This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Fix maximum line length
[perl5.git] / pod / perlpodspec.pod
CommitLineData
49781f4a 1=encoding utf8
8a93676d
SB
2
3=head1 NAME
4
5perlpodspec - Plain Old Documentation: format specification and notes
6
7=head1 DESCRIPTION
8
9This document is detailed notes on the Pod markup language. Most
10people will only have to read L<perlpod|perlpod> to know how to write
11in Pod, but this document may answer some incidental questions to do
12with parsing and rendering Pod.
13
14In this document, "must" / "must not", "should" /
15"should not", and "may" have their conventional (cf. RFC 2119)
16meanings: "X must do Y" means that if X doesn't do Y, it's against
17this specification, and should really be fixed. "X should do Y"
18means that it's recommended, but X may fail to do Y, if there's a
19good reason. "X may do Y" is merely a note that X can do Y at
20will (although it is up to the reader to detect any connotation of
21"and I think it would be I<nice> if X did Y" versus "it wouldn't
22really I<bother> me if X did Y").
23
24Notably, when I say "the parser should do Y", the
25parser may fail to do Y, if the calling application explicitly
26requests that the parser I<not> do Y. I often phrase this as
27"the parser should, by default, do Y." This doesn't I<require>
28the parser to provide an option for turning off whatever
29feature Y is (like expanding tabs in verbatim paragraphs), although
30it implicates that such an option I<may> be provided.
31
32=head1 Pod Definitions
33
ac036724 34Pod is embedded in files, typically Perl source files, although you
8a93676d
SB
35can write a file that's nothing but Pod.
36
37A B<line> in a file consists of zero or more non-newline characters,
38terminated by either a newline or the end of the file.
39
40A B<newline sequence> is usually a platform-dependent concept, but
41Pod parsers should understand it to mean any of CR (ASCII 13), LF
42(ASCII 10), or a CRLF (ASCII 13 followed immediately by ASCII 10), in
43addition to any other system-specific meaning. The first CR/CRLF/LF
44sequence in the file may be used as the basis for identifying the
45newline sequence for parsing the rest of the file.
46
47A B<blank line> is a line consisting entirely of zero or more spaces
48(ASCII 32) or tabs (ASCII 9), and terminated by a newline or end-of-file.
49A B<non-blank line> is a line containing one or more characters other
50than space or tab (and terminated by a newline or end-of-file).
51
52(I<Note:> Many older Pod parsers did not accept a line consisting of
ac036724 53spaces/tabs and then a newline as a blank line. The only lines they
8a93676d
SB
54considered blank were lines consisting of I<no characters at all>,
55terminated by a newline.)
56
57B<Whitespace> is used in this document as a blanket term for spaces,
58tabs, and newline sequences. (By itself, this term usually refers
59to literal whitespace. That is, sequences of whitespace characters
60in Pod source, as opposed to "EE<lt>32>", which is a formatting
61code that I<denotes> a whitespace character.)
62
63A B<Pod parser> is a module meant for parsing Pod (regardless of
64whether this involves calling callbacks or building a parse tree or
65directly formatting it). A B<Pod formatter> (or B<Pod translator>)
66is a module or program that converts Pod to some other format (HTML,
67plaintext, TeX, PostScript, RTF). A B<Pod processor> might be a
68formatter or translator, or might be a program that does something
353c6505 69else with the Pod (like counting words, scanning for index points,
8a93676d
SB
70etc.).
71
72Pod content is contained in B<Pod blocks>. A Pod block starts with a
1bca558f 73line that matches C<m/\A=[a-zA-Z]/>, and continues up to the next line
ac036724 74that matches C<m/\A=cut/> or up to the end of the file if there is
8a93676d
SB
75no C<m/\A=cut/> line.
76
77=for comment
78 The current perlsyn says:
79 [beginquote]
80 Note that pod translators should look at only paragraphs beginning
81 with a pod directive (it makes parsing easier), whereas the compiler
82 actually knows to look for pod escapes even in the middle of a
83 paragraph. This means that the following secret stuff will be ignored
84 by both the compiler and the translators.
85 $a=3;
86 =secret stuff
87 warn "Neither POD nor CODE!?"
88 =cut back
89 print "got $a\n";
90 You probably shouldn't rely upon the warn() being podded out forever.
91 Not all pod translators are well-behaved in this regard, and perhaps
92 the compiler will become pickier.
93 [endquote]
94 I think that those paragraphs should just be removed; paragraph-based
95 parsing seems to have been largely abandoned, because of the hassle
96 with non-empty blank lines messing up what people meant by "paragraph".
97 Even if the "it makes parsing easier" bit were especially true,
98 it wouldn't be worth the confusion of having perl and pod2whatever
99 actually disagree on what can constitute a Pod block.
100
e1a97e07
KW
101Note that a parser is not expected to distinguish between something that
102looks like pod, but is in a quoted string, such as a here document.
103
8a93676d
SB
104Within a Pod block, there are B<Pod paragraphs>. A Pod paragraph
105consists of non-blank lines of text, separated by one or more blank
106lines.
107
108For purposes of Pod processing, there are four types of paragraphs in
109a Pod block:
110
111=over
112
113=item *
114
115A command paragraph (also called a "directive"). The first line of
116this paragraph must match C<m/\A=[a-zA-Z]/>. Command paragraphs are
117typically one line, as in:
118
119 =head1 NOTES
120
121 =item *
122
123But they may span several (non-blank) lines:
124
125 =for comment
126 Hm, I wonder what it would look like if
127 you tried to write a BNF for Pod from this.
210b36aa 128
8a93676d
SB
129 =head3 Dr. Strangelove, or: How I Learned to
130 Stop Worrying and Love the Bomb
131
132I<Some> command paragraphs allow formatting codes in their content
133(i.e., after the part that matches C<m/\A=[a-zA-Z]\S*\s*/>), as in:
134
135 =head1 Did You Remember to C<use strict;>?
136
137In other words, the Pod processing handler for "head1" will apply the
138same processing to "Did You Remember to CE<lt>use strict;>?" that it
ac036724 139would to an ordinary paragraph (i.e., formatting codes like
8a93676d
SB
140"CE<lt>...>") are parsed and presumably formatted appropriately, and
141whitespace in the form of literal spaces and/or tabs is not
142significant.
143
144=item *
145
146A B<verbatim paragraph>. The first line of this paragraph must be a
147literal space or tab, and this paragraph must not be inside a "=begin
148I<identifier>", ... "=end I<identifier>" sequence unless
149"I<identifier>" begins with a colon (":"). That is, if a paragraph
150starts with a literal space or tab, but I<is> inside a
151"=begin I<identifier>", ... "=end I<identifier>" region, then it's
152a data paragraph, unless "I<identifier>" begins with a colon.
153
154Whitespace I<is> significant in verbatim paragraphs (although, in
155processing, tabs are probably expanded).
156
157=item *
158
159An B<ordinary paragraph>. A paragraph is an ordinary paragraph
160if its first line matches neither C<m/\A=[a-zA-Z]/> nor
161C<m/\A[ \t]/>, I<and> if it's not inside a "=begin I<identifier>",
162... "=end I<identifier>" sequence unless "I<identifier>" begins with
163a colon (":").
164
165=item *
166
167A B<data paragraph>. This is a paragraph that I<is> inside a "=begin
168I<identifier>" ... "=end I<identifier>" sequence where
169"I<identifier>" does I<not> begin with a literal colon (":"). In
170some sense, a data paragraph is not part of Pod at all (i.e.,
171effectively it's "out-of-band"), since it's not subject to most kinds
172of Pod parsing; but it is specified here, since Pod
173parsers need to be able to call an event for it, or store it in some
174form in a parse tree, or at least just parse I<around> it.
175
176=back
177
178For example: consider the following paragraphs:
179
180 # <- that's the 0th column
181
182 =head1 Foo
210b36aa 183
8a93676d 184 Stuff
210b36aa 185
8a93676d 186 $foo->bar
210b36aa 187
8a93676d
SB
188 =cut
189
190Here, "=head1 Foo" and "=cut" are command paragraphs because the first
191line of each matches C<m/\A=[a-zA-Z]/>. "I<[space][space]>$foo->bar"
192is a verbatim paragraph, because its first line starts with a literal
193whitespace character (and there's no "=begin"..."=end" region around).
194
195The "=begin I<identifier>" ... "=end I<identifier>" commands stop
6fbdb1cc 196paragraphs that they surround from being parsed as ordinary or verbatim
8a93676d
SB
197paragraphs, if I<identifier> doesn't begin with a colon. This
198is discussed in detail in the section
199L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
200
201=head1 Pod Commands
202
203This section is intended to supplement and clarify the discussion in
204L<perlpod/"Command Paragraph">. These are the currently recognized
205Pod commands:
206
207=over
208
209=item "=head1", "=head2", "=head3", "=head4"
210
211This command indicates that the text in the remainder of the paragraph
212is a heading. That text may contain formatting codes. Examples:
213
214 =head1 Object Attributes
210b36aa 215
8a93676d
SB
216 =head3 What B<Not> to Do!
217
218=item "=pod"
219
220This command indicates that this paragraph begins a Pod block. (If we
221are already in the middle of a Pod block, this command has no effect at
222all.) If there is any text in this command paragraph after "=pod",
223it must be ignored. Examples:
224
225 =pod
210b36aa 226
8a93676d 227 This is a plain Pod paragraph.
210b36aa 228
8a93676d
SB
229 =pod This text is ignored.
230
231=item "=cut"
232
233This command indicates that this line is the end of this previously
234started Pod block. If there is any text after "=cut" on the line, it must be
235ignored. Examples:
236
237 =cut
238
239 =cut The documentation ends here.
240
241 =cut
242 # This is the first line of program text.
243 sub foo { # This is the second.
244
659cfd94 245It is an error to try to I<start> a Pod block with a "=cut" command. In
8a93676d
SB
246that case, the Pod processor must halt parsing of the input file, and
247must by default emit a warning.
248
249=item "=over"
250
251This command indicates that this is the start of a list/indent
252region. If there is any text following the "=over", it must consist
253of only a nonzero positive numeral. The semantics of this numeral is
254explained in the L</"About =over...=back Regions"> section, further
255below. Formatting codes are not expanded. Examples:
256
257 =over 3
210b36aa 258
8a93676d 259 =over 3.5
210b36aa 260
8a93676d
SB
261 =over
262
263=item "=item"
264
265This command indicates that an item in a list begins here. Formatting
266codes are processed. The semantics of the (optional) text in the
267remainder of this paragraph are
268explained in the L</"About =over...=back Regions"> section, further
269below. Examples:
270
271 =item
210b36aa 272
8a93676d 273 =item *
210b36aa 274
8a93676d 275 =item *
210b36aa 276
8a93676d 277 =item 14
210b36aa 278
8a93676d 279 =item 3.
210b36aa 280
8a93676d 281 =item C<< $thing->stuff(I<dodad>) >>
210b36aa 282
8a93676d
SB
283 =item For transporting us beyond seas to be tried for pretended
284 offenses
210b36aa 285
8a93676d
SB
286 =item He is at this time transporting large armies of foreign
287 mercenaries to complete the works of death, desolation and
288 tyranny, already begun with circumstances of cruelty and perfidy
289 scarcely paralleled in the most barbarous ages, and totally
290 unworthy the head of a civilized nation.
291
292=item "=back"
293
294This command indicates that this is the end of the region begun
295by the most recent "=over" command. It permits no text after the
296"=back" command.
297
298=item "=begin formatname"
299
93592fd5
RS
300=item "=begin formatname parameter"
301
8a93676d
SB
302This marks the following paragraphs (until the matching "=end
303formatname") as being for some special kind of processing. Unless
304"formatname" begins with a colon, the contained non-command
305paragraphs are data paragraphs. But if "formatname" I<does> begin
306with a colon, then non-command paragraphs are ordinary paragraphs
307or data paragraphs. This is discussed in detail in the section
308L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
309
310It is advised that formatnames match the regexp
c85e9b4c 311C<m/\A:?[-a-zA-Z0-9_]+\z/>. Everything following whitespace after the
93592fd5
RS
312formatname is a parameter that may be used by the formatter when dealing
313with this region. This parameter must not be repeated in the "=end"
314paragraph. Implementors should anticipate future expansion in the
315semantics and syntax of the first parameter to "=begin"/"=end"/"=for".
8a93676d
SB
316
317=item "=end formatname"
318
319This marks the end of the region opened by the matching
320"=begin formatname" region. If "formatname" is not the formatname
321of the most recent open "=begin formatname" region, then this
322is an error, and must generate an error message. This
323is discussed in detail in the section
324L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
325
326=item "=for formatname text..."
327
328This is synonymous with:
329
330 =begin formatname
210b36aa 331
8a93676d 332 text...
210b36aa 333
8a93676d
SB
334 =end formatname
335
336That is, it creates a region consisting of a single paragraph; that
337paragraph is to be treated as a normal paragraph if "formatname"
338begins with a ":"; if "formatname" I<doesn't> begin with a colon,
339then "text..." will constitute a data paragraph. There is no way
340to use "=for formatname text..." to express "text..." as a verbatim
341paragraph.
342
a179871b
SB
343=item "=encoding encodingname"
344
345This command, which should occur early in the document (at least
1e54db1a 346before any non-US-ASCII data!), declares that this document is
a179871b 347encoded in the encoding I<encodingname>, which must be
6fbdb1cc 348an encoding name that L<Encode> recognizes. (Encode's list
8a3f7e95 349of supported encodings, in L<Encode::Supported>, is useful here.)
a179871b
SB
350If the Pod parser cannot decode the declared encoding, it
351should emit a warning and may abort parsing the document
352altogether.
353
354A document having more than one "=encoding" line should be
355considered an error. Pod processors may silently tolerate this if
356the not-first "=encoding" lines are just duplicates of the
6fbdb1cc
RS
357first one (e.g., if there's a "=encoding utf8" line, and later on
358another "=encoding utf8" line). But Pod processors should complain if
a179871b
SB
359there are contradictory "=encoding" lines in the same document
360(e.g., if there is a "=encoding utf8" early in the document and
361"=encoding big5" later). Pod processors that recognize BOMs
362may also complain if they see an "=encoding" line
1e54db1a
JH
363that contradicts the BOM (e.g., if a document with a UTF-16LE
364BOM has an "=encoding shiftjis" line).
a179871b 365
8a93676d
SB
366=back
367
368If a Pod processor sees any command other than the ones listed
369above (like "=head", or "=haed1", or "=stuff", or "=cuttlefish",
370or "=w123"), that processor must by default treat this as an
371error. It must not process the paragraph beginning with that
372command, must by default warn of this as an error, and may
373abort the parse. A Pod parser may allow a way for particular
374applications to add to the above list of known commands, and to
375stipulate, for each additional command, whether formatting
376codes should be processed.
377
378Future versions of this specification may add additional
379commands.
380
381
382
383=head1 Pod Formatting Codes
384
385(Note that in previous drafts of this document and of perlpod,
386formatting codes were referred to as "interior sequences", and
387this term may still be found in the documentation for Pod parsers,
388and in error messages from Pod processors.)
389
390There are two syntaxes for formatting codes:
391
392=over
393
394=item *
395
396A formatting code starts with a capital letter (just US-ASCII [A-Z])
397followed by a "<", any number of characters, and ending with the first
398matching ">". Examples:
399
400 That's what I<you> think!
401
d8ff3e95 402 What's C<CORE::dump()> for?
8a93676d
SB
403
404 X<C<chmod> and C<unlink()> Under Different Operating Systems>
405
406=item *
407
408A formatting code starts with a capital letter (just US-ASCII [A-Z])
409followed by two or more "<"'s, one or more whitespace characters,
410any number of characters, one or more whitespace characters,
411and ending with the first matching sequence of two or more ">"'s, where
412the number of ">"'s equals the number of "<"'s in the opening of this
413formatting code. Examples:
414
415 That's what I<< you >> think!
416
417 C<<< open(X, ">>thing.dat") || die $! >>>
418
419 B<< $foo->bar(); >>
420
421With this syntax, the whitespace character(s) after the "CE<lt><<"
1bca558f 422and before the ">>>" (or whatever letter) are I<not> renderable. They
8a93676d
SB
423do not signify whitespace, are merely part of the formatting codes
424themselves. That is, these are all synonymous:
425
426 C<thing>
427 C<< thing >>
428 C<< thing >>
429 C<<< thing >>>
430 C<<<<
431 thing
432 >>>>
433
434and so on.
435
a3d78747
RS
436Finally, the multiple-angle-bracket form does I<not> alter the interpretation
437of nested formatting codes, meaning that the following four example lines are
438identical in meaning:
439
440 B<example: C<$a E<lt>=E<gt> $b>>
441
442 B<example: C<< $a <=> $b >>>
443
444 B<example: C<< $a E<lt>=E<gt> $b >>>
445
446 B<<< example: C<< $a E<lt>=E<gt> $b >> >>>
447
8a93676d
SB
448=back
449
450In parsing Pod, a notably tricky part is the correct parsing of
451(potentially nested!) formatting codes. Implementors should
452consult the code in the C<parse_text> routine in Pod::Parser as an
453example of a correct implementation.
454
455=over
456
457=item C<IE<lt>textE<gt>> -- italic text
458
459See the brief discussion in L<perlpod/"Formatting Codes">.
460
461=item C<BE<lt>textE<gt>> -- bold text
462
463See the brief discussion in L<perlpod/"Formatting Codes">.
464
465=item C<CE<lt>codeE<gt>> -- code text
466
467See the brief discussion in L<perlpod/"Formatting Codes">.
468
469=item C<FE<lt>filenameE<gt>> -- style for filenames
470
471See the brief discussion in L<perlpod/"Formatting Codes">.
472
473=item C<XE<lt>topic nameE<gt>> -- an index entry
474
475See the brief discussion in L<perlpod/"Formatting Codes">.
476
477This code is unusual in that most formatters completely discard
478this code and its content. Other formatters will render it with
479invisible codes that can be used in building an index of
480the current document.
481
482=item C<ZE<lt>E<gt>> -- a null (zero-effect) formatting code
483
484Discussed briefly in L<perlpod/"Formatting Codes">.
485
c195f169 486This code is unusual in that it should have no content. That is,
8a93676d
SB
487a processor may complain if it sees C<ZE<lt>potatoesE<gt>>. Whether
488or not it complains, the I<potatoes> text should ignored.
489
490=item C<LE<lt>nameE<gt>> -- a hyperlink
491
492The complicated syntaxes of this code are discussed at length in
493L<perlpod/"Formatting Codes">, and implementation details are
494discussed below, in L</"About LE<lt>...E<gt> Codes">. Parsing the
495contents of LE<lt>content> is tricky. Notably, the content has to be
496checked for whether it looks like a URL, or whether it has to be split
497on literal "|" and/or "/" (in the right order!), and so on,
498I<before> EE<lt>...> codes are resolved.
499
500=item C<EE<lt>escapeE<gt>> -- a character escape
501
502See L<perlpod/"Formatting Codes">, and several points in
503L</Notes on Implementing Pod Processors>.
504
505=item C<SE<lt>textE<gt>> -- text contains non-breaking spaces
506
507This formatting code is syntactically simple, but semantically
508complex. What it means is that each space in the printable
3e666715 509content of this code signifies a non-breaking space.
8a93676d
SB
510
511Consider:
512
513 C<$x ? $y : $z>
514
515 S<C<$x ? $y : $z>>
516
517Both signify the monospace (c[ode] style) text consisting of
518"$x", one space, "?", one space, ":", one space, "$z". The
519difference is that in the latter, with the S code, those spaces
3e666715 520are not "normal" spaces, but instead are non-breaking spaces.
8a93676d
SB
521
522=back
523
524
525If a Pod processor sees any formatting code other than the ones
526listed above (as in "NE<lt>...>", or "QE<lt>...>", etc.), that
527processor must by default treat this as an error.
528A Pod parser may allow a way for particular
529applications to add to the above list of known formatting codes;
530a Pod parser might even allow a way to stipulate, for each additional
531command, whether it requires some form of special processing, as
532LE<lt>...> does.
533
534Future versions of this specification may add additional
535formatting codes.
536
537Historical note: A few older Pod processors would not see a ">" as
538closing a "CE<lt>" code, if the ">" was immediately preceded by
539a "-". This was so that this:
540
541 C<$foo->bar>
542
543would parse as equivalent to this:
544
75f15e9f 545 C<$foo-E<gt>bar>
8a93676d
SB
546
547instead of as equivalent to a "C" formatting code containing
548only "$foo-", and then a "bar>" outside the "C" formatting code. This
549problem has since been solved by the addition of syntaxes like this:
550
551 C<< $foo->bar >>
552
553Compliant parsers must not treat "->" as special.
554
555Formatting codes absolutely cannot span paragraphs. If a code is
556opened in one paragraph, and no closing code is found by the end of
557that paragraph, the Pod parser must close that formatting code,
558and should complain (as in "Unterminated I code in the paragraph
559starting at line 123: 'Time objects are not...'"). So these
560two paragraphs:
561
562 I<I told you not to do this!
210b36aa 563
8a93676d
SB
564 Don't make me say it again!>
565
566...must I<not> be parsed as two paragraphs in italics (with the I
567code starting in one paragraph and starting in another.) Instead,
568the first paragraph should generate a warning, but that aside, the
569above code must parse as if it were:
570
571 I<I told you not to do this!>
210b36aa 572
8a93676d
SB
573 Don't make me say it again!E<gt>
574
575(In SGMLish jargon, all Pod commands are like block-level
576elements, whereas all Pod formatting codes are like inline-level
577elements.)
578
579
580
581=head1 Notes on Implementing Pod Processors
582
583The following is a long section of miscellaneous requirements
584and suggestions to do with Pod processing.
585
586=over
587
588=item *
589
590Pod formatters should tolerate lines in verbatim blocks that are of
591any length, even if that means having to break them (possibly several
592times, for very long lines) to avoid text running off the side of the
593page. Pod formatters may warn of such line-breaking. Such warnings
594are particularly appropriate for lines are over 100 characters long, which
595are usually not intentional.
596
597=item *
598
599Pod parsers must recognize I<all> of the three well-known newline
600formats: CR, LF, and CRLF. See L<perlport|perlport>.
601
602=item *
603
604Pod parsers should accept input lines that are of any length.
605
606=item *
607
608Since Perl recognizes a Unicode Byte Order Mark at the start of files
609as signaling that the file is Unicode encoded as in UTF-16 (whether
610big-endian or little-endian) or UTF-8, Pod parsers should do the
611same. Otherwise, the character encoding should be understood as
612being UTF-8 if the first highbit byte sequence in the file seems
8f226aee
DW
613valid as a UTF-8 sequence, or otherwise as CP-1252 (earlier versions of
614this specification used Latin-1 instead of CP-1252).
8a93676d
SB
615
616Future versions of this specification may specify
617how Pod can accept other encodings. Presumably treatment of other
618encodings in Pod parsing would be as in XML parsing: whatever the
619encoding declared by a particular Pod file, content is to be
620stored in memory as Unicode characters.
621
622=item *
623
624The well known Unicode Byte Order Marks are as follows: if the
625file begins with the two literal byte values 0xFE 0xFF, this is
626the BOM for big-endian UTF-16. If the file begins with the two
627literal byte value 0xFF 0xFE, this is the BOM for little-endian
df0c7995
KW
628UTF-16. On an ASCII platform, if the file begins with the three literal
629byte values
8a93676d 6300xEF 0xBB 0xBF, this is the BOM for UTF-8.
e8a0e562 631A mechanism portable to EBCDIC platforms is to:
df0c7995
KW
632
633 my $utf8_bom = "\x{FEFF}";
634 utf8::encode($utf8_bom);
8a93676d
SB
635
636=for comment
637 use bytes; print map sprintf(" 0x%02X", ord $_), split '', "\x{feff}";
638 0xEF 0xBB 0xBF
639
640=for comment
1e54db1a 641 If toke.c is modified to support UTF-32, add mention of those here.
8a93676d
SB
642
643=item *
644
df0c7995
KW
645A naive, but often sufficient heuristic on ASCII platforms, for testing
646the first highbit
8a93676d
SB
647byte-sequence in a BOM-less file (whether in code or in Pod!), to see
648whether that sequence is valid as UTF-8 (RFC 2279) is to check whether
9a5b9407 649that the first byte in the sequence is in the range 0xC2 - 0xFD
8a93676d
SB
650I<and> whether the next byte is in the range
6510x80 - 0xBF. If so, the parser may conclude that this file is in
652UTF-8, and all highbit sequences in the file should be assumed to
653be UTF-8. Otherwise the parser should treat the file as being
df0c7995
KW
654in CP-1252. (A better check, and which works on EBCDIC platforms as
655well, is to pass a copy of the sequence to
9a5b9407
KW
656L<utf8::decode()|utf8> which performs a full validity check on the
657sequence and returns TRUE if it is valid UTF-8, FALSE otherwise. This
658function is always pre-loaded, is fast because it is written in C, and
659will only get called at most once, so you don't need to avoid it out of
660performance concerns.)
661In the unlikely circumstance that the first highbit
8a93676d
SB
662sequence in a truly non-UTF-8 file happens to appear to be UTF-8, one
663can cater to our heuristic (as well as any more intelligent heuristic)
664by prefacing that line with a comment line containing a highbit
665sequence that is clearly I<not> valid as UTF-8. A line consisting
666of simply "#", an e-acute, and any non-highbit byte,
667is sufficient to establish this file's encoding.
668
669=for comment
670 If/WHEN some brave soul makes these heuristics into a generic
fae2c0fb 671 text-file class (or PerlIO layer?), we can presumably delete
8a93676d 672 mention of these icky details from this file, and can instead
fae2c0fb 673 tell people to just use appropriate class/layer.
8a93676d 674 Auto-recognition of newline sequences would be another desirable
fae2c0fb 675 feature of such a class/layer.
8a93676d
SB
676 HINT HINT HINT.
677
678=for comment
679 "The probability that a string of characters
680 in any other encoding appears as valid UTF-8 is low" - RFC2279
681
682=item *
683
8a93676d
SB
684Pod processors must treat a "=for [label] [content...]" paragraph as
685meaning the same thing as a "=begin [label]" paragraph, content, and
686an "=end [label]" paragraph. (The parser may conflate these two
687constructs, or may leave them distinct, in the expectation that the
688formatter will nevertheless treat them the same.)
689
690=item *
691
692When rendering Pod to a format that allows comments (i.e., to nearly
693any format other than plaintext), a Pod formatter must insert comment
694text identifying its name and version number, and the name and
695version numbers of any modules it might be using to process the Pod.
696Minimal examples:
697
555bd962 698 %% POD::Pod2PS v3.14159, using POD::Parser v1.92
210b36aa 699
555bd962 700 <!-- Pod::HTML v3.14159, using POD::Parser v1.92 -->
210b36aa 701
555bd962 702 {\doccomm generated by Pod::Tree::RTF 3.14159 using Pod::Tree 1.08}
210b36aa 703
555bd962 704 .\" Pod::Man version 3.14159, using POD::Parser version 1.92
8a93676d
SB
705
706Formatters may also insert additional comments, including: the
707release date of the Pod formatter program, the contact address for
708the author(s) of the formatter, the current time, the name of input
709file, the formatting options in effect, version of Perl used, etc.
710
711Formatters may also choose to note errors/warnings as comments,
712besides or instead of emitting them otherwise (as in messages to
713STDERR, or C<die>ing).
714
715=item *
716
717Pod parsers I<may> emit warnings or error messages ("Unknown E code
718EE<lt>zslig>!") to STDERR (whether through printing to STDERR, or
719C<warn>ing/C<carp>ing, or C<die>ing/C<croak>ing), but I<must> allow
720suppressing all such STDERR output, and instead allow an option for
721reporting errors/warnings
722in some other way, whether by triggering a callback, or noting errors
723in some attribute of the document object, or some similarly unobtrusive
724mechanism -- or even by appending a "Pod Errors" section to the end of
725the parsed form of the document.
726
727=item *
728
729In cases of exceptionally aberrant documents, Pod parsers may abort the
730parse. Even then, using C<die>ing/C<croak>ing is to be avoided; where
731possible, the parser library may simply close the input file
732and add text like "*** Formatting Aborted ***" to the end of the
733(partial) in-memory document.
734
735=item *
736
737In paragraphs where formatting codes (like EE<lt>...>, BE<lt>...>)
738are understood (i.e., I<not> verbatim paragraphs, but I<including>
739ordinary paragraphs, and command paragraphs that produce renderable
740text, like "=head1"), literal whitespace should generally be considered
741"insignificant", in that one literal space has the same meaning as any
742(nonzero) number of literal spaces, literal newlines, and literal tabs
743(as long as this produces no blank lines, since those would terminate
744the paragraph). Pod parsers should compact literal whitespace in each
745processed paragraph, but may provide an option for overriding this
746(since some processing tasks do not require it), or may follow
747additional special rules (for example, specially treating
748period-space-space or period-newline sequences).
749
750=item *
751
752Pod parsers should not, by default, try to coerce apostrophe (') and
753quote (") into smart quotes (little 9's, 66's, 99's, etc), nor try to
754turn backtick (`) into anything else but a single backtick character
353c6505 755(distinct from an open quote character!), nor "--" into anything but
8a93676d
SB
756two minus signs. They I<must never> do any of those things to text
757in CE<lt>...> formatting codes, and never I<ever> to text in verbatim
758paragraphs.
759
760=item *
761
762When rendering Pod to a format that has two kinds of hyphens (-), one
3e666715 763that's a non-breaking hyphen, and another that's a breakable hyphen
8a93676d
SB
764(as in "object-oriented", which can be split across lines as
765"object-", newline, "oriented"), formatters are encouraged to
3e666715 766generally translate "-" to non-breaking hyphen, but may apply
8a93676d
SB
767heuristics to convert some of these to breaking hyphens.
768
769=item *
770
771Pod formatters should make reasonable efforts to keep words of Perl
772code from being broken across lines. For example, "Foo::Bar" in some
773formatting systems is seen as eligible for being broken across lines
774as "Foo::" newline "Bar" or even "Foo::-" newline "Bar". This should
775be avoided where possible, either by disabling all line-breaking in
776mid-word, or by wrapping particular words with internal punctuation
777in "don't break this across lines" codes (which in some formats may
778not be a single code, but might be a matter of inserting non-breaking
779zero-width spaces between every pair of characters in a word.)
780
781=item *
782
783Pod parsers should, by default, expand tabs in verbatim paragraphs as
784they are processed, before passing them to the formatter or other
785processor. Parsers may also allow an option for overriding this.
786
787=item *
788
789Pod parsers should, by default, remove newlines from the end of
790ordinary and verbatim paragraphs before passing them to the
791formatter. For example, while the paragraph you're reading now
792could be considered, in Pod source, to end with (and contain)
793the newline(s) that end it, it should be processed as ending with
794(and containing) the period character that ends this sentence.
795
796=item *
797
798Pod parsers, when reporting errors, should make some effort to report
799an approximate line number ("Nested EE<lt>>'s in Paragraph #52, near
800line 633 of Thing/Foo.pm!"), instead of merely noting the paragraph
801number ("Nested EE<lt>>'s in Paragraph #52 of Thing/Foo.pm!"). Where
802this is problematic, the paragraph number should at least be
803accompanied by an excerpt from the paragraph ("Nested EE<lt>>'s in
804Paragraph #52 of Thing/Foo.pm, which begins 'Read/write accessor for
805the CE<lt>interest rate> attribute...'").
806
807=item *
808
809Pod parsers, when processing a series of verbatim paragraphs one
810after another, should consider them to be one large verbatim
811paragraph that happens to contain blank lines. I.e., these two
d1be9408 812lines, which have a blank line between them:
8a93676d
SB
813
814 use Foo;
815
816 print Foo->VERSION
817
818should be unified into one paragraph ("\tuse Foo;\n\n\tprint
819Foo->VERSION") before being passed to the formatter or other
820processor. Parsers may also allow an option for overriding this.
821
822While this might be too cumbersome to implement in event-based Pod
823parsers, it is straightforward for parsers that return parse trees.
824
825=item *
826
827Pod formatters, where feasible, are advised to avoid splitting short
828verbatim paragraphs (under twelve lines, say) across pages.
829
830=item *
831
832Pod parsers must treat a line with only spaces and/or tabs on it as a
833"blank line" such as separates paragraphs. (Some older parsers
834recognized only two adjacent newlines as a "blank line" but would not
835recognize a newline, a space, and a newline, as a blank line. This
836is noncompliant behavior.)
837
838=item *
839
840Authors of Pod formatters/processors should make every effort to
841avoid writing their own Pod parser. There are already several in
842CPAN, with a wide range of interface styles -- and one of them,
33874d2e 843Pod::Simple, comes with modern versions of Perl.
8a93676d
SB
844
845=item *
846
847Characters in Pod documents may be conveyed either as literals, or by
848number in EE<lt>n> codes, or by an equivalent mnemonic, as in
bd940430
KW
849EE<lt>eacute> which is exactly equivalent to EE<lt>233>. The numbers
850are the Latin1/Unicode values, even on EBCDIC platforms.
851
852When referring to characters by using a EE<lt>n> numeric code, numbers
853in the range 32-126 refer to those well known US-ASCII characters (also
854defined there by Unicode, with the same meaning), which all Pod
df0c7995
KW
855formatters must render faithfully. Characters whose EE<lt>E<gt> numbers
856are in the ranges 0-31 and 127-159 should not be used (neither as
857literals,
858nor as EE<lt>number> codes), except for the literal byte-sequences for
859newline (ASCII 13, ASCII 13 10, or ASCII 10), and tab (ASCII 9).
bd940430
KW
860
861Numbers in the range 160-255 refer to Latin-1 characters (also
862defined there by Unicode, with the same meaning). Numbers above
8a93676d
SB
863255 should be understood to refer to Unicode characters.
864
865=item *
866
867Be warned
868that some formatters cannot reliably render characters outside 32-126;
869and many are able to handle 32-126 and 160-255, but nothing above
870255.
871
872=item *
873
874Besides the well-known "EE<lt>lt>" and "EE<lt>gt>" codes for
875less-than and greater-than, Pod parsers must understand "EE<lt>sol>"
876for "/" (solidus, slash), and "EE<lt>verbar>" for "|" (vertical bar,
877pipe). Pod parsers should also understand "EE<lt>lchevron>" and
878"EE<lt>rchevron>" as legacy codes for characters 171 and 187, i.e.,
879"left-pointing double angle quotation mark" = "left pointing
880guillemet" and "right-pointing double angle quotation mark" = "right
881pointing guillemet". (These look like little "<<" and ">>", and they
882are now preferably expressed with the HTML/XHTML codes "EE<lt>laquo>"
883and "EE<lt>raquo>".)
884
885=item *
886
887Pod parsers should understand all "EE<lt>html>" codes as defined
888in the entity declarations in the most recent XHTML specification at
889C<www.W3.org>. Pod parsers must understand at least the entities
890that define characters in the range 160-255 (Latin-1). Pod parsers,
891when faced with some unknown "EE<lt>I<identifier>>" code,
892shouldn't simply replace it with nullstring (by default, at least),
893but may pass it through as a string consisting of the literal characters
894E, less-than, I<identifier>, greater-than. Or Pod parsers may offer the
895alternative option of processing such unknown
896"EE<lt>I<identifier>>" codes by firing an event especially
897for such codes, or by adding a special node-type to the in-memory
898document tree. Such "EE<lt>I<identifier>>" may have special meaning
899to some processors, or some processors may choose to add them to
900a special error report.
901
902=item *
903
904Pod parsers must also support the XHTML codes "EE<lt>quot>" for
905character 34 (doublequote, "), "EE<lt>amp>" for character 38
906(ampersand, &), and "EE<lt>apos>" for character 39 (apostrophe, ').
907
908=item *
909
1bca558f 910Note that in all cases of "EE<lt>whateverE<gt>", I<whatever> (whether
8a93676d 911an htmlname, or a number in any base) must consist only of
817141f8 912alphanumeric characters -- that is, I<whatever> must match
1bca558f 913C<m/\A\w+\z/>. So S<"EE<lt> 0 1 2 3 E<gt>"> is invalid, because
8a93676d
SB
914it contains spaces, which aren't alphanumeric characters. This
915presumably does not I<need> special treatment by a Pod processor;
1bca558f 916S<" 0 1 2 3 "> doesn't look like a number in any base, so it would
8a93676d 917presumably be looked up in the table of HTML-like names. Since
1bca558f 918there isn't (and cannot be) an HTML-like entity called S<" 0 1 2 3 ">,
8a93676d 919this will be treated as an error. However, Pod processors may
1bca558f 920treat S<"EE<lt> 0 1 2 3 E<gt>"> or "EE<lt>e-acute>" as I<syntactically>
8a93676d
SB
921invalid, potentially earning a different error message than the
922error message (or warning, or event) generated by a merely unknown
923(but theoretically valid) htmlname, as in "EE<lt>qacute>"
924[sic]. However, Pod parsers are not required to make this
925distinction.
926
927=item *
928
929Note that EE<lt>number> I<must not> be interpreted as simply
930"codepoint I<number> in the current/native character set". It always
931means only "the character represented by codepoint I<number> in
932Unicode." (This is identical to the semantics of &#I<number>; in XML.)
933
934This will likely require many formatters to have tables mapping from
935treatable Unicode codepoints (such as the "\xE9" for the e-acute
936character) to the escape sequences or codes necessary for conveying
937such sequences in the target output format. A converter to *roff
938would, for example know that "\xE9" (whether conveyed literally, or via
939a EE<lt>...> sequence) is to be conveyed as "e\\*'".
8939ba94 940Similarly, a program rendering Pod in a Mac OS application window, would
8a93676d 941presumably need to know that "\xE9" maps to codepoint 142 in MacRoman
8939ba94 942encoding that (at time of writing) is native for Mac OS. Such
8a93676d
SB
943Unicode2whatever mappings are presumably already widely available for
944common output formats. (Such mappings may be incomplete! Implementers
945are not expected to bend over backwards in an attempt to render
946Cherokee syllabics, Etruscan runes, Byzantine musical symbols, or any
947of the other weird things that Unicode can encode.) And
948if a Pod document uses a character not found in such a mapping, the
949formatter should consider it an unrenderable character.
950
951=item *
952
953If, surprisingly, the implementor of a Pod formatter can't find a
954satisfactory pre-existing table mapping from Unicode characters to
955escapes in the target format (e.g., a decent table of Unicode
956characters to *roff escapes), it will be necessary to build such a
957table. If you are in this circumstance, you should begin with the
958characters in the range 0x00A0 - 0x00FF, which is mostly the heavily
959used accented characters. Then proceed (as patience permits and
960fastidiousness compels) through the characters that the (X)HTML
961standards groups judged important enough to merit mnemonics
962for. These are declared in the (X)HTML specifications at the
963www.W3.org site. At time of writing (September 2001), the most recent
964entity declaration files are:
965
966 http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent
967 http://www.w3.org/TR/xhtml1/DTD/xhtml-special.ent
968 http://www.w3.org/TR/xhtml1/DTD/xhtml-symbol.ent
969
970Then you can progress through any remaining notable Unicode characters
971in the range 0x2000-0x204D (consult the character tables at
972www.unicode.org), and whatever else strikes your fancy. For example,
973in F<xhtml-symbol.ent>, there is the entry:
974
975 <!ENTITY infin "&#8734;"> <!-- infinity, U+221E ISOtech -->
976
977While the mapping "infin" to the character "\x{221E}" will (hopefully)
978have been already handled by the Pod parser, the presence of the
979character in this file means that it's reasonably important enough to
980include in a formatter's table that maps from notable Unicode characters
981to the codes necessary for rendering them. So for a Unicode-to-*roff
982mapping, for example, this would merit the entry:
983
984 "\x{221E}" => '\(in',
985
986It is eagerly hoped that in the future, increasing numbers of formats
987(and formatters) will support Unicode characters directly (as (X)HTML
988does with C<&infin;>, C<&#8734;>, or C<&#x221E;>), reducing the need
989for idiosyncratic mappings of Unicode-to-I<my_escapes>.
990
991=item *
992
353c6505 993It is up to individual Pod formatter to display good judgement when
8a93676d
SB
994confronted with an unrenderable character (which is distinct from an
995unknown EE<lt>thing> sequence that the parser couldn't resolve to
996anything, renderable or not). It is good practice to map Latin letters
997with diacritics (like "EE<lt>eacute>"/"EE<lt>233>") to the corresponding
998unaccented US-ASCII letters (like a simple character 101, "e"), but
210b36aa 999clearly this is often not feasible, and an unrenderable character may
8a93676d
SB
1000be represented as "?", or the like. In attempting a sane fallback
1001(as from EE<lt>233> to "e"), Pod formatters may use the
1002%Latin1Code_to_fallback table in L<Pod::Escapes|Pod::Escapes>, or
1003L<Text::Unidecode|Text::Unidecode>, if available.
1004
1005For example, this Pod text:
1006
1007 magic is enabled if you set C<$Currency> to 'E<euro>'.
1008
1009may be rendered as:
1010"magic is enabled if you set C<$Currency> to 'I<?>'" or as
1011"magic is enabled if you set C<$Currency> to 'B<[euro]>'", or as
1012"magic is enabled if you set C<$Currency> to '[x20AC]', etc.
1013
1014A Pod formatter may also note, in a comment or warning, a list of what
1015unrenderable characters were encountered.
1016
1017=item *
1018
1019EE<lt>...> may freely appear in any formatting code (other than
1020in another EE<lt>...> or in an ZE<lt>>). That is, "XE<lt>The
1021EE<lt>euro>1,000,000 Solution>" is valid, as is "LE<lt>The
1022EE<lt>euro>1,000,000 Solution|Million::Euros>".
1023
1024=item *
1025
3e666715 1026Some Pod formatters output to formats that implement non-breaking
8a93676d 1027spaces as an individual character (which I'll call "NBSP"), and
3e666715 1028others output to formats that implement non-breaking spaces just as
8a93676d
SB
1029spaces wrapped in a "don't break this across lines" code. Note that
1030at the level of Pod, both sorts of codes can occur: Pod can contain a
1031NBSP character (whether as a literal, or as a "EE<lt>160>" or
1032"EE<lt>nbsp>" code); and Pod can contain "SE<lt>foo
1033IE<lt>barE<gt> baz>" codes, where "mere spaces" (character 32) in
3e666715 1034such codes are taken to represent non-breaking spaces. Pod
8a93676d
SB
1035parsers should consider supporting the optional parsing of "SE<lt>foo
1036IE<lt>barE<gt> baz>" as if it were
1037"fooI<NBSP>IE<lt>barE<gt>I<NBSP>baz", and, going the other way, the
1038optional parsing of groups of words joined by NBSP's as if each group
1039were in a SE<lt>...> code, so that formatters may use the
1040representation that maps best to what the output format demands.
1041
1042=item *
1043
210b36aa 1044Some processors may find that the C<SE<lt>...E<gt>> code is easiest to
8a93676d
SB
1045implement by replacing each space in the parse tree under the content
1046of the S, with an NBSP. But note: the replacement should apply I<not> to
1047spaces in I<all> text, but I<only> to spaces in I<printable> text. (This
1048distinction may or may not be evident in the particular tree/event
1049model implemented by the Pod parser.) For example, consider this
1050unusual case:
1051
1052 S<L</Autoloaded Functions>>
1053
1054This means that the space in the middle of the visible link text must
1055not be broken across lines. In other words, it's the same as this:
1056
1057 L<"AutoloadedE<160>Functions"/Autoloaded Functions>
1058
1059However, a misapplied space-to-NBSP replacement could (wrongly)
1060produce something equivalent to this:
1061
1062 L<"AutoloadedE<160>Functions"/AutoloadedE<160>Functions>
1063
1064...which is almost definitely not going to work as a hyperlink (assuming
1065this formatter outputs a format supporting hypertext).
1066
1067Formatters may choose to just not support the S format code,
1068especially in cases where the output format simply has no NBSP
1069character/code and no code for "don't break this stuff across lines".
1070
1071=item *
1072
1073Besides the NBSP character discussed above, implementors are reminded
1074of the existence of the other "special" character in Latin-1, the
210b36aa 1075"soft hyphen" character, also known as "discretionary hyphen",
8a93676d
SB
1076i.e. C<EE<lt>173E<gt>> = C<EE<lt>0xADE<gt>> =
1077C<EE<lt>shyE<gt>>). This character expresses an optional hyphenation
1078point. That is, it normally renders as nothing, but may render as a
1079"-" if a formatter breaks the word at that point. Pod formatters
1080should, as appropriate, do one of the following: 1) render this with
1081a code with the same meaning (e.g., "\-" in RTF), 2) pass it through
1082in the expectation that the formatter understands this character as
1083such, or 3) delete it.
1084
1085For example:
1086
1087 sigE<shy>action
1088 manuE<shy>script
1089 JarkE<shy>ko HieE<shy>taE<shy>nieE<shy>mi
1090
1091These signal to a formatter that if it is to hyphenate "sigaction"
1092or "manuscript", then it should be done as
1093"sig-I<[linebreak]>action" or "manu-I<[linebreak]>script"
1094(and if it doesn't hyphenate it, then the C<EE<lt>shyE<gt>> doesn't
1095show up at all). And if it is
1096to hyphenate "Jarkko" and/or "Hietaniemi", it can do
1097so only at the points where there is a C<EE<lt>shyE<gt>> code.
1098
1099In practice, it is anticipated that this character will not be used
1100often, but formatters should either support it, or delete it.
1101
1102=item *
1103
1104If you think that you want to add a new command to Pod (like, say, a
1105"=biblio" command), consider whether you could get the same
1106effect with a for or begin/end sequence: "=for biblio ..." or "=begin
1107biblio" ... "=end biblio". Pod processors that don't understand
1108"=for biblio", etc, will simply ignore it, whereas they may complain
1109loudly if they see "=biblio".
1110
1111=item *
1112
1113Throughout this document, "Pod" has been the preferred spelling for
1114the name of the documentation format. One may also use "POD" or
da75cd15 1115"pod". For the documentation that is (typically) in the Pod
8a93676d
SB
1116format, you may use "pod", or "Pod", or "POD". Understanding these
1117distinctions is useful; but obsessing over how to spell them, usually
1118is not.
1119
1120=back
1121
1122
1123
1124
1125
1126=head1 About LE<lt>...E<gt> Codes
1127
1128As you can tell from a glance at L<perlpod|perlpod>, the LE<lt>...>
1129code is the most complex of the Pod formatting codes. The points below
1130will hopefully clarify what it means and how processors should deal
1131with it.
1132
1133=over
1134
1135=item *
1136
1137In parsing an LE<lt>...> code, Pod parsers must distinguish at least
1138four attributes:
1139
1140=over
1141
1142=item First:
1143
1bca558f 1144The link-text. If there is none, this must be C<undef>. (E.g., in
8a93676d
SB
1145"LE<lt>Perl Functions|perlfunc>", the link-text is "Perl Functions".
1146In "LE<lt>Time::HiRes>" and even "LE<lt>|Time::HiRes>", there is no
1147link text. Note that link text may contain formatting.)
1148
1149=item Second:
1150
ac036724 1151The possibly inferred link-text; i.e., if there was no real link
8a93676d
SB
1152text, then this is the text that we'll infer in its place. (E.g., for
1153"LE<lt>Getopt::Std>", the inferred link text is "Getopt::Std".)
1154
1155=item Third:
1156
1bca558f 1157The name or URL, or C<undef> if none. (E.g., in "LE<lt>Perl
ac036724 1158Functions|perlfunc>", the name (also sometimes called the page)
1bca558f 1159is "perlfunc". In "LE<lt>/CAVEATS>", the name is C<undef>.)
8a93676d
SB
1160
1161=item Fourth:
1162
1bca558f 1163The section (AKA "item" in older perlpods), or C<undef> if none. E.g.,
f41e638c 1164in "LE<lt>Getopt::Std/DESCRIPTIONE<gt>", "DESCRIPTION" is the section. (Note
8a93676d
SB
1165that this is not the same as a manpage section like the "5" in "man 5
1166crontab". "Section Foo" in the Pod sense means the part of the text
6edf2346 1167that's introduced by the heading or item whose text is "Foo".)
8a93676d
SB
1168
1169=back
1170
1171Pod parsers may also note additional attributes including:
1172
1173=over
1174
1175=item Fifth:
1176
1177A flag for whether item 3 (if present) is a URL (like
1178"http://lists.perl.org" is), in which case there should be no section
1179attribute; a Pod name (like "perldoc" and "Getopt::Std" are); or
1180possibly a man page name (like "crontab(5)" is).
1181
1182=item Sixth:
1183
1184The raw original LE<lt>...> content, before text is split on
1185"|", "/", etc, and before EE<lt>...> codes are expanded.
1186
1187=back
1188
1189(The above were numbered only for concise reference below. It is not
1190a requirement that these be passed as an actual list or array.)
1191
1192For example:
1193
1194 L<Foo::Bar>
555bd962
BG
1195 => undef, # link text
1196 "Foo::Bar", # possibly inferred link text
1197 "Foo::Bar", # name
1198 undef, # section
1199 'pod', # what sort of link
1200 "Foo::Bar" # original content
8a93676d
SB
1201
1202 L<Perlport's section on NL's|perlport/Newlines>
555bd962
BG
1203 => "Perlport's section on NL's", # link text
1204 "Perlport's section on NL's", # possibly inferred link text
1205 "perlport", # name
1206 "Newlines", # section
1207 'pod', # what sort of link
1208 "Perlport's section on NL's|perlport/Newlines"
1209 # original content
8a93676d
SB
1210
1211 L<perlport/Newlines>
555bd962
BG
1212 => undef, # link text
1213 '"Newlines" in perlport', # possibly inferred link text
1214 "perlport", # name
1215 "Newlines", # section
1216 'pod', # what sort of link
1217 "perlport/Newlines" # original content
8a93676d
SB
1218
1219 L<crontab(5)/"DESCRIPTION">
555bd962
BG
1220 => undef, # link text
1221 '"DESCRIPTION" in crontab(5)', # possibly inferred link text
1222 "crontab(5)", # name
1223 "DESCRIPTION", # section
1224 'man', # what sort of link
1225 'crontab(5)/"DESCRIPTION"' # original content
8a93676d
SB
1226
1227 L</Object Attributes>
555bd962
BG
1228 => undef, # link text
1229 '"Object Attributes"', # possibly inferred link text
1230 undef, # name
1231 "Object Attributes", # section
1232 'pod', # what sort of link
1233 "/Object Attributes" # original content
8a93676d 1234
71c89d21 1235 L<https://www.perl.org/>
555bd962 1236 => undef, # link text
a7b1b289
MM
1237 "https://www.perl.org/", # possibly inferred link text
1238 "https://www.perl.org/", # name
555bd962
BG
1239 undef, # section
1240 'url', # what sort of link
71c89d21 1241 "https://www.perl.org/" # original content
8a93676d 1242
71c89d21 1243 L<Perl.org|https://www.perl.org/>
555bd962 1244 => "Perl.org", # link text
a7b1b289
MM
1245 "https://www.perl.org/", # possibly inferred link text
1246 "https://www.perl.org/", # name
555bd962
BG
1247 undef, # section
1248 'url', # what sort of link
71c89d21 1249 "Perl.org|https://www.perl.org/" # original content
f6e963e4 1250
8a93676d
SB
1251Note that you can distinguish URL-links from anything else by the
1252fact that they match C<m/\A\w+:[^:\s]\S*\z/>. So
1253C<LE<lt>http://www.perl.comE<gt>> is a URL, but
1254C<LE<lt>HTTP::ResponseE<gt>> isn't.
1255
1256=item *
1257
1258In case of LE<lt>...> codes with no "text|" part in them,
1259older formatters have exhibited great variation in actually displaying
1260the link or cross reference. For example, LE<lt>crontab(5)> would render
1261as "the C<crontab(5)> manpage", or "in the C<crontab(5)> manpage"
1262or just "C<crontab(5)>".
1263
1264Pod processors must now treat "text|"-less links as follows:
1265
1266 L<name> => L<name|name>
1267 L</section> => L<"section"|/section>
1268 L<name/section> => L<"section" in name|name/section>
1269
1270=item *
1271
1272Note that section names might contain markup. I.e., if a section
1273starts with:
1274
1275 =head2 About the C<-M> Operator
1276
1277or with:
1278
1279 =item About the C<-M> Operator
1280
1281then a link to it would look like this:
1282
1283 L<somedoc/About the C<-M> Operator>
1284
1285Formatters may choose to ignore the markup for purposes of resolving
1286the link and use only the renderable characters in the section name,
1287as in:
1288
1289 <h1><a name="About_the_-M_Operator">About the <code>-M</code>
1290 Operator</h1>
210b36aa 1291
8a93676d 1292 ...
210b36aa 1293
8a93676d
SB
1294 <a href="somedoc#About_the_-M_Operator">About the <code>-M</code>
1295 Operator" in somedoc</a>
1296
1297=item *
1298
1299Previous versions of perlpod distinguished C<LE<lt>name/"section"E<gt>>
1300links from C<LE<lt>name/itemE<gt>> links (and their targets). These
1301have been merged syntactically and semantically in the current
1302specification, and I<section> can refer either to a "=headI<n> Heading
1303Content" command or to a "=item Item Content" command. This
1304specification does not specify what behavior should be in the case
1305of a given document having several things all seeming to produce the
1306same I<section> identifier (e.g., in HTML, several things all producing
1307the same I<anchorname> in <a name="I<anchorname>">...</a>
1308elements). Where Pod processors can control this behavior, they should
1309use the first such anchor. That is, C<LE<lt>Foo/BarE<gt>> refers to the
1310I<first> "Bar" section in Foo.
1311
1312But for some processors/formats this cannot be easily controlled; as
1313with the HTML example, the behavior of multiple ambiguous
1314<a name="I<anchorname>">...</a> is most easily just left up to
1315browsers to decide.
1316
1317=item *
1318
8a93676d
SB
1319In a C<LE<lt>text|...E<gt>> code, text may contain formatting codes
1320for formatting or for EE<lt>...> escapes, as in:
1321
1322 L<B<ummE<234>stuff>|...>
1323
1324For C<LE<lt>...E<gt>> codes without a "name|" part, only
ac036724 1325C<EE<lt>...E<gt>> and C<ZE<lt>E<gt>> codes may occur. That is,
1326authors should not use "C<LE<lt>BE<lt>Foo::BarE<gt>E<gt>>".
8a93676d
SB
1327
1328Note, however, that formatting codes and ZE<lt>>'s can occur in any
1329and all parts of an LE<lt>...> (i.e., in I<name>, I<section>, I<text>,
1330and I<url>).
1331
1332Authors must not nest LE<lt>...> codes. For example, "LE<lt>The
1333LE<lt>Foo::Bar> man page>" should be treated as an error.
1334
1335=item *
1336
1337Note that Pod authors may use formatting codes inside the "text"
1338part of "LE<lt>text|name>" (and so on for LE<lt>text|/"sec">).
1339
1340In other words, this is valid:
1341
1342 Go read L<the docs on C<$.>|perlvar/"$.">
1343
1344Some output formats that do allow rendering "LE<lt>...>" codes as
1345hypertext, might not allow the link-text to be formatted; in
1346that case, formatters will have to just ignore that formatting.
1347
1348=item *
1349
1350At time of writing, C<LE<lt>nameE<gt>> values are of two types:
1351either the name of a Pod page like C<LE<lt>Foo::BarE<gt>> (which
1352might be a real Perl module or program in an @INC / PATH
e1020413 1353directory, or a .pod file in those places); or the name of a Unix
8a93676d 1354man page, like C<LE<lt>crontab(5)E<gt>>. In theory, C<LE<lt>chmodE<gt>>
62a78fcb 1355is ambiguous between a Pod page called "chmod", or the Unix man page
8a93676d
SB
1356"chmod" (in whatever man-section). However, the presence of a string
1357in parens, as in "crontab(5)", is sufficient to signal that what
1358is being discussed is not a Pod page, and so is presumably a
e1020413 1359Unix man page. The distinction is of no importance to many
8a93676d
SB
1360Pod processors, but some processors that render to hypertext formats
1361may need to distinguish them in order to know how to render a
1362given C<LE<lt>fooE<gt>> code.
1363
1364=item *
1365
b41aadf2
RS
1366Previous versions of perlpod allowed for a C<LE<lt>sectionE<gt>> syntax (as in
1367C<LE<lt>Object AttributesE<gt>>), which was not easily distinguishable from
1368C<LE<lt>nameE<gt>> syntax and for C<LE<lt>"section"E<gt>> which was only
1369slightly less ambiguous. This syntax is no longer in the specification, and
1370has been replaced by the C<LE<lt>/sectionE<gt>> syntax (where the slash was
1371formerly optional). Pod parsers should tolerate the C<LE<lt>"section"E<gt>>
1372syntax, for a while at least. The suggested heuristic for distinguishing
1373C<LE<lt>sectionE<gt>> from C<LE<lt>nameE<gt>> is that if it contains any
1374whitespace, it's a I<section>. Pod processors should warn about this being
1375deprecated syntax.
8a93676d
SB
1376
1377=back
1378
1379=head1 About =over...=back Regions
1380
1381"=over"..."=back" regions are used for various kinds of list-like
1382structures. (I use the term "region" here simply as a collective
1383term for everything from the "=over" to the matching "=back".)
1384
1385=over
1386
1387=item *
1388
1389The non-zero numeric I<indentlevel> in "=over I<indentlevel>" ...
1390"=back" is used for giving the formatter a clue as to how many
1391"spaces" (ems, or roughly equivalent units) it should tab over,
1392although many formatters will have to convert this to an absolute
1393measurement that may not exactly match with the size of spaces (or M's)
1394in the document's base font. Other formatters may have to completely
1395ignore the number. The lack of any explicit I<indentlevel> parameter is
1396equivalent to an I<indentlevel> value of 4. Pod processors may
1397complain if I<indentlevel> is present but is not a positive number
1398matching C<m/\A(\d*\.)?\d+\z/>.
1399
1400=item *
1401
1402Authors of Pod formatters are reminded that "=over" ... "=back" may
1403map to several different constructs in your output format. For
1404example, in converting Pod to (X)HTML, it can map to any of
1405<ul>...</ul>, <ol>...</ol>, <dl>...</dl>, or
1406<blockquote>...</blockquote>. Similarly, "=item" can map to <li> or
1407<dt>.
1408
1409=item *
1410
1411Each "=over" ... "=back" region should be one of the following:
1412
1413=over
1414
1415=item *
1416
1417An "=over" ... "=back" region containing only "=item *" commands,
1418each followed by some number of ordinary/verbatim paragraphs, other
1419nested "=over" ... "=back" regions, "=for..." paragraphs, and
1420"=begin"..."=end" regions.
1421
1422(Pod processors must tolerate a bare "=item" as if it were "=item
1423*".) Whether "*" is rendered as a literal asterisk, an "o", or as
1424some kind of real bullet character, is left up to the Pod formatter,
1425and may depend on the level of nesting.
1426
1427=item *
1428
1429An "=over" ... "=back" region containing only
1430C<m/\A=item\s+\d+\.?\s*\z/> paragraphs, each one (or each group of them)
1431followed by some number of ordinary/verbatim paragraphs, other nested
1432"=over" ... "=back" regions, "=for..." paragraphs, and/or
1433"=begin"..."=end" codes. Note that the numbers must start at 1
1434in each section, and must proceed in order and without skipping
1435numbers.
1436
1437(Pod processors must tolerate lines like "=item 1" as if they were
1438"=item 1.", with the period.)
1439
1440=item *
1441
1442An "=over" ... "=back" region containing only "=item [text]"
1443commands, each one (or each group of them) followed by some number of
1444ordinary/verbatim paragraphs, other nested "=over" ... "=back"
1445regions, or "=for..." paragraphs, and "=begin"..."=end" regions.
1446
1447The "=item [text]" paragraph should not match
1448C<m/\A=item\s+\d+\.?\s*\z/> or C<m/\A=item\s+\*\s*\z/>, nor should it
1449match just C<m/\A=item\s*\z/>.
1450
1451=item *
1452
1453An "=over" ... "=back" region containing no "=item" paragraphs at
1454all, and containing only some number of
1455ordinary/verbatim paragraphs, and possibly also some nested "=over"
1456... "=back" regions, "=for..." paragraphs, and "=begin"..."=end"
1457regions. Such an itemless "=over" ... "=back" region in Pod is
1458equivalent in meaning to a "<blockquote>...</blockquote>" element in
1459HTML.
1460
1461=back
1462
1463Note that with all the above cases, you can determine which type of
1464"=over" ... "=back" you have, by examining the first (non-"=cut",
1465non-"=pod") Pod paragraph after the "=over" command.
1466
1467=item *
1468
1469Pod formatters I<must> tolerate arbitrarily large amounts of text
1470in the "=item I<text...>" paragraph. In practice, most such
1471paragraphs are short, as in:
1472
1473 =item For cutting off our trade with all parts of the world
1474
1475But they may be arbitrarily long:
1476
1477 =item For transporting us beyond seas to be tried for pretended
1478 offenses
1479
1480 =item He is at this time transporting large armies of foreign
1481 mercenaries to complete the works of death, desolation and
1482 tyranny, already begun with circumstances of cruelty and perfidy
1483 scarcely paralleled in the most barbarous ages, and totally
1484 unworthy the head of a civilized nation.
1485
1486=item *
1487
1488Pod processors should tolerate "=item *" / "=item I<number>" commands
1489with no accompanying paragraph. The middle item is an example:
1490
1491 =over
210b36aa 1492
8a93676d 1493 =item 1
210b36aa 1494
8a93676d 1495 Pick up dry cleaning.
210b36aa 1496
8a93676d 1497 =item 2
210b36aa 1498
8a93676d 1499 =item 3
210b36aa 1500
8a93676d 1501 Stop by the store. Get Abba Zabas, Stoli, and cheap lawn chairs.
210b36aa 1502
8a93676d
SB
1503 =back
1504
1505=item *
1506
1507No "=over" ... "=back" region can contain headings. Processors may
1508treat such a heading as an error.
1509
1510=item *
1511
1512Note that an "=over" ... "=back" region should have some
1513content. That is, authors should not have an empty region like this:
1514
1515 =over
210b36aa 1516
8a93676d
SB
1517 =back
1518
1519Pod processors seeing such a contentless "=over" ... "=back" region,
1520may ignore it, or may report it as an error.
1521
1522=item *
1523
1524Processors must tolerate an "=over" list that goes off the end of the
1525document (i.e., which has no matching "=back"), but they may warn
1526about such a list.
1527
1528=item *
1529
1530Authors of Pod formatters should note that this construct:
1531
1532 =item Neque
1533
1534 =item Porro
1535
1536 =item Quisquam Est
210b36aa 1537
8a93676d
SB
1538 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1539 velit, sed quia non numquam eius modi tempora incidunt ut
1540 labore et dolore magnam aliquam quaerat voluptatem.
1541
1542 =item Ut Enim
1543
1544is semantically ambiguous, in a way that makes formatting decisions
1545a bit difficult. On the one hand, it could be mention of an item
1546"Neque", mention of another item "Porro", and mention of another
1547item "Quisquam Est", with just the last one requiring the explanatory
1548paragraph "Qui dolorem ipsum quia dolor..."; and then an item
1549"Ut Enim". In that case, you'd want to format it like so:
1550
1551 Neque
210b36aa 1552
8a93676d 1553 Porro
210b36aa 1554
8a93676d
SB
1555 Quisquam Est
1556 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1557 velit, sed quia non numquam eius modi tempora incidunt ut
1558 labore et dolore magnam aliquam quaerat voluptatem.
1559
1560 Ut Enim
1561
1562But it could equally well be a discussion of three (related or equivalent)
1563items, "Neque", "Porro", and "Quisquam Est", followed by a paragraph
1564explaining them all, and then a new item "Ut Enim". In that case, you'd
1565probably want to format it like so:
1566
1567 Neque
1568 Porro
1569 Quisquam Est
1570 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1571 velit, sed quia non numquam eius modi tempora incidunt ut
1572 labore et dolore magnam aliquam quaerat voluptatem.
1573
1574 Ut Enim
1575
353c6505 1576But (for the foreseeable future), Pod does not provide any way for Pod
8a93676d
SB
1577authors to distinguish which grouping is meant by the above
1578"=item"-cluster structure. So formatters should format it like so:
1579
1580 Neque
1581
1582 Porro
1583
1584 Quisquam Est
1585
1586 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1587 velit, sed quia non numquam eius modi tempora incidunt ut
1588 labore et dolore magnam aliquam quaerat voluptatem.
1589
1590 Ut Enim
1591
210b36aa 1592That is, there should be (at least roughly) equal spacing between
8a93676d
SB
1593items as between paragraphs (although that spacing may well be less
1594than the full height of a line of text). This leaves it to the reader
1595to use (con)textual cues to figure out whether the "Qui dolorem
1596ipsum..." paragraph applies to the "Quisquam Est" item or to all three
1597items "Neque", "Porro", and "Quisquam Est". While not an ideal
1598situation, this is preferable to providing formatting cues that may
1599be actually contrary to the author's intent.
1600
1601=back
1602
1603
1604
1605=head1 About Data Paragraphs and "=begin/=end" Regions
1606
1607Data paragraphs are typically used for inlining non-Pod data that is
1608to be used (typically passed through) when rendering the document to
1609a specific format:
1610
1611 =begin rtf
210b36aa 1612
8a93676d 1613 \par{\pard\qr\sa4500{\i Printed\~\chdate\~\chtime}\par}
210b36aa 1614
8a93676d
SB
1615 =end rtf
1616
1617The exact same effect could, incidentally, be achieved with a single
1618"=for" paragraph:
1619
1620 =for rtf \par{\pard\qr\sa4500{\i Printed\~\chdate\~\chtime}\par}
1621
1622(Although that is not formally a data paragraph, it has the same
1623meaning as one, and Pod parsers may parse it as one.)
1624
1625Another example of a data paragraph:
1626
1627 =begin html
210b36aa 1628
8a93676d 1629 I like <em>PIE</em>!
210b36aa 1630
8a93676d 1631 <hr>Especially pecan pie!
210b36aa 1632
8a93676d
SB
1633 =end html
1634
1635If these were ordinary paragraphs, the Pod parser would try to
1636expand the "EE<lt>/em>" (in the first paragraph) as a formatting
1637code, just like "EE<lt>lt>" or "EE<lt>eacute>". But since this
1638is in a "=begin I<identifier>"..."=end I<identifier>" region I<and>
1639the identifier "html" doesn't begin have a ":" prefix, the contents
1640of this region are stored as data paragraphs, instead of being
1641processed as ordinary paragraphs (or if they began with a spaces
1642and/or tabs, as verbatim paragraphs).
1643
1644As a further example: At time of writing, no "biblio" identifier is
1645supported, but suppose some processor were written to recognize it as
1646a way of (say) denoting a bibliographic reference (necessarily
1647containing formatting codes in ordinary paragraphs). The fact that
1648"biblio" paragraphs were meant for ordinary processing would be
1649indicated by prefacing each "biblio" identifier with a colon:
1650
1651 =begin :biblio
1652
1653 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1654 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1655
1656 =end :biblio
1657
1658This would signal to the parser that paragraphs in this begin...end
1659region are subject to normal handling as ordinary/verbatim paragraphs
1660(while still tagged as meant only for processors that understand the
1661"biblio" identifier). The same effect could be had with:
1662
1663 =for :biblio
1664 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1665 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1666
1667The ":" on these identifiers means simply "process this stuff
1668normally, even though the result will be for some special target".
1669I suggest that parser APIs report "biblio" as the target identifier,
1670but also report that it had a ":" prefix. (And similarly, with the
1671above "html", report "html" as the target identifier, and note the
1672I<lack> of a ":" prefix.)
1673
1674Note that a "=begin I<identifier>"..."=end I<identifier>" region where
1675I<identifier> begins with a colon, I<can> contain commands. For example:
1676
1677 =begin :biblio
210b36aa 1678
8a93676d 1679 Wirth's classic is available in several editions, including:
210b36aa 1680
8a93676d
SB
1681 =for comment
1682 hm, check abebooks.com for how much used copies cost.
210b36aa 1683
8a93676d 1684 =over
210b36aa 1685
8a93676d 1686 =item
210b36aa 1687
8a93676d
SB
1688 Wirth, Niklaus. 1975. I<Algorithmen und Datenstrukturen.>
1689 Teubner, Stuttgart. [Yes, it's in German.]
210b36aa 1690
8a93676d 1691 =item
210b36aa 1692
8a93676d
SB
1693 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1694 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
210b36aa 1695
8a93676d 1696 =back
210b36aa 1697
8a93676d
SB
1698 =end :biblio
1699
1700Note, however, a "=begin I<identifier>"..."=end I<identifier>"
1701region where I<identifier> does I<not> begin with a colon, should not
1702directly contain "=head1" ... "=head4" commands, nor "=over", nor "=back",
1703nor "=item". For example, this may be considered invalid:
1704
1705 =begin somedata
210b36aa 1706
8a93676d 1707 This is a data paragraph.
210b36aa 1708
8a93676d 1709 =head1 Don't do this!
210b36aa 1710
8a93676d 1711 This is a data paragraph too.
210b36aa 1712
8a93676d
SB
1713 =end somedata
1714
1715A Pod processor may signal that the above (specifically the "=head1"
1716paragraph) is an error. Note, however, that the following should
1717I<not> be treated as an error:
1718
1719 =begin somedata
210b36aa 1720
8a93676d 1721 This is a data paragraph.
210b36aa 1722
8a93676d 1723 =cut
210b36aa 1724
8a93676d
SB
1725 # Yup, this isn't Pod anymore.
1726 sub excl { (rand() > .5) ? "hoo!" : "hah!" }
210b36aa 1727
8a93676d 1728 =pod
210b36aa 1729
8a93676d 1730 This is a data paragraph too.
210b36aa 1731
8a93676d
SB
1732 =end somedata
1733
1734And this too is valid:
1735
1736 =begin someformat
210b36aa 1737
8a93676d 1738 This is a data paragraph.
210b36aa 1739
8a93676d 1740 And this is a data paragraph.
210b36aa 1741
8a93676d 1742 =begin someotherformat
210b36aa 1743
8a93676d 1744 This is a data paragraph too.
210b36aa 1745
8a93676d 1746 And this is a data paragraph too.
210b36aa 1747
8a93676d
SB
1748 =begin :yetanotherformat
1749
1750 =head2 This is a command paragraph!
1751
1752 This is an ordinary paragraph!
210b36aa 1753
8a93676d 1754 And this is a verbatim paragraph!
210b36aa 1755
8a93676d 1756 =end :yetanotherformat
210b36aa 1757
8a93676d 1758 =end someotherformat
210b36aa 1759
8a93676d 1760 Another data paragraph!
210b36aa 1761
8a93676d
SB
1762 =end someformat
1763
1764The contents of the above "=begin :yetanotherformat" ...
1765"=end :yetanotherformat" region I<aren't> data paragraphs, because
1766the immediately containing region's identifier (":yetanotherformat")
1767begins with a colon. In practice, most regions that contain
1768data paragraphs will contain I<only> data paragraphs; however,
1769the above nesting is syntactically valid as Pod, even if it is
1770rare. However, the handlers for some formats, like "html",
1771will accept only data paragraphs, not nested regions; and they may
1772complain if they see (targeted for them) nested regions, or commands,
1773other than "=end", "=pod", and "=cut".
1774
1775Also consider this valid structure:
1776
1777 =begin :biblio
210b36aa 1778
8a93676d 1779 Wirth's classic is available in several editions, including:
210b36aa 1780
8a93676d 1781 =over
210b36aa 1782
8a93676d 1783 =item
210b36aa 1784
8a93676d
SB
1785 Wirth, Niklaus. 1975. I<Algorithmen und Datenstrukturen.>
1786 Teubner, Stuttgart. [Yes, it's in German.]
210b36aa 1787
8a93676d 1788 =item
210b36aa 1789
8a93676d
SB
1790 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1791 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1792
1793 =back
210b36aa 1794
8a93676d 1795 Buy buy buy!
210b36aa 1796
8a93676d 1797 =begin html
210b36aa 1798
8a93676d 1799 <img src='wirth_spokesmodeling_book.png'>
210b36aa 1800
8a93676d 1801 <hr>
210b36aa 1802
8a93676d 1803 =end html
210b36aa 1804
8a93676d 1805 Now now now!
210b36aa 1806
8a93676d
SB
1807 =end :biblio
1808
1809There, the "=begin html"..."=end html" region is nested inside
1810the larger "=begin :biblio"..."=end :biblio" region. Note that the
1811content of the "=begin html"..."=end html" region is data
1812paragraph(s), because the immediately containing region's identifier
1813("html") I<doesn't> begin with a colon.
1814
1815Pod parsers, when processing a series of data paragraphs one
1816after another (within a single region), should consider them to
1817be one large data paragraph that happens to contain blank lines. So
1818the content of the above "=begin html"..."=end html" I<may> be stored
1819as two data paragraphs (one consisting of
1820"<img src='wirth_spokesmodeling_book.png'>\n"
1821and another consisting of "<hr>\n"), but I<should> be stored as
1822a single data paragraph (consisting of
1823"<img src='wirth_spokesmodeling_book.png'>\n\n<hr>\n").
1824
1825Pod processors should tolerate empty
1826"=begin I<something>"..."=end I<something>" regions,
1827empty "=begin :I<something>"..."=end :I<something>" regions, and
1828contentless "=for I<something>" and "=for :I<something>"
1829paragraphs. I.e., these should be tolerated:
1830
1831 =for html
210b36aa 1832
8a93676d 1833 =begin html
210b36aa 1834
8a93676d 1835 =end html
210b36aa 1836
8a93676d 1837 =begin :biblio
210b36aa 1838
8a93676d
SB
1839 =end :biblio
1840
1841Incidentally, note that there's no easy way to express a data
1842paragraph starting with something that looks like a command. Consider:
1843
1844 =begin stuff
210b36aa 1845
8a93676d 1846 =shazbot
210b36aa 1847
8a93676d
SB
1848 =end stuff
1849
1850There, "=shazbot" will be parsed as a Pod command "shazbot", not as a data
1851paragraph "=shazbot\n". However, you can express a data paragraph consisting
1852of "=shazbot\n" using this code:
1853
1854 =for stuff =shazbot
1855
1856The situation where this is necessary, is presumably quite rare.
1857
1858Note that =end commands must match the currently open =begin command. That
1859is, they must properly nest. For example, this is valid:
1860
1861 =begin outer
210b36aa 1862
8a93676d 1863 X
210b36aa 1864
8a93676d 1865 =begin inner
210b36aa 1866
8a93676d 1867 Y
210b36aa 1868
8a93676d 1869 =end inner
210b36aa 1870
8a93676d 1871 Z
210b36aa 1872
8a93676d
SB
1873 =end outer
1874
1875while this is invalid:
1876
1877 =begin outer
210b36aa 1878
8a93676d 1879 X
210b36aa 1880
8a93676d 1881 =begin inner
210b36aa 1882
8a93676d 1883 Y
210b36aa 1884
8a93676d 1885 =end outer
210b36aa 1886
8a93676d 1887 Z
210b36aa 1888
8a93676d 1889 =end inner
210b36aa 1890
8a93676d
SB
1891This latter is improper because when the "=end outer" command is seen, the
1892currently open region has the formatname "inner", not "outer". (It just
1893happens that "outer" is the format name of a higher-up region.) This is
1894an error. Processors must by default report this as an error, and may halt
210b36aa 1895processing the document containing that error. A corollary of this is that
ac036724 1896regions cannot "overlap". That is, the latter block above does not represent
8a93676d
SB
1897a region called "outer" which contains X and Y, overlapping a region called
1898"inner" which contains Y and Z. But because it is invalid (as all
1899apparently overlapping regions would be), it doesn't represent that, or
1900anything at all.
1901
1902Similarly, this is invalid:
1903
1904 =begin thing
210b36aa 1905
8a93676d
SB
1906 =end hting
1907
1908This is an error because the region is opened by "thing", and the "=end"
1909tries to close "hting" [sic].
1910
1911This is also invalid:
1912
1913 =begin thing
210b36aa 1914
8a93676d
SB
1915 =end
1916
1917This is invalid because every "=end" command must have a formatname
1918parameter.
1919
1920=head1 SEE ALSO
1921
1922L<perlpod>, L<perlsyn/"PODs: Embedded Documentation">,
1923L<podchecker>
1924
1925=head1 AUTHOR
1926
1927Sean M. Burke
1928
1929=cut
1930
1931