This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perlapi: Clarify Latin1 and ISO-8859-1
[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}
32require 'regen/regen_lib.pl';
e8e591c9 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->())) {
94bdecf9
JH
113 last DOC if $doc =~ /^=\w+/;
114 if ($doc =~ m:^\*/$:) {
115 warn "=cut missing? $file:$line:$doc";;
116 last DOC;
117 }
118 $docs .= $doc;
119 }
120 $docs = "\n$docs" if $docs and $docs !~ /^\n/;
5ce57792
NC
121
122 # Check the consistency of the flags
123 my ($embed_where, $inline_where);
124 my ($embed_may_change, $inline_may_change);
125
126 my $docref = delete $funcflags{$name};
127 if ($docref and %$docref) {
128 $embed_where = $docref->{flags} =~ /A/ ? 'api' : 'guts';
129 $embed_may_change = $docref->{flags} =~ /M/;
d4e99c76 130 $flags .= 'D' if $docref->{flags} =~ /D/;
5ce57792
NC
131 } else {
132 $missing{$name} = $file;
94bdecf9 133 }
5ce57792
NC
134 if ($flags =~ /m/) {
135 $inline_where = $flags =~ /A/ ? 'api' : 'guts';
136 $inline_may_change = $flags =~ /x/;
137
138 if (defined $embed_where && $inline_where ne $embed_where) {
139 warn "Function '$name' inconsistency: embed.fnc says $embed_where, Pod says $inline_where";
140 }
141
142 if (defined $embed_may_change
143 && $inline_may_change ne $embed_may_change) {
144 my $message = "Function '$name' inconsistency: ";
145 if ($embed_may_change) {
146 $message .= "embed.fnc says 'may change', Pod does not";
147 } else {
148 $message .= "Pod says 'may change', embed.fnc does not";
149 }
150 warn $message;
151 }
152 } elsif (!defined $embed_where) {
153 warn "Unable to place $name!\n";
154 next;
155 } else {
156 $inline_where = $embed_where;
157 $flags .= 'x' if $embed_may_change;
158 @args = @{$docref->{args}};
159 $ret = $docref->{retval};
94bdecf9 160 }
5ce57792 161
7a6610ca
DM
162 if (exists $docs{$inline_where}{$curheader}{$name}) {
163 warn "$0: duplicate API entry for '$name' in $inline_where/$curheader\n";
164 next;
165 }
5ce57792
NC
166 $docs{$inline_where}{$curheader}{$name}
167 = [$flags, $docs, $ret, $file, @args];
168
151c3fe5
KW
169 # Create a special entry with an empty-string name for the
170 # heading-level documentation.
171 if (defined $header_doc) {
172 $docs{$inline_where}{$curheader}{""} = $header_doc;
173 undef $header_doc;
174 }
175
94bdecf9 176 if (defined $doc) {
e509e693 177 if ($doc =~ /^=(?:for|head)/) {
94bdecf9
JH
178 $in = $doc;
179 redo FUNC;
180 }
181 } else {
182 warn "$file:$line:$in";
183 }
184 }
185 }
186}
187
188sub docout ($$$) { # output the docs for one function
189 my($fh, $name, $docref) = @_;
190 my($flags, $docs, $ret, $file, @args) = @$docref;
d8c40edc 191 $name =~ s/\s*$//;
94bdecf9 192
d4e99c76
KW
193 if ($flags =~ /D/) {
194 $docs = "\n\nDEPRECATED! It is planned to remove this function from a
195future release of Perl. Do not use it for new code; remove it from
196existing code.\n\n$docs";
197 }
198 else {
58a428bb
KW
199 $docs = "\n\nNOTE: this function is experimental and may change or be
200removed without notice.\n\n$docs" if $flags =~ /x/;
d4e99c76 201 }
94bdecf9
JH
202 $docs .= "NOTE: the perl_ form of this function is deprecated.\n\n"
203 if $flags =~ /p/;
5afac1eb
BM
204 $docs .= "NOTE: this function must be explicitly called as Perl_$name with an aTHX_ parameter.\n\n"
205 if $flags =~ /o/;
94bdecf9 206
d8c40edc 207 print $fh "=item $name\nX<$name>\n$docs";
94bdecf9
JH
208
209 if ($flags =~ /U/) { # no usage
210 # nothing
211 } elsif ($flags =~ /s/) { # semicolon ("dTHR;")
212 print $fh "\t\t$name;\n\n";
213 } elsif ($flags =~ /n/) { # no args
214 print $fh "\t$ret\t$name\n\n";
215 } else { # full usage
dee6204d
FC
216 my $p = $flags =~ /o/; # no #define foo Perl_foo
217 my $n = "Perl_"x$p . $name;
218 my $large_ret = length $ret > 7;
219 my $indent_size = 7+8 # nroff: 7 under =head + 8 under =item
220 +8+($large_ret ? 1 + length $ret : 8)
221 +length($n) + 1;
222 my $indent;
223 print $fh "\t$ret" . ($large_ret ? ' ' : "\t") . "$n(";
06e9ce89
FC
224 my $long_args;
225 for (@args) {
d2086f64 226 if ($indent_size + 2 + length > 79) {
06e9ce89
FC
227 $long_args=1;
228 $indent_size -= length($n) - 3;
229 last;
230 }
231 }
232 my $args = '';
233 if ($p) {
234 $args = @args ? "pTHX_ " : "pTHX";
235 if ($long_args) { print $fh $args; $args = '' }
236 }
237 $long_args and print $fh "\n";
238 my $first = !$long_args;
dee6204d
FC
239 while () {
240 if (!@args or
241 length $args
d2086f64 242 && $indent_size + 3 + length($args[0]) + length $args > 79
dee6204d
FC
243 ) {
244 print $fh
245 $first ? '' : (
246 $indent //=
247 "\t".($large_ret ? " " x (1+length $ret) : "\t")
06e9ce89 248 ." "x($long_args ? 4 : 1 + length $n)
dee6204d
FC
249 ),
250 $args, (","x($args ne 'pTHX_ ') . "\n")x!!@args;
251 $args = $first = '';
252 }
253 @args or last;
254 $args .= ", "x!!(length $args && $args ne 'pTHX_ ')
255 . shift @args;
256 }
06e9ce89 257 if ($long_args) { print $fh "\n", substr $indent, 0, -4 }
dee6204d 258 print $fh ")\n\n";
94bdecf9
JH
259 }
260 print $fh "=for hackers\nFound in file $file\n\n";
261}
262
f83c6033
KW
263sub sort_helper {
264 # Do a case-insensitive dictionary sort, with only alphabetics
265 # significant, falling back to using everything for determinancy
1354d57e 266 return (uc($a =~ s/[[:^alpha:]]//r) cmp uc($b =~ s/[[:^alpha:]]//r))
f83c6033
KW
267 || uc($a) cmp uc($b)
268 || $a cmp $b;
269}
270
7b73ff98 271sub output {
5a0155e6 272 my ($podname, $header, $dochash, $missing, $footer) = @_;
7882b24a
NC
273 my $fh = open_new("pod/$podname.pod", undef,
274 {by => "$0 extracting documentation",
f1f44974 275 from => 'the C source files'}, 1);
e0492643 276
7882b24a 277 print $fh $header;
e0492643 278
7b73ff98 279 my $key;
f83c6033 280 for $key (sort sort_helper keys %$dochash) {
7b73ff98 281 my $section = $dochash->{$key};
151c3fe5
KW
282 print $fh "\n=head1 $key\n\n";
283
284 # Output any heading-level documentation and delete so won't get in
285 # the way later
286 if (exists $section->{""}) {
287 print $fh $section->{""} . "\n";
288 delete $section->{""};
289 }
290 print $fh "=over 8\n\n";
291
f83c6033 292 for my $key (sort sort_helper keys %$section) {
7b73ff98
NC
293 docout($fh, $key, $section->{$key});
294 }
295 print $fh "\n=back\n";
296 }
297
5a0155e6 298 if (@$missing) {
a23e6e20 299 print $fh "\n=head1 Undocumented functions\n\n";
2616800a 300 print $fh $podname eq 'perlapi' ? <<'_EOB_' : <<'_EOB_';
474d0ac8 301The following functions have been flagged as part of the public API,
72d33970 302but are currently undocumented. Use them at your own risk, as the
ba4591a5
KW
303interfaces are subject to change. Functions that are not listed in this
304document are not intended for public use, and should NOT be used under any
305circumstances.
306
5a4fed09
KW
307If you feel you need to use one of these functions, first send email to
308L<perl5-porters@perl.org|mailto:perl5-porters@perl.org>. It may be
309that there is a good reason for the function not being documented, and it
310should be removed from this list; or it may just be that no one has gotten
311around to documenting it. In the latter case, you will be asked to submit a
312patch to document the function. Once your patch is accepted, it will indicate
313that the interface is stable (unless it is explicitly marked otherwise) and
314usable by you.
cf5f2f8f
KW
315
316=over
317
318_EOB_
2616800a
FC
319The following functions are currently undocumented. If you use one of
320them, you may wish to consider creating and submitting documentation for
321it.
322
323=over
324
325_EOB_
cf5f2f8f
KW
326 for my $missing (sort @$missing) {
327 print $fh "=item $missing\nX<$missing>\n\n";
5a0155e6 328 }
cf5f2f8f
KW
329 print $fh "=back\n\n";
330}
7882b24a 331 print $fh $footer, "=cut\n";
5a0155e6 332
7882b24a 333 read_only_bottom_close_and_rename($fh);
cd093254
MM
334}
335
e8e591c9
NC
336foreach (@{(setup_embed())[0]}) {
337 next if @$_ < 2;
338 my ($flags, $retval, $func, @args) = @$_;
339 s/\b(?:NN|NULLOK)\b\s+//g for @args;
bc350081 340
5ce57792
NC
341 $funcflags{$func} = {
342 flags => $flags,
343 retval => $retval,
344 args => \@args,
345 };
346}
347
5ce57792
NC
348# glob() picks up docs from extra .c or .h files that may be in unclean
349# development trees.
741c0772
NC
350open my $fh, '<', 'MANIFEST'
351 or die "Can't open MANIFEST: $!";
352while (my $line = <$fh>) {
353 next unless my ($file) = $line =~ /^(\S+\.[ch])\t/;
5ce57792 354
5ce57792
NC
355 open F, "< $file" or die "Cannot open $file for docs: $!\n";
356 $curheader = "Functions in file $file\n";
357 autodoc(\*F,$file);
358 close F or die "Error closing $file: $!\n";
359}
741c0772 360close $fh or die "Error whilst reading MANIFEST: $!";
5ce57792
NC
361
362for (sort keys %funcflags) {
363 next unless $funcflags{$_}{flags} =~ /d/;
364 warn "no docs for $_\n"
bc350081 365}
94bdecf9 366
5ce57792
NC
367foreach (sort keys %missing) {
368 next if $macro{$_};
369 # Heuristics for known not-a-function macros:
370 next if /^[A-Z]/;
371 next if /^dj?[A-Z]/;
372
373 warn "Function '$_', documented in $missing{$_}, not listed in embed.fnc";
94bdecf9
JH
374}
375
5ce57792
NC
376# walk table providing an array of components in each line to
377# subroutine, printing the result
378
8c869419
KW
379# List of funcs in the public API that aren't also marked as experimental nor
380# deprecated.
381my @missing_api = grep $funcflags{$_}{flags} =~ /A/ && $funcflags{$_}{flags} !~ /[MD]/ && !$docs{api}{$_}, keys %funcflags;
5a0155e6 382output('perlapi', <<'_EOB_', $docs{api}, \@missing_api, <<'_EOE_');
94bdecf9
JH
383=head1 NAME
384
385perlapi - autogenerated documentation for the perl public API
386
387=head1 DESCRIPTION
d8c40edc 388X<Perl API> X<API> X<api>
94bdecf9
JH
389
390This file contains the documentation of the perl public API generated by
ef9741a5 391F<embed.pl>, specifically a listing of functions, macros, flags, and variables
cf5f2f8f
KW
392that may be used by extension writers. L<At the end|/Undocumented functions>
393is a list of functions which have yet to be documented. The interfaces of
37a519b2 394those are subject to change without notice. Anything not listed here is
cf5f2f8f
KW
395not part of the public API, and should not be used by extension writers at
396all. For these reasons, blindly using functions listed in proto.h is to be
397avoided when writing extensions.
94bdecf9 398
dc6b0978
KW
399In Perl, unlike C, a string of characters may generally contain embedded
400C<NUL> characters. Sometimes in the documentation a Perl string is referred
401to as a "buffer" to distinguish it from a C string, but sometimes they are
402both just referred to as strings.
403
94bdecf9 404Note that all Perl API global variables must be referenced with the C<PL_>
37a519b2
KW
405prefix. Again, those not listed here are not to be used by extension writers,
406and can be changed or removed without notice; same with macros.
407Some macros are provided for compatibility with the older,
94bdecf9
JH
408unadorned names, but this support may be disabled in a future release.
409
2bbc8d55
SP
410Perl was originally written to handle US-ASCII only (that is characters
411whose ordinal numbers are in the range 0 - 127).
412And documentation and comments may still use the term ASCII, when
413sometimes in fact the entire range from 0 - 255 is meant.
414
14d32fa9
KW
415The non-ASCII characters below 256 can have various meanings, depending on
416various things. (See, most notably, L<perllocale>.) But usually the whole
417range can be referred to as ISO-8859-1. Often, the term "Latin-1" (or
418"Latin1") is used as an equivalent for ISO-8859-1. But some people treat
419"Latin1" as referring just to the characters in the range 160 through 255.
420This documentation uses "Latin1" and "Latin-1" to refer to all 256 characters.
421
89d7b922
KW
422Note that Perl can be compiled and run under either ASCII or EBCDIC (See
423L<perlebcdic>). Most of the documentation (and even comments in the code)
2bbc8d55
SP
424ignore the EBCDIC possibility.
425For almost all purposes the differences are transparent.
426As an example, under EBCDIC,
427instead of UTF-8, UTF-EBCDIC is used to encode Unicode strings, and so
428whenever this documentation refers to C<utf8>
429(and variants of that name, including in function names),
430it also (essentially transparently) means C<UTF-EBCDIC>.
431But the ordinals of characters differ between ASCII, EBCDIC, and
c0236afe
KW
432the UTF- encodings, and a string encoded in UTF-EBCDIC may occupy a different
433number of bytes than in UTF-8.
2bbc8d55 434
2bbc8d55 435The listing below is alphabetical, case insensitive.
94bdecf9
JH
436
437_EOB_
438
94bdecf9
JH
439=head1 AUTHORS
440
441Until May 1997, this document was maintained by Jeff Okamoto
442<okamoto@corp.hp.com>. It is now maintained as part of Perl itself.
443
444With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
445Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
446Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
447Stephen McCamant, and Gurusamy Sarathy.
448
449API Listing originally by Dean Roehrich <roehrich@cray.com>.
450
451Updated to be autogenerated from comments in the source by Benjamin Stuhl.
452
453=head1 SEE ALSO
454
b92fc6c1 455L<perlguts>, L<perlxs>, L<perlxstut>, L<perlintern>
94bdecf9
JH
456
457_EOE_
458
79fc8511
FC
459# List of non-static internal functions
460my @missing_guts =
461 grep $funcflags{$_}{flags} !~ /[As]/ && !$docs{guts}{$_}, keys %funcflags;
5a0155e6
TC
462
463output('perlintern', <<'END', $docs{guts}, \@missing_guts, <<'END');
94bdecf9
JH
464=head1 NAME
465
466perlintern - autogenerated documentation of purely B<internal>
467 Perl functions
468
469=head1 DESCRIPTION
d8c40edc 470X<internal Perl functions> X<interpreter functions>
94bdecf9
JH
471
472This file is the autogenerated documentation of functions in the
473Perl interpreter that are documented using Perl's internal documentation
154e47c8 474format but are not marked as part of the Perl API. In other words,
94bdecf9
JH
475B<they are not for use in extensions>!
476
477END
478
94bdecf9
JH
479=head1 AUTHORS
480
481The autodocumentation system was originally added to the Perl core by
154e47c8 482Benjamin Stuhl. Documentation is by whoever was kind enough to
94bdecf9
JH
483document their functions.
484
485=head1 SEE ALSO
486
b92fc6c1 487L<perlguts>, L<perlapi>
94bdecf9
JH
488
489END