This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
autodoc.pl: Refactor slightly
[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
56a0c332 25use strict;
a64c954a 26
7882b24a
NC
27if (@ARGV) {
28 my $workdir = shift;
29 chdir $workdir
30 or die "Couldn't chdir to '$workdir': $!";
31}
3d7c117d
MB
32require './regen/regen_lib.pl';
33require './regen/embed_lib.pl';
7882b24a 34
94bdecf9 35#
346f75ff 36# See database of global and static function prototypes in embed.fnc
94bdecf9
JH
37# This is used to generate prototype headers under various configurations,
38# export symbols lists for different platforms, and macros to provide an
39# implicit interpreter context argument.
40#
41
6a235718 42my %docs;
5ce57792
NC
43my %funcflags;
44my %macro = (
45 ax => 1,
46 items => 1,
47 ix => 1,
48 svtype => 1,
49 );
50my %missing;
94bdecf9
JH
51
52my $curheader = "Unknown section";
53
54sub autodoc ($$) { # parse a file and extract documentation info
55 my($fh,$file) = @_;
151c3fe5 56 my($in, $doc, $line, $header_doc);
f554dfc5
MH
57
58 # Count lines easier
59 my $get_next_line = sub { $line++; return <$fh> };
60
94bdecf9 61FUNC:
f554dfc5 62 while (defined($in = $get_next_line->())) {
5ce57792
NC
63 if ($in =~ /^#\s*define\s+([A-Za-z_][A-Za-z_0-9]+)\(/ &&
64 ($file ne 'embed.h' || $file ne 'proto.h')) {
65 $macro{$1} = $file;
66 next FUNC;
67 }
94bdecf9
JH
68 if ($in=~ /^=head1 (.*)/) {
69 $curheader = $1;
151c3fe5 70
f554dfc5
MH
71 # If the next non-space line begins with a word char, then it is
72 # the start of heading-ldevel documentation.
73 if (defined($doc = $get_next_line->())) {
74 # Skip over empty lines
75 while ($doc =~ /^\s+$/) {
76 if (! defined($doc = $get_next_line->())) {
77 next FUNC;
78 }
79 }
80
151c3fe5
KW
81 if ($doc !~ /^\w/) {
82 $in = $doc;
83 redo FUNC;
84 }
85 $header_doc = $doc;
151c3fe5
KW
86
87 # Continue getting the heading-level documentation until read
88 # in any pod directive (or as a fail-safe, find a closing
89 # comment to this pod in a C language file
90HDR_DOC:
f554dfc5 91 while (defined($doc = $get_next_line->())) {
151c3fe5
KW
92 if ($doc =~ /^=\w/) {
93 $in = $doc;
94 redo FUNC;
95 }
151c3fe5
KW
96
97 if ($doc =~ m:^\s*\*/$:) {
98 warn "=cut missing? $file:$line:$doc";;
99 last HDR_DOC;
100 }
101 $header_doc .= $doc;
102 }
103 }
94bdecf9
JH
104 next FUNC;
105 }
78c9d763 106 if ($in =~ /^=for\s+apidoc\s+(.*?)\s*\n/) {
94bdecf9
JH
107 my $proto = $1;
108 $proto = "||$proto" unless $proto =~ /\|/;
109 my($flags, $ret, $name, @args) = split /\|/, $proto;
110 my $docs = "";
111DOC:
f554dfc5 112 while (defined($doc = $get_next_line->())) {
72d4186d
KW
113
114 # Other pod commands are considered part of the current
115 # function's docs, so can have lists, etc.
116 last DOC if $doc =~ /^=(cut|for\s+apidoc|head)/;
94bdecf9
JH
117 if ($doc =~ m:^\*/$:) {
118 warn "=cut missing? $file:$line:$doc";;
119 last DOC;
72d4186d 120 }
94bdecf9
JH
121 $docs .= $doc;
122 }
123 $docs = "\n$docs" if $docs and $docs !~ /^\n/;
5ce57792
NC
124
125 # Check the consistency of the flags
126 my ($embed_where, $inline_where);
127 my ($embed_may_change, $inline_may_change);
128
d40e888a
KW
129 my $embed_docref = delete $funcflags{$name};
130 if ($embed_docref and %$embed_docref) {
131 $embed_where = $embed_docref->{flags} =~ /A/ ? 'api' : 'guts';
132 $embed_may_change = $embed_docref->{flags} =~ /M/;
133 $flags .= 'D' if $embed_docref->{flags} =~ /D/;
5ce57792
NC
134 } else {
135 $missing{$name} = $file;
94bdecf9 136 }
5ce57792
NC
137 if ($flags =~ /m/) {
138 $inline_where = $flags =~ /A/ ? 'api' : 'guts';
139 $inline_may_change = $flags =~ /x/;
140
141 if (defined $embed_where && $inline_where ne $embed_where) {
142 warn "Function '$name' inconsistency: embed.fnc says $embed_where, Pod says $inline_where";
143 }
144
145 if (defined $embed_may_change
146 && $inline_may_change ne $embed_may_change) {
147 my $message = "Function '$name' inconsistency: ";
148 if ($embed_may_change) {
149 $message .= "embed.fnc says 'may change', Pod does not";
150 } else {
151 $message .= "Pod says 'may change', embed.fnc does not";
152 }
153 warn $message;
154 }
155 } elsif (!defined $embed_where) {
156 warn "Unable to place $name!\n";
157 next;
158 } else {
159 $inline_where = $embed_where;
160 $flags .= 'x' if $embed_may_change;
d40e888a
KW
161 @args = @{$embed_docref->{args}};
162 $ret = $embed_docref->{retval};
94bdecf9 163 }
5ce57792 164
7a6610ca
DM
165 if (exists $docs{$inline_where}{$curheader}{$name}) {
166 warn "$0: duplicate API entry for '$name' in $inline_where/$curheader\n";
167 next;
168 }
5ce57792
NC
169 $docs{$inline_where}{$curheader}{$name}
170 = [$flags, $docs, $ret, $file, @args];
171
151c3fe5
KW
172 # Create a special entry with an empty-string name for the
173 # heading-level documentation.
174 if (defined $header_doc) {
175 $docs{$inline_where}{$curheader}{""} = $header_doc;
176 undef $header_doc;
177 }
178
94bdecf9 179 if (defined $doc) {
e509e693 180 if ($doc =~ /^=(?:for|head)/) {
94bdecf9
JH
181 $in = $doc;
182 redo FUNC;
183 }
184 } else {
185 warn "$file:$line:$in";
186 }
187 }
188 }
189}
190
191sub docout ($$$) { # output the docs for one function
192 my($fh, $name, $docref) = @_;
193 my($flags, $docs, $ret, $file, @args) = @$docref;
d8c40edc 194 $name =~ s/\s*$//;
94bdecf9 195
d4e99c76
KW
196 if ($flags =~ /D/) {
197 $docs = "\n\nDEPRECATED! It is planned to remove this function from a
198future release of Perl. Do not use it for new code; remove it from
199existing code.\n\n$docs";
200 }
201 else {
58a428bb
KW
202 $docs = "\n\nNOTE: this function is experimental and may change or be
203removed without notice.\n\n$docs" if $flags =~ /x/;
d4e99c76 204 }
94bdecf9
JH
205 $docs .= "NOTE: the perl_ form of this function is deprecated.\n\n"
206 if $flags =~ /p/;
5afac1eb
BM
207 $docs .= "NOTE: this function must be explicitly called as Perl_$name with an aTHX_ parameter.\n\n"
208 if $flags =~ /o/;
94bdecf9 209
d8c40edc 210 print $fh "=item $name\nX<$name>\n$docs";
94bdecf9
JH
211
212 if ($flags =~ /U/) { # no usage
213 # nothing
05ca4832
KW
214 } else {
215 if ($flags =~ /s/) { # semicolon ("dTHR;")
216 print $fh "\t\t$name;";
94bdecf9 217 } elsif ($flags =~ /n/) { # no args
05ca4832 218 print $fh "\t$ret\t$name";
94bdecf9 219 } else { # full usage
dee6204d
FC
220 my $p = $flags =~ /o/; # no #define foo Perl_foo
221 my $n = "Perl_"x$p . $name;
222 my $large_ret = length $ret > 7;
223 my $indent_size = 7+8 # nroff: 7 under =head + 8 under =item
224 +8+($large_ret ? 1 + length $ret : 8)
225 +length($n) + 1;
226 my $indent;
227 print $fh "\t$ret" . ($large_ret ? ' ' : "\t") . "$n(";
06e9ce89
FC
228 my $long_args;
229 for (@args) {
d2086f64 230 if ($indent_size + 2 + length > 79) {
06e9ce89
FC
231 $long_args=1;
232 $indent_size -= length($n) - 3;
233 last;
234 }
235 }
236 my $args = '';
237 if ($p) {
238 $args = @args ? "pTHX_ " : "pTHX";
239 if ($long_args) { print $fh $args; $args = '' }
240 }
241 $long_args and print $fh "\n";
242 my $first = !$long_args;
dee6204d
FC
243 while () {
244 if (!@args or
245 length $args
d2086f64 246 && $indent_size + 3 + length($args[0]) + length $args > 79
dee6204d
FC
247 ) {
248 print $fh
249 $first ? '' : (
250 $indent //=
251 "\t".($large_ret ? " " x (1+length $ret) : "\t")
06e9ce89 252 ." "x($long_args ? 4 : 1 + length $n)
dee6204d
FC
253 ),
254 $args, (","x($args ne 'pTHX_ ') . "\n")x!!@args;
255 $args = $first = '';
256 }
257 @args or last;
258 $args .= ", "x!!(length $args && $args ne 'pTHX_ ')
259 . shift @args;
260 }
06e9ce89 261 if ($long_args) { print $fh "\n", substr $indent, 0, -4 }
05ca4832
KW
262 print $fh ")";
263 }
264 print $fh "\n\n";
94bdecf9
JH
265 }
266 print $fh "=for hackers\nFound in file $file\n\n";
267}
268
f83c6033
KW
269sub sort_helper {
270 # Do a case-insensitive dictionary sort, with only alphabetics
271 # significant, falling back to using everything for determinancy
1354d57e 272 return (uc($a =~ s/[[:^alpha:]]//r) cmp uc($b =~ s/[[:^alpha:]]//r))
f83c6033
KW
273 || uc($a) cmp uc($b)
274 || $a cmp $b;
275}
276
7b73ff98 277sub output {
5a0155e6 278 my ($podname, $header, $dochash, $missing, $footer) = @_;
6a4c4cd4
DM
279 #
280 # strip leading '|' from each line which had been used to hide
281 # pod from pod checkers.
282 s/^\|//gm for $header, $footer;
283
7882b24a
NC
284 my $fh = open_new("pod/$podname.pod", undef,
285 {by => "$0 extracting documentation",
f1f44974 286 from => 'the C source files'}, 1);
e0492643 287
7882b24a 288 print $fh $header;
e0492643 289
7b73ff98 290 my $key;
f83c6033 291 for $key (sort sort_helper keys %$dochash) {
7b73ff98 292 my $section = $dochash->{$key};
151c3fe5
KW
293 print $fh "\n=head1 $key\n\n";
294
295 # Output any heading-level documentation and delete so won't get in
296 # the way later
297 if (exists $section->{""}) {
298 print $fh $section->{""} . "\n";
299 delete $section->{""};
300 }
301 print $fh "=over 8\n\n";
302
f83c6033 303 for my $key (sort sort_helper keys %$section) {
7b73ff98
NC
304 docout($fh, $key, $section->{$key});
305 }
306 print $fh "\n=back\n";
307 }
308
5a0155e6 309 if (@$missing) {
a23e6e20 310 print $fh "\n=head1 Undocumented functions\n\n";
2616800a 311 print $fh $podname eq 'perlapi' ? <<'_EOB_' : <<'_EOB_';
474d0ac8 312The following functions have been flagged as part of the public API,
72d33970 313but are currently undocumented. Use them at your own risk, as the
ba4591a5
KW
314interfaces are subject to change. Functions that are not listed in this
315document are not intended for public use, and should NOT be used under any
316circumstances.
317
5a4fed09
KW
318If you feel you need to use one of these functions, first send email to
319L<perl5-porters@perl.org|mailto:perl5-porters@perl.org>. It may be
320that there is a good reason for the function not being documented, and it
321should be removed from this list; or it may just be that no one has gotten
322around to documenting it. In the latter case, you will be asked to submit a
323patch to document the function. Once your patch is accepted, it will indicate
324that the interface is stable (unless it is explicitly marked otherwise) and
325usable by you.
cf5f2f8f 326_EOB_
2616800a
FC
327The following functions are currently undocumented. If you use one of
328them, you may wish to consider creating and submitting documentation for
329it.
2616800a 330_EOB_
6a4c4cd4
DM
331 print $fh "\n=over\n\n";
332
cf5f2f8f
KW
333 for my $missing (sort @$missing) {
334 print $fh "=item $missing\nX<$missing>\n\n";
5a0155e6 335 }
cf5f2f8f
KW
336 print $fh "=back\n\n";
337}
7882b24a 338 print $fh $footer, "=cut\n";
5a0155e6 339
7882b24a 340 read_only_bottom_close_and_rename($fh);
cd093254
MM
341}
342
e8e591c9
NC
343foreach (@{(setup_embed())[0]}) {
344 next if @$_ < 2;
345 my ($flags, $retval, $func, @args) = @$_;
346 s/\b(?:NN|NULLOK)\b\s+//g for @args;
bc350081 347
5ce57792
NC
348 $funcflags{$func} = {
349 flags => $flags,
350 retval => $retval,
351 args => \@args,
352 };
353}
354
5ce57792
NC
355# glob() picks up docs from extra .c or .h files that may be in unclean
356# development trees.
741c0772
NC
357open my $fh, '<', 'MANIFEST'
358 or die "Can't open MANIFEST: $!";
359while (my $line = <$fh>) {
360 next unless my ($file) = $line =~ /^(\S+\.[ch])\t/;
5ce57792 361
1ae6ead9 362 open F, '<', $file or die "Cannot open $file for docs: $!\n";
5ce57792
NC
363 $curheader = "Functions in file $file\n";
364 autodoc(\*F,$file);
365 close F or die "Error closing $file: $!\n";
366}
741c0772 367close $fh or die "Error whilst reading MANIFEST: $!";
5ce57792
NC
368
369for (sort keys %funcflags) {
370 next unless $funcflags{$_}{flags} =~ /d/;
371 warn "no docs for $_\n"
bc350081 372}
94bdecf9 373
5ce57792
NC
374foreach (sort keys %missing) {
375 next if $macro{$_};
376 # Heuristics for known not-a-function macros:
377 next if /^[A-Z]/;
378 next if /^dj?[A-Z]/;
379
380 warn "Function '$_', documented in $missing{$_}, not listed in embed.fnc";
94bdecf9
JH
381}
382
5ce57792
NC
383# walk table providing an array of components in each line to
384# subroutine, printing the result
385
8c869419
KW
386# List of funcs in the public API that aren't also marked as experimental nor
387# deprecated.
388my @missing_api = grep $funcflags{$_}{flags} =~ /A/ && $funcflags{$_}{flags} !~ /[MD]/ && !$docs{api}{$_}, keys %funcflags;
5a0155e6 389output('perlapi', <<'_EOB_', $docs{api}, \@missing_api, <<'_EOE_');
6a4c4cd4
DM
390|=encoding UTF-8
391|
392|=head1 NAME
393|
394|perlapi - autogenerated documentation for the perl public API
395|
396|=head1 DESCRIPTION
397|X<Perl API> X<API> X<api>
398|
399|This file contains the documentation of the perl public API generated by
400|F<embed.pl>, specifically a listing of functions, macros, flags, and variables
401|that may be used by extension writers. L<At the end|/Undocumented functions>
402|is a list of functions which have yet to be documented. The interfaces of
403|those are subject to change without notice. Anything not listed here is
404|not part of the public API, and should not be used by extension writers at
405|all. For these reasons, blindly using functions listed in proto.h is to be
406|avoided when writing extensions.
407|
408|In Perl, unlike C, a string of characters may generally contain embedded
409|C<NUL> characters. Sometimes in the documentation a Perl string is referred
410|to as a "buffer" to distinguish it from a C string, but sometimes they are
411|both just referred to as strings.
412|
413|Note that all Perl API global variables must be referenced with the C<PL_>
414|prefix. Again, those not listed here are not to be used by extension writers,
415|and can be changed or removed without notice; same with macros.
416|Some macros are provided for compatibility with the older,
417|unadorned names, but this support may be disabled in a future release.
418|
419|Perl was originally written to handle US-ASCII only (that is characters
420|whose ordinal numbers are in the range 0 - 127).
421|And documentation and comments may still use the term ASCII, when
422|sometimes in fact the entire range from 0 - 255 is meant.
423|
424|The non-ASCII characters below 256 can have various meanings, depending on
425|various things. (See, most notably, L<perllocale>.) But usually the whole
426|range can be referred to as ISO-8859-1. Often, the term "Latin-1" (or
427|"Latin1") is used as an equivalent for ISO-8859-1. But some people treat
428|"Latin1" as referring just to the characters in the range 128 through 255, or
429|somethimes from 160 through 255.
430|This documentation uses "Latin1" and "Latin-1" to refer to all 256 characters.
431|
432|Note that Perl can be compiled and run under either ASCII or EBCDIC (See
433|L<perlebcdic>). Most of the documentation (and even comments in the code)
434|ignore the EBCDIC possibility.
435|For almost all purposes the differences are transparent.
436|As an example, under EBCDIC,
437|instead of UTF-8, UTF-EBCDIC is used to encode Unicode strings, and so
438|whenever this documentation refers to C<utf8>
439|(and variants of that name, including in function names),
440|it also (essentially transparently) means C<UTF-EBCDIC>.
441|But the ordinals of characters differ between ASCII, EBCDIC, and
442|the UTF- encodings, and a string encoded in UTF-EBCDIC may occupy a different
443|number of bytes than in UTF-8.
444|
445|The listing below is alphabetical, case insensitive.
446|
94bdecf9 447_EOB_
6a4c4cd4
DM
448|
449|=head1 AUTHORS
450|
451|Until May 1997, this document was maintained by Jeff Okamoto
452|<okamoto@corp.hp.com>. It is now maintained as part of Perl itself.
453|
454|With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
455|Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
456|Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
457|Stephen McCamant, and Gurusamy Sarathy.
458|
459|API Listing originally by Dean Roehrich <roehrich@cray.com>.
460|
461|Updated to be autogenerated from comments in the source by Benjamin Stuhl.
462|
463|=head1 SEE ALSO
464|
465|L<perlguts>, L<perlxs>, L<perlxstut>, L<perlintern>
466|
94bdecf9
JH
467_EOE_
468
79fc8511
FC
469# List of non-static internal functions
470my @missing_guts =
471 grep $funcflags{$_}{flags} !~ /[As]/ && !$docs{guts}{$_}, keys %funcflags;
5a0155e6
TC
472
473output('perlintern', <<'END', $docs{guts}, \@missing_guts, <<'END');
6a4c4cd4
DM
474|=head1 NAME
475|
476|perlintern - autogenerated documentation of purely B<internal>
477| Perl functions
478|
479|=head1 DESCRIPTION
480|X<internal Perl functions> X<interpreter functions>
481|
482|This file is the autogenerated documentation of functions in the
483|Perl interpreter that are documented using Perl's internal documentation
484|format but are not marked as part of the Perl API. In other words,
485|B<they are not for use in extensions>!
486|
94bdecf9 487END
6a4c4cd4
DM
488|
489|=head1 AUTHORS
490|
491|The autodocumentation system was originally added to the Perl core by
492|Benjamin Stuhl. Documentation is by whoever was kind enough to
493|document their functions.
494|
495|=head1 SEE ALSO
496|
497|L<perlguts>, L<perlapi>
498|
94bdecf9 499END