
7b1f0a98
KW
38my @specialized_docs = sort qw( perlguts
39 perlxs
40 perlxstut
b87d9527 41 perlclib
a1a5a9c8 42 warnings
678f21a2 43 perlapio
31b83e34 44 perlcall
0f292d69 45 perlfilter
6a6f8717 46 perlmroapi
7b1f0a98
KW
47 );
48my $other_places_api = join " ", map { "L<$_>" } sort @specialized_docs, 'perlintern';
49my $other_places_intern = join " ", map { "L<$_>" } sort @specialized_docs, 'perlapi';
50
b87d9527
KW
51@specialized_docs = map { "L<$_>" } sort @specialized_docs;
52$specialized_docs[-1] =~ s/^/and /;
53my $specialized_docs = join ", ", @specialized_docs;
54
94bdecf9 55#
346f75ff 56# See database of global and static function prototypes in embed.fnc
94bdecf9
JH
57# This is used to generate prototype headers under various configurations,
58# export symbols lists for different platforms, and macros to provide an
59# implicit interpreter context argument.
60#
61
6a235718 62my %docs;
5ce57792 63my %funcflags;
5ce57792 64my %missing;
94bdecf9
JH
65
66my $curheader = "Unknown section";
67
68sub autodoc ($$) { # parse a file and extract documentation info
69 my($fh,$file) = @_;
151c3fe5 70 my($in, $doc, $line, $header_doc);
f554dfc5
MH
71
72 # Count lines easier
73 my $get_next_line = sub { $line++; return <$fh> };
74
94bdecf9 75FUNC:
f554dfc5 76 while (defined($in = $get_next_line->())) {
94bdecf9
JH
77 if ($in=~ /^=head1 (.*)/) {
78 $curheader = $1;
151c3fe5 79
f554dfc5 80 # If the next non-space line begins with a word char, then it is
1c82f4a4 81 # the start of heading-level documentation.
20046047 82 if (defined($doc = $get_next_line->())) {
f554dfc5
MH
83 # Skip over empty lines
84 while ($doc =~ /^\s+$/) {
85 if (! defined($doc = $get_next_line->())) {
86 next FUNC;
87 }
88 }
89
151c3fe5
KW
90 if ($doc !~ /^\w/) {
91 $in = $doc;
92 redo FUNC;
93 }
94 $header_doc = $doc;
151c3fe5
KW
95
96 # Continue getting the heading-level documentation until read
97 # in any pod directive (or as a fail-safe, find a closing
98 # comment to this pod in a C language file
99HDR_DOC:
f554dfc5 100 while (defined($doc = $get_next_line->())) {
151c3fe5
KW
101 if ($doc =~ /^=\w/) {
102 $in = $doc;
103 redo FUNC;
104 }
151c3fe5
KW
105
106 if ($doc =~ m:^\s*\*/$:) {
107 warn "=cut missing? $file:$line:$doc";;
108 last HDR_DOC;
109 }
110 $header_doc .= $doc;
111 }
112 }
94bdecf9
JH
113 next FUNC;
114 }
20046047
KE
115 if ($in =~ /^=for\s+apidoc\s+(.*?)\s*\n/) {
116 my $proto_in_file = $1;
117 my $proto = $proto_in_file;
118 $proto = "||$proto" unless $proto =~ /\|/;
119 my($flags, $ret, $name, @args) = split /\s*\|\s*/, $proto;
256dda50
TC
120 $name or die <<EOS;
121Bad apidoc at $file line $.:
122 $in
123Expected:
124 =for apidoc flags|returntype|name|arg|arg|...
125 =for apidoc flags|returntype|name
126 =for apidoc name
127EOS
6523e108
KW
128 next FUNC if $flags =~ /h/;
129
0a60f600
KW
130 warn ("'$name' not \\w+ in '$proto_in_file' in $file")
131 if $flags !~ /N/ && $name !~ / ^ [_[:alpha:]] \w* $ /x;
20046047 132 my $docs = "";
94bdecf9 133DOC:
20046047 134 while (defined($doc = $get_next_line->())) {
72d4186d
KW
135
136 # Other pod commands are considered part of the current
137 # function's docs, so can have lists, etc.
138 last DOC if $doc =~ /^=(cut|for\s+apidoc|head)/;
20046047
KE
139 if ($doc =~ m:^\*/$:) {
140 warn "=cut missing? $file:$line:$doc";;
141 last DOC;
72d4186d 142 }
20046047
KE
143 $docs .= $doc;
144 }
145 $docs = "\n$docs" if $docs and $docs !~ /^\n/;
5ce57792 146
20046047 147 # If the entry is also in embed.fnc, it should be defined
8902d554 148 # completely there, but not here
20046047
KE
149 my $embed_docref = delete $funcflags{$name};
150 if ($embed_docref and %$embed_docref) {
8902d554
KW
151 warn "embed.fnc entry overrides redundant information in"
152 . " '$proto_in_file' in $file" if $flags || $ret || @args;
153 $flags = $embed_docref->{'flags'};
5514c4f1
KW
154 warn "embed.fnc entry '$name' missing 'd' flag"
155 unless $flags =~ /d/;
6523e108 156 next FUNC if $flags =~ /h/;
8902d554 157 $ret = $embed_docref->{'retval'};
20046047 158 @args = @{$embed_docref->{args}};
5514c4f1
KW
159 } elsif ($flags !~ /m/) { # Not in embed.fnc, is missing if not a
160 # macro
20046047
KE
161 $missing{$name} = $file;
162 }
5ce57792 163
8902d554 164 my $inline_where = $flags =~ /A/ ? 'api' : 'guts';
5ce57792 165
20046047 166 if (exists $docs{$inline_where}{$curheader}{$name}) {
7a6610ca
DM
167 warn "$0: duplicate API entry for '$name' in $inline_where/$curheader\n";
168 next;
169 }
20046047
KE
170 $docs{$inline_where}{$curheader}{$name}
171 = [$flags, $docs, $ret, $file, @args];
5ce57792 172
151c3fe5
KW
173 # Create a special entry with an empty-string name for the
174 # heading-level documentation.
20046047 175 if (defined $header_doc) {
151c3fe5
KW
176 $docs{$inline_where}{$curheader}{""} = $header_doc;
177 undef $header_doc;
178 }
179
20046047
KE
180 if (defined $doc) {
181 if ($doc =~ /^=(?:for|head)/) {
182 $in = $doc;
183 redo FUNC;
184 }
185 } else {
186 warn "$file:$line:$in";
187 }
188 }
94bdecf9
JH
189 }
190}
191
192sub docout ($$$) { # output the docs for one function
193 my($fh, $name, $docref) = @_;
194 my($flags, $docs, $ret, $file, @args) = @$docref;
d8c40edc 195 $name =~ s/\s*$//;
94bdecf9 196
d4e99c76
KW
197 if ($flags =~ /D/) {
198 $docs = "\n\nDEPRECATED! It is planned to remove this function from a
199future release of Perl. Do not use it for new code; remove it from
200existing code.\n\n$docs";
201 }
202 else {
58a428bb
KW
203 $docs = "\n\nNOTE: this function is experimental and may change or be
204removed without notice.\n\n$docs" if $flags =~ /x/;
d4e99c76 205 }
54c193ae
KW
206
207 # Is Perl_, but no #define foo # Perl_foo
208 my $p = $flags =~ /p/ && $flags =~ /o/ && $flags !~ /M/;
209
94bdecf9 210 $docs .= "NOTE: the perl_ form of this function is deprecated.\n\n"
20046047 211 if $flags =~ /O/;
54c193ae
KW
212 if ($p) {
213 $docs .= "NOTE: this function must be explicitly called as Perl_$name";
d7cc3209 214 $docs .= " with an aTHX_ parameter" if $flags !~ /T/;
54c193ae
KW
215 $docs .= ".\n\n"
216 }
94bdecf9 217
d8c40edc 218 print $fh "=item $name\nX<$name>\n$docs";
94bdecf9
JH
219
220 if ($flags =~ /U/) { # no usage
8b5ff177 221 warn("U and s flags are incompatible") if $flags =~ /s/;
20046047 222 # nothing
05ca4832 223 } else {
8b5ff177 224 if ($flags =~ /n/) { # no args
1fcde0e9
KW
225 warn("n flag without m") unless $flags =~ /m/;
226 warn("n flag but apparently has args") if @args;
1ded1f42
KW
227 print $fh "\t$ret\t$name";
228 } else { # full usage
1ded1f42
KW
229 my $n = "Perl_"x$p . $name;
230 my $large_ret = length $ret > 7;
231 my $indent_size = 7+8 # nroff: 7 under =head + 8 under =item
232 +8+($large_ret ? 1 + length $ret : 8)
233 +length($n) + 1;
234 my $indent;
235 print $fh "\t$ret" . ($large_ret ? ' ' : "\t") . "$n(";
236 my $long_args;
237 for (@args) {
238 if ($indent_size + 2 + length > 79) {
239 $long_args=1;
240 $indent_size -= length($n) - 3;
241 last;
242 }
243 }
244 my $args = '';
d7cc3209 245 if ($p && $flags !~ /T/) {
1ded1f42
KW
246 $args = @args ? "pTHX_ " : "pTHX";
247 if ($long_args) { print $fh $args; $args = '' }
248 }
249 $long_args and print $fh "\n";
250 my $first = !$long_args;
251 while () {
252 if (!@args or
253 length $args
254 && $indent_size + 3 + length($args[0]) + length $args > 79
255 ) {
256 print $fh
257 $first ? '' : (
258 $indent //=
259 "\t".($large_ret ? " " x (1+length $ret) : "\t")
260 ." "x($long_args ? 4 : 1 + length $n)
261 ),
262 $args, (","x($args ne 'pTHX_ ') . "\n")x!!@args;
263 $args = $first = '';
264 }
265 @args or last;
266 $args .= ", "x!!(length $args && $args ne 'pTHX_ ')
267 . shift @args;
268 }
269 if ($long_args) { print $fh "\n", substr $indent, 0, -4 }
270 print $fh ")";
271 }
8b5ff177 272 print $fh ";" if $flags =~ /s/; # semicolon "dTHR;"
1ded1f42 273 print $fh "\n\n";
94bdecf9
JH
274 }
275 print $fh "=for hackers\nFound in file $file\n\n";
276}
277
f83c6033
KW
278sub sort_helper {
279 # Do a case-insensitive dictionary sort, with only alphabetics
280 # significant, falling back to using everything for determinancy
1354d57e 281 return (uc($a =~ s/[[:^alpha:]]//r) cmp uc($b =~ s/[[:^alpha:]]//r))
f83c6033
KW
282 || uc($a) cmp uc($b)
283 || $a cmp $b;
284}
285
7b73ff98 286sub output {
5a0155e6 287 my ($podname, $header, $dochash, $missing, $footer) = @_;
6a4c4cd4
DM
288 #
289 # strip leading '|' from each line which had been used to hide
290 # pod from pod checkers.
291 s/^\|//gm for $header, $footer;
292
7882b24a 293 my $fh = open_new("pod/$podname.pod", undef,
20046047 294 {by => "$0 extracting documentation",
f1f44974 295 from => 'the C source files'}, 1);
e0492643 296
7882b24a 297 print $fh $header;
e0492643 298
7b73ff98 299 my $key;
f83c6033 300 for $key (sort sort_helper keys %$dochash) {
20046047
KE
301 my $section = $dochash->{$key};
302 print $fh "\n=head1 $key\n\n";
151c3fe5
KW
303
304 # Output any heading-level documentation and delete so won't get in
305 # the way later
306 if (exists $section->{""}) {
307 print $fh $section->{""} . "\n";
308 delete $section->{""};
309 }
20046047 310 print $fh "=over 8\n\n";
151c3fe5 311
20046047
KE
312 for my $key (sort sort_helper keys %$section) {
313 docout($fh, $key, $section->{$key});
314 }
315 print $fh "\n=back\n";
7b73ff98
NC
316 }
317
5a0155e6 318 if (@$missing) {
a23e6e20 319 print $fh "\n=head1 Undocumented functions\n\n";
2616800a 320 print $fh $podname eq 'perlapi' ? <<'_EOB_' : <<'_EOB_';
474d0ac8 321The following functions have been flagged as part of the public API,
72d33970 322but are currently undocumented. Use them at your own risk, as the
ba4591a5
KW
323interfaces are subject to change. Functions that are not listed in this
324document are not intended for public use, and should NOT be used under any
325circumstances.
326
5a4fed09
KW
327If you feel you need to use one of these functions, first send email to
328L<perl5-porters@perl.org|mailto:perl5-porters@perl.org>. It may be
329that there is a good reason for the function not being documented, and it
330should be removed from this list; or it may just be that no one has gotten
331around to documenting it. In the latter case, you will be asked to submit a
332patch to document the function. Once your patch is accepted, it will indicate
333that the interface is stable (unless it is explicitly marked otherwise) and
334usable by you.
cf5f2f8f 335_EOB_
2616800a
FC
336The following functions are currently undocumented. If you use one of
337them, you may wish to consider creating and submitting documentation for
338it.
2616800a 339_EOB_
6a4c4cd4
DM
340 print $fh "\n=over\n\n";
341
cf5f2f8f
KW
342 for my $missing (sort @$missing) {
343 print $fh "=item $missing\nX<$missing>\n\n";
5a0155e6 344 }
cf5f2f8f
KW
345 print $fh "=back\n\n";
346}
7882b24a 347 print $fh $footer, "=cut\n";
5a0155e6 348
7882b24a 349 read_only_bottom_close_and_rename($fh);
cd093254
MM
350}
351
e8e591c9
NC
352foreach (@{(setup_embed())[0]}) {
353 next if @$_ < 2;
354 my ($flags, $retval, $func, @args) = @$_;
355 s/\b(?:NN|NULLOK)\b\s+//g for @args;
bc350081 356
5ce57792 357 $funcflags{$func} = {
20046047
KE
358 flags => $flags,
359 retval => $retval,
360 args => \@args,
361 };
5ce57792
NC
362}
363
5ce57792
NC
364# glob() picks up docs from extra .c or .h files that may be in unclean
365# development trees.
741c0772
NC
366open my $fh, '<', 'MANIFEST'
367 or die "Can't open MANIFEST: $!";
368while (my $line = <$fh>) {
b87d9527 369 next unless my ($file) = $line =~ /^(\S+\.(?:[ch]|pod))\t/;
5ce57792 370
1ae6ead9 371 open F, '<', $file or die "Cannot open $file for docs: $!\n";
5ce57792
NC
372 $curheader = "Functions in file $file\n";
373 autodoc(\*F,$file);
374 close F or die "Error closing $file: $!\n";
375}
741c0772 376close $fh or die "Error whilst reading MANIFEST: $!";
5ce57792
NC
377
378for (sort keys %funcflags) {
379 next unless $funcflags{$_}{flags} =~ /d/;
6523e108 380 next if $funcflags{$_}{flags} =~ /h/;
5ce57792 381 warn "no docs for $_\n"
bc350081 382}
94bdecf9 383
5ce57792 384foreach (sort keys %missing) {
5ce57792 385 warn "Function '$_', documented in $missing{$_}, not listed in embed.fnc";
94bdecf9
JH
386}
387
5ce57792
NC
388# walk table providing an array of components in each line to
389# subroutine, printing the result
390
8c869419
KW
391# List of funcs in the public API that aren't also marked as experimental nor
392# deprecated.
b87d9527
KW
393my @missing_api = grep $funcflags{$_}{flags} =~ /A/
394 && $funcflags{$_}{flags} !~ /[xD]/
395 && !$docs{api}{$_}, keys %funcflags;
396output('perlapi', <<"_EOB_", $docs{api}, \@missing_api, <<"_EOE_");
6a4c4cd4
DM
397|=encoding UTF-8
398|
399|=head1 NAME
400|
401|perlapi - autogenerated documentation for the perl public API
402|
403|=head1 DESCRIPTION
404|X<Perl API> X<API> X<api>
405|
b87d9527
KW
406|This file contains most of the documentation of the perl public API, as
407|generated by F<embed.pl>. Specifically, it is a listing of functions,
408|macros, flags, and variables that may be used by extension writers. Some
409|specialized items are instead documented in $specialized_docs.
410|
411|L<At the end|/Undocumented functions> is a list of functions which have yet
412|to be documented. Patches welcome! The interfaces of these are subject to
413|change without notice.
414|
415|Anything not listed here is not part of the public API, and should not be
416|used by extension writers at all. For these reasons, blindly using functions
417|listed in proto.h is to be avoided when writing extensions.
6a4c4cd4
DM
418|
419|In Perl, unlike C, a string of characters may generally contain embedded
420|C<NUL> characters. Sometimes in the documentation a Perl string is referred
421|to as a "buffer" to distinguish it from a C string, but sometimes they are
422|both just referred to as strings.
423|
424|Note that all Perl API global variables must be referenced with the C<PL_>
425|prefix. Again, those not listed here are not to be used by extension writers,
426|and can be changed or removed without notice; same with macros.
427|Some macros are provided for compatibility with the older,
428|unadorned names, but this support may be disabled in a future release.
429|
430|Perl was originally written to handle US-ASCII only (that is characters
431|whose ordinal numbers are in the range 0 - 127).
432|And documentation and comments may still use the term ASCII, when
433|sometimes in fact the entire range from 0 - 255 is meant.
434|
435|The non-ASCII characters below 256 can have various meanings, depending on
436|various things. (See, most notably, L<perllocale>.) But usually the whole
437|range can be referred to as ISO-8859-1. Often, the term "Latin-1" (or
438|"Latin1") is used as an equivalent for ISO-8859-1. But some people treat
439|"Latin1" as referring just to the characters in the range 128 through 255, or
440|somethimes from 160 through 255.
441|This documentation uses "Latin1" and "Latin-1" to refer to all 256 characters.
442|
443|Note that Perl can be compiled and run under either ASCII or EBCDIC (See
444|L<perlebcdic>). Most of the documentation (and even comments in the code)
445|ignore the EBCDIC possibility.
446|For almost all purposes the differences are transparent.
447|As an example, under EBCDIC,
448|instead of UTF-8, UTF-EBCDIC is used to encode Unicode strings, and so
449|whenever this documentation refers to C<utf8>
450|(and variants of that name, including in function names),
451|it also (essentially transparently) means C<UTF-EBCDIC>.
452|But the ordinals of characters differ between ASCII, EBCDIC, and
453|the UTF- encodings, and a string encoded in UTF-EBCDIC may occupy a different
454|number of bytes than in UTF-8.
455|
456|The listing below is alphabetical, case insensitive.
457|
94bdecf9 458_EOB_
6a4c4cd4
DM
459|
460|=head1 AUTHORS
461|
462|Until May 1997, this document was maintained by Jeff Okamoto
7b1f0a98 463|<okamoto\@corp.hp.com>. It is now maintained as part of Perl itself.
6a4c4cd4
DM
464|
465|With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
466|Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
467|Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
468|Stephen McCamant, and Gurusamy Sarathy.
469|
7b1f0a98 470|API Listing originally by Dean Roehrich <roehrich\@cray.com>.
6a4c4cd4
DM
471|
472|Updated to be autogenerated from comments in the source by Benjamin Stuhl.
473|
474|=head1 SEE ALSO
475|
7b1f0a98 476$other_places_api
94bdecf9
JH
477_EOE_
478
79fc8511
FC
479# List of non-static internal functions
480my @missing_guts =
9f589e47 481 grep $funcflags{$_}{flags} !~ /[AS]/ && !$docs{guts}{$_}, keys %funcflags;
5a0155e6 482
7b1f0a98 483output('perlintern', <<'_EOB_', $docs{guts}, \@missing_guts, <<"_EOE_");
6a4c4cd4
DM
484|=head1 NAME
485|
486|perlintern - autogenerated documentation of purely B<internal>
20046047 487|Perl functions
6a4c4cd4
DM
488|
489|=head1 DESCRIPTION
490|X<internal Perl functions> X<interpreter functions>
491|
492|This file is the autogenerated documentation of functions in the
493|Perl interpreter that are documented using Perl's internal documentation
494|format but are not marked as part of the Perl API. In other words,
495|B<they are not for use in extensions>!
496|
7b1f0a98 497_EOB_
6a4c4cd4
DM
498|
499|=head1 AUTHORS
500|
501|The autodocumentation system was originally added to the Perl core by
502|Benjamin Stuhl. Documentation is by whoever was kind enough to
503|document their functions.
504|
505|=head1 SEE ALSO
506|
7b1f0a98
KW
507$other_places_intern
508_EOE_