5 perlpodspec - Plain Old Documentation: format specification and notes
9 This document is detailed notes on the Pod markup language. Most
10 people will only have to read L<perlpod|perlpod> to know how to write
11 in Pod, but this document may answer some incidental questions to do
12 with parsing and rendering Pod.
14 In this document, "must" / "must not", "should" /
15 "should not", and "may" have their conventional (cf. RFC 2119)
16 meanings: "X must do Y" means that if X doesn't do Y, it's against
17 this specification, and should really be fixed. "X should do Y"
18 means that it's recommended, but X may fail to do Y, if there's a
19 good reason. "X may do Y" is merely a note that X can do Y at
20 will (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
22 really I<bother> me if X did Y").
24 Notably, when I say "the parser should do Y", the
25 parser may fail to do Y, if the calling application explicitly
26 requests 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>
28 the parser to provide an option for turning off whatever
29 feature Y is (like expanding tabs in verbatim paragraphs), although
30 it implicates that such an option I<may> be provided.
32 =head1 Pod Definitions
34 Pod is embedded in files, typically Perl source files, although you
35 can write a file that's nothing but Pod.
37 A B<line> in a file consists of zero or more non-newline characters,
38 terminated by either a newline or the end of the file.
40 A B<newline sequence> is usually a platform-dependent concept, but
41 Pod 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
43 addition to any other system-specific meaning. The first CR/CRLF/LF
44 sequence in the file may be used as the basis for identifying the
45 newline sequence for parsing the rest of the file.
47 A 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.
49 A B<non-blank line> is a line containing one or more characters other
50 than space or tab (and terminated by a newline or end-of-file).
52 (I<Note:> Many older Pod parsers did not accept a line consisting of
53 spaces/tabs and then a newline as a blank line. The only lines they
54 considered blank were lines consisting of I<no characters at all>,
55 terminated by a newline.)
57 B<Whitespace> is used in this document as a blanket term for spaces,
58 tabs, and newline sequences. (By itself, this term usually refers
59 to literal whitespace. That is, sequences of whitespace characters
60 in Pod source, as opposed to "EE<lt>32>", which is a formatting
61 code that I<denotes> a whitespace character.)
63 A B<Pod parser> is a module meant for parsing Pod (regardless of
64 whether this involves calling callbacks or building a parse tree or
65 directly formatting it). A B<Pod formatter> (or B<Pod translator>)
66 is a module or program that converts Pod to some other format (HTML,
67 plaintext, TeX, PostScript, RTF). A B<Pod processor> might be a
68 formatter or translator, or might be a program that does something
69 else with the Pod (like counting words, scanning for index points,
72 Pod content is contained in B<Pod blocks>. A Pod block starts with a
73 line that matches C<m/\A=[a-zA-Z]/>, and continues up to the next line
74 that matches C<m/\A=cut/> or up to the end of the file if there is
78 The current perlsyn says:
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.
87 warn "Neither POD nor CODE!?"
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.
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.
101 Note that a parser is not expected to distinguish between something that
102 looks like pod, but is in a quoted string, such as a here document.
104 Within a Pod block, there are B<Pod paragraphs>. A Pod paragraph
105 consists of non-blank lines of text, separated by one or more blank
108 For purposes of Pod processing, there are four types of paragraphs in
115 A command paragraph (also called a "directive"). The first line of
116 this paragraph must match C<m/\A=[a-zA-Z]/>. Command paragraphs are
117 typically one line, as in:
123 But they may span several (non-blank) lines:
126 Hm, I wonder what it would look like if
127 you tried to write a BNF for Pod from this.
129 =head3 Dr. Strangelove, or: How I Learned to
130 Stop Worrying and Love the Bomb
132 I<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:
135 =head1 Did You Remember to C<use strict;>?
137 In other words, the Pod processing handler for "head1" will apply the
138 same processing to "Did You Remember to CE<lt>use strict;>?" that it
139 would to an ordinary paragraph (i.e., formatting codes like
140 "CE<lt>...>") are parsed and presumably formatted appropriately, and
141 whitespace in the form of literal spaces and/or tabs is not
146 A B<verbatim paragraph>. The first line of this paragraph must be a
147 literal space or tab, and this paragraph must not be inside a "=begin
148 I<identifier>", ... "=end I<identifier>" sequence unless
149 "I<identifier>" begins with a colon (":"). That is, if a paragraph
150 starts with a literal space or tab, but I<is> inside a
151 "=begin I<identifier>", ... "=end I<identifier>" region, then it's
152 a data paragraph, unless "I<identifier>" begins with a colon.
154 Whitespace I<is> significant in verbatim paragraphs (although, in
155 processing, tabs are probably expanded).
159 An B<ordinary paragraph>. A paragraph is an ordinary paragraph
160 if its first line matches neither C<m/\A=[a-zA-Z]/> nor
161 C<m/\A[ \t]/>, I<and> if it's not inside a "=begin I<identifier>",
162 ... "=end I<identifier>" sequence unless "I<identifier>" begins with
167 A B<data paragraph>. This is a paragraph that I<is> inside a "=begin
168 I<identifier>" ... "=end I<identifier>" sequence where
169 "I<identifier>" does I<not> begin with a literal colon (":"). In
170 some sense, a data paragraph is not part of Pod at all (i.e.,
171 effectively it's "out-of-band"), since it's not subject to most kinds
172 of Pod parsing; but it is specified here, since Pod
173 parsers need to be able to call an event for it, or store it in some
174 form in a parse tree, or at least just parse I<around> it.
178 For example: consider the following paragraphs:
180 # <- that's the 0th column
190 Here, "=head1 Foo" and "=cut" are command paragraphs because the first
191 line of each matches C<m/\A=[a-zA-Z]/>. "I<[space][space]>$foo->bar"
192 is a verbatim paragraph, because its first line starts with a literal
193 whitespace character (and there's no "=begin"..."=end" region around).
195 The "=begin I<identifier>" ... "=end I<identifier>" commands stop
196 paragraphs that they surround from being parsed as ordinary or verbatim
197 paragraphs, if I<identifier> doesn't begin with a colon. This
198 is discussed in detail in the section
199 L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
203 This section is intended to supplement and clarify the discussion in
204 L<perlpod/"Command Paragraph">. These are the currently recognized
209 =item "=head1", "=head2", "=head3", "=head4", "=head5", "=head6"
211 This command indicates that the text in the remainder of the paragraph
212 is a heading. That text may contain formatting codes. Examples:
214 =head1 Object Attributes
216 =head3 What B<Not> to Do!
218 Both C<=head5> and C<=head6> were added in 2020 and might not be
219 supported on all Pod parsers. L<Pod::Simple> 3.41 was released on October
220 2020 and supports both of these providing support for all
221 L<Pod::Simple>-based Pod parsers.
225 This command indicates that this paragraph begins a Pod block. (If we
226 are already in the middle of a Pod block, this command has no effect at
227 all.) If there is any text in this command paragraph after "=pod",
228 it must be ignored. Examples:
232 This is a plain Pod paragraph.
234 =pod This text is ignored.
238 This command indicates that this line is the end of this previously
239 started Pod block. If there is any text after "=cut" on the line, it must be
244 =cut The documentation ends here.
247 # This is the first line of program text.
248 sub foo { # This is the second.
250 It is an error to try to I<start> a Pod block with a "=cut" command. In
251 that case, the Pod processor must halt parsing of the input file, and
252 must by default emit a warning.
256 This command indicates that this is the start of a list/indent
257 region. If there is any text following the "=over", it must consist
258 of only a nonzero positive numeral. The semantics of this numeral is
259 explained in the L</"About =over...=back Regions"> section, further
260 below. Formatting codes are not expanded. Examples:
270 This command indicates that an item in a list begins here. Formatting
271 codes are processed. The semantics of the (optional) text in the
272 remainder of this paragraph are
273 explained in the L</"About =over...=back Regions"> section, further
286 =item C<< $thing->stuff(I<dodad>) >>
288 =item For transporting us beyond seas to be tried for pretended
291 =item He is at this time transporting large armies of foreign
292 mercenaries to complete the works of death, desolation and
293 tyranny, already begun with circumstances of cruelty and perfidy
294 scarcely paralleled in the most barbarous ages, and totally
295 unworthy the head of a civilized nation.
299 This command indicates that this is the end of the region begun
300 by the most recent "=over" command. It permits no text after the
303 =item "=begin formatname"
305 =item "=begin formatname parameter"
307 This marks the following paragraphs (until the matching "=end
308 formatname") as being for some special kind of processing. Unless
309 "formatname" begins with a colon, the contained non-command
310 paragraphs are data paragraphs. But if "formatname" I<does> begin
311 with a colon, then non-command paragraphs are ordinary paragraphs
312 or data paragraphs. This is discussed in detail in the section
313 L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
315 It is advised that formatnames match the regexp
316 C<m/\A:?[-a-zA-Z0-9_]+\z/>. Everything following whitespace after the
317 formatname is a parameter that may be used by the formatter when dealing
318 with this region. This parameter must not be repeated in the "=end"
319 paragraph. Implementors should anticipate future expansion in the
320 semantics and syntax of the first parameter to "=begin"/"=end"/"=for".
322 =item "=end formatname"
324 This marks the end of the region opened by the matching
325 "=begin formatname" region. If "formatname" is not the formatname
326 of the most recent open "=begin formatname" region, then this
327 is an error, and must generate an error message. This
328 is discussed in detail in the section
329 L</About Data Paragraphs and "=beginE<sol>=end" Regions>.
331 =item "=for formatname text..."
333 This is synonymous with:
341 That is, it creates a region consisting of a single paragraph; that
342 paragraph is to be treated as a normal paragraph if "formatname"
343 begins with a ":"; if "formatname" I<doesn't> begin with a colon,
344 then "text..." will constitute a data paragraph. There is no way
345 to use "=for formatname text..." to express "text..." as a verbatim
348 =item "=encoding encodingname"
350 This command, which should occur early in the document (at least
351 before any non-US-ASCII data!), declares that this document is
352 encoded in the encoding I<encodingname>, which must be
353 an encoding name that L<Encode> recognizes. (Encode's list
354 of supported encodings, in L<Encode::Supported>, is useful here.)
355 If the Pod parser cannot decode the declared encoding, it
356 should emit a warning and may abort parsing the document
359 A document having more than one "=encoding" line should be
360 considered an error. Pod processors may silently tolerate this if
361 the not-first "=encoding" lines are just duplicates of the
362 first one (e.g., if there's a "=encoding utf8" line, and later on
363 another "=encoding utf8" line). But Pod processors should complain if
364 there are contradictory "=encoding" lines in the same document
365 (e.g., if there is a "=encoding utf8" early in the document and
366 "=encoding big5" later). Pod processors that recognize BOMs
367 may also complain if they see an "=encoding" line
368 that contradicts the BOM (e.g., if a document with a UTF-16LE
369 BOM has an "=encoding shiftjis" line).
373 If a Pod processor sees any command other than the ones listed
374 above (like "=head", or "=haed1", or "=stuff", or "=cuttlefish",
375 or "=w123"), that processor must by default treat this as an
376 error. It must not process the paragraph beginning with that
377 command, must by default warn of this as an error, and may
378 abort the parse. A Pod parser may allow a way for particular
379 applications to add to the above list of known commands, and to
380 stipulate, for each additional command, whether formatting
381 codes should be processed.
383 Future versions of this specification may add additional
388 =head1 Pod Formatting Codes
390 (Note that in previous drafts of this document and of perlpod,
391 formatting codes were referred to as "interior sequences", and
392 this term may still be found in the documentation for Pod parsers,
393 and in error messages from Pod processors.)
395 There are two syntaxes for formatting codes:
401 A formatting code starts with a capital letter (just US-ASCII [A-Z])
402 followed by a "<", any number of characters, and ending with the first
403 matching ">". Examples:
405 That's what I<you> think!
407 What's C<CORE::dump()> for?
409 X<C<chmod> and C<unlink()> Under Different Operating Systems>
413 A formatting code starts with a capital letter (just US-ASCII [A-Z])
414 followed by two or more "<"'s, one or more whitespace characters,
415 any number of characters, one or more whitespace characters,
416 and ending with the first matching sequence of two or more ">"'s, where
417 the number of ">"'s equals the number of "<"'s in the opening of this
418 formatting code. Examples:
420 That's what I<< you >> think!
422 C<<< open(X, ">>thing.dat") || die $! >>>
426 With this syntax, the whitespace character(s) after the "CE<lt><<"
427 and before the ">>>" (or whatever letter) are I<not> renderable. They
428 do not signify whitespace, are merely part of the formatting codes
429 themselves. That is, these are all synonymous:
441 Finally, the multiple-angle-bracket form does I<not> alter the interpretation
442 of nested formatting codes, meaning that the following four example lines are
443 identical in meaning:
445 B<example: C<$a E<lt>=E<gt> $b>>
447 B<example: C<< $a <=> $b >>>
449 B<example: C<< $a E<lt>=E<gt> $b >>>
451 B<<< example: C<< $a E<lt>=E<gt> $b >> >>>
455 In parsing Pod, a notably tricky part is the correct parsing of
456 (potentially nested!) formatting codes. Implementors should
457 consult the code in the C<parse_text> routine in Pod::Parser as an
458 example of a correct implementation.
462 =item C<IE<lt>textE<gt>> -- italic text
464 See the brief discussion in L<perlpod/"Formatting Codes">.
466 =item C<BE<lt>textE<gt>> -- bold text
468 See the brief discussion in L<perlpod/"Formatting Codes">.
470 =item C<CE<lt>codeE<gt>> -- code text
472 See the brief discussion in L<perlpod/"Formatting Codes">.
474 =item C<FE<lt>filenameE<gt>> -- style for filenames
476 See the brief discussion in L<perlpod/"Formatting Codes">.
478 =item C<XE<lt>topic nameE<gt>> -- an index entry
480 See the brief discussion in L<perlpod/"Formatting Codes">.
482 This code is unusual in that most formatters completely discard
483 this code and its content. Other formatters will render it with
484 invisible codes that can be used in building an index of
485 the current document.
487 =item C<ZE<lt>E<gt>> -- a null (zero-effect) formatting code
489 Discussed briefly in L<perlpod/"Formatting Codes">.
491 This code is unusual in that it should have no content. That is,
492 a processor may complain if it sees C<ZE<lt>potatoesE<gt>>. Whether
493 or not it complains, the I<potatoes> text should ignored.
495 =item C<LE<lt>nameE<gt>> -- a hyperlink
497 The complicated syntaxes of this code are discussed at length in
498 L<perlpod/"Formatting Codes">, and implementation details are
499 discussed below, in L</"About LE<lt>...E<gt> Codes">. Parsing the
500 contents of LE<lt>content> is tricky. Notably, the content has to be
501 checked for whether it looks like a URL, or whether it has to be split
502 on literal "|" and/or "/" (in the right order!), and so on,
503 I<before> EE<lt>...> codes are resolved.
505 =item C<EE<lt>escapeE<gt>> -- a character escape
507 See L<perlpod/"Formatting Codes">, and several points in
508 L</Notes on Implementing Pod Processors>.
510 =item C<SE<lt>textE<gt>> -- text contains non-breaking spaces
512 This formatting code is syntactically simple, but semantically
513 complex. What it means is that each space in the printable
514 content of this code signifies a non-breaking space.
522 Both signify the monospace (c[ode] style) text consisting of
523 "$x", one space, "?", one space, ":", one space, "$z". The
524 difference is that in the latter, with the S code, those spaces
525 are not "normal" spaces, but instead are non-breaking spaces.
530 If a Pod processor sees any formatting code other than the ones
531 listed above (as in "NE<lt>...>", or "QE<lt>...>", etc.), that
532 processor must by default treat this as an error.
533 A Pod parser may allow a way for particular
534 applications to add to the above list of known formatting codes;
535 a Pod parser might even allow a way to stipulate, for each additional
536 command, whether it requires some form of special processing, as
539 Future versions of this specification may add additional
542 Historical note: A few older Pod processors would not see a ">" as
543 closing a "CE<lt>" code, if the ">" was immediately preceded by
544 a "-". This was so that this:
548 would parse as equivalent to this:
552 instead of as equivalent to a "C" formatting code containing
553 only "$foo-", and then a "bar>" outside the "C" formatting code. This
554 problem has since been solved by the addition of syntaxes like this:
558 Compliant parsers must not treat "->" as special.
560 Formatting codes absolutely cannot span paragraphs. If a code is
561 opened in one paragraph, and no closing code is found by the end of
562 that paragraph, the Pod parser must close that formatting code,
563 and should complain (as in "Unterminated I code in the paragraph
564 starting at line 123: 'Time objects are not...'"). So these
567 I<I told you not to do this!
569 Don't make me say it again!>
571 ...must I<not> be parsed as two paragraphs in italics (with the I
572 code starting in one paragraph and starting in another.) Instead,
573 the first paragraph should generate a warning, but that aside, the
574 above code must parse as if it were:
576 I<I told you not to do this!>
578 Don't make me say it again!E<gt>
580 (In SGMLish jargon, all Pod commands are like block-level
581 elements, whereas all Pod formatting codes are like inline-level
586 =head1 Notes on Implementing Pod Processors
588 The following is a long section of miscellaneous requirements
589 and suggestions to do with Pod processing.
595 Pod formatters should tolerate lines in verbatim blocks that are of
596 any length, even if that means having to break them (possibly several
597 times, for very long lines) to avoid text running off the side of the
598 page. Pod formatters may warn of such line-breaking. Such warnings
599 are particularly appropriate for lines are over 100 characters long, which
600 are usually not intentional.
604 Pod parsers must recognize I<all> of the three well-known newline
605 formats: CR, LF, and CRLF. See L<perlport|perlport>.
609 Pod parsers should accept input lines that are of any length.
613 Since Perl recognizes a Unicode Byte Order Mark at the start of files
614 as signaling that the file is Unicode encoded as in UTF-16 (whether
615 big-endian or little-endian) or UTF-8, Pod parsers should do the
616 same. Otherwise, the character encoding should be understood as
617 being UTF-8 if the first highbit byte sequence in the file seems
618 valid as a UTF-8 sequence, or otherwise as CP-1252 (earlier versions of
619 this specification used Latin-1 instead of CP-1252).
621 Future versions of this specification may specify
622 how Pod can accept other encodings. Presumably treatment of other
623 encodings in Pod parsing would be as in XML parsing: whatever the
624 encoding declared by a particular Pod file, content is to be
625 stored in memory as Unicode characters.
629 The well known Unicode Byte Order Marks are as follows: if the
630 file begins with the two literal byte values 0xFE 0xFF, this is
631 the BOM for big-endian UTF-16. If the file begins with the two
632 literal byte value 0xFF 0xFE, this is the BOM for little-endian
633 UTF-16. On an ASCII platform, if the file begins with the three literal
635 0xEF 0xBB 0xBF, this is the BOM for UTF-8.
636 A mechanism portable to EBCDIC platforms is to:
638 my $utf8_bom = "\x{FEFF}";
639 utf8::encode($utf8_bom);
642 use bytes; print map sprintf(" 0x%02X", ord $_), split '', "\x{feff}";
646 If toke.c is modified to support UTF-32, add mention of those here.
650 A naive, but often sufficient heuristic on ASCII platforms, for testing
652 byte-sequence in a BOM-less file (whether in code or in Pod!), to see
653 whether that sequence is valid as UTF-8 (RFC 2279) is to check whether
654 that the first byte in the sequence is in the range 0xC2 - 0xFD
655 I<and> whether the next byte is in the range
656 0x80 - 0xBF. If so, the parser may conclude that this file is in
657 UTF-8, and all highbit sequences in the file should be assumed to
658 be UTF-8. Otherwise the parser should treat the file as being
659 in CP-1252. (A better check, and which works on EBCDIC platforms as
660 well, is to pass a copy of the sequence to
661 L<utf8::decode()|utf8> which performs a full validity check on the
662 sequence and returns TRUE if it is valid UTF-8, FALSE otherwise. This
663 function is always pre-loaded, is fast because it is written in C, and
664 will only get called at most once, so you don't need to avoid it out of
665 performance concerns.)
666 In the unlikely circumstance that the first highbit
667 sequence in a truly non-UTF-8 file happens to appear to be UTF-8, one
668 can cater to our heuristic (as well as any more intelligent heuristic)
669 by prefacing that line with a comment line containing a highbit
670 sequence that is clearly I<not> valid as UTF-8. A line consisting
671 of simply "#", an e-acute, and any non-highbit byte,
672 is sufficient to establish this file's encoding.
675 If/WHEN some brave soul makes these heuristics into a generic
676 text-file class (or PerlIO layer?), we can presumably delete
677 mention of these icky details from this file, and can instead
678 tell people to just use appropriate class/layer.
679 Auto-recognition of newline sequences would be another desirable
680 feature of such a class/layer.
684 "The probability that a string of characters
685 in any other encoding appears as valid UTF-8 is low" - RFC2279
689 Pod processors must treat a "=for [label] [content...]" paragraph as
690 meaning the same thing as a "=begin [label]" paragraph, content, and
691 an "=end [label]" paragraph. (The parser may conflate these two
692 constructs, or may leave them distinct, in the expectation that the
693 formatter will nevertheless treat them the same.)
697 When rendering Pod to a format that allows comments (i.e., to nearly
698 any format other than plaintext), a Pod formatter must insert comment
699 text identifying its name and version number, and the name and
700 version numbers of any modules it might be using to process the Pod.
703 %% POD::Pod2PS v3.14159, using POD::Parser v1.92
705 <!-- Pod::HTML v3.14159, using POD::Parser v1.92 -->
707 {\doccomm generated by Pod::Tree::RTF 3.14159 using Pod::Tree 1.08}
709 .\" Pod::Man version 3.14159, using POD::Parser version 1.92
711 Formatters may also insert additional comments, including: the
712 release date of the Pod formatter program, the contact address for
713 the author(s) of the formatter, the current time, the name of input
714 file, the formatting options in effect, version of Perl used, etc.
716 Formatters may also choose to note errors/warnings as comments,
717 besides or instead of emitting them otherwise (as in messages to
718 STDERR, or C<die>ing).
722 Pod parsers I<may> emit warnings or error messages ("Unknown E code
723 EE<lt>zslig>!") to STDERR (whether through printing to STDERR, or
724 C<warn>ing/C<carp>ing, or C<die>ing/C<croak>ing), but I<must> allow
725 suppressing all such STDERR output, and instead allow an option for
726 reporting errors/warnings
727 in some other way, whether by triggering a callback, or noting errors
728 in some attribute of the document object, or some similarly unobtrusive
729 mechanism -- or even by appending a "Pod Errors" section to the end of
730 the parsed form of the document.
734 In cases of exceptionally aberrant documents, Pod parsers may abort the
735 parse. Even then, using C<die>ing/C<croak>ing is to be avoided; where
736 possible, the parser library may simply close the input file
737 and add text like "*** Formatting Aborted ***" to the end of the
738 (partial) in-memory document.
742 In paragraphs where formatting codes (like EE<lt>...>, BE<lt>...>)
743 are understood (i.e., I<not> verbatim paragraphs, but I<including>
744 ordinary paragraphs, and command paragraphs that produce renderable
745 text, like "=head1"), literal whitespace should generally be considered
746 "insignificant", in that one literal space has the same meaning as any
747 (nonzero) number of literal spaces, literal newlines, and literal tabs
748 (as long as this produces no blank lines, since those would terminate
749 the paragraph). Pod parsers should compact literal whitespace in each
750 processed paragraph, but may provide an option for overriding this
751 (since some processing tasks do not require it), or may follow
752 additional special rules (for example, specially treating
753 period-space-space or period-newline sequences).
757 Pod parsers should not, by default, try to coerce apostrophe (') and
758 quote (") into smart quotes (little 9's, 66's, 99's, etc), nor try to
759 turn backtick (`) into anything else but a single backtick character
760 (distinct from an open quote character!), nor "--" into anything but
761 two minus signs. They I<must never> do any of those things to text
762 in CE<lt>...> formatting codes, and never I<ever> to text in verbatim
767 When rendering Pod to a format that has two kinds of hyphens (-), one
768 that's a non-breaking hyphen, and another that's a breakable hyphen
769 (as in "object-oriented", which can be split across lines as
770 "object-", newline, "oriented"), formatters are encouraged to
771 generally translate "-" to non-breaking hyphen, but may apply
772 heuristics to convert some of these to breaking hyphens.
776 Pod formatters should make reasonable efforts to keep words of Perl
777 code from being broken across lines. For example, "Foo::Bar" in some
778 formatting systems is seen as eligible for being broken across lines
779 as "Foo::" newline "Bar" or even "Foo::-" newline "Bar". This should
780 be avoided where possible, either by disabling all line-breaking in
781 mid-word, or by wrapping particular words with internal punctuation
782 in "don't break this across lines" codes (which in some formats may
783 not be a single code, but might be a matter of inserting non-breaking
784 zero-width spaces between every pair of characters in a word.)
788 Pod parsers should, by default, expand tabs in verbatim paragraphs as
789 they are processed, before passing them to the formatter or other
790 processor. Parsers may also allow an option for overriding this.
794 Pod parsers should, by default, remove newlines from the end of
795 ordinary and verbatim paragraphs before passing them to the
796 formatter. For example, while the paragraph you're reading now
797 could be considered, in Pod source, to end with (and contain)
798 the newline(s) that end it, it should be processed as ending with
799 (and containing) the period character that ends this sentence.
803 Pod parsers, when reporting errors, should make some effort to report
804 an approximate line number ("Nested EE<lt>>'s in Paragraph #52, near
805 line 633 of Thing/Foo.pm!"), instead of merely noting the paragraph
806 number ("Nested EE<lt>>'s in Paragraph #52 of Thing/Foo.pm!"). Where
807 this is problematic, the paragraph number should at least be
808 accompanied by an excerpt from the paragraph ("Nested EE<lt>>'s in
809 Paragraph #52 of Thing/Foo.pm, which begins 'Read/write accessor for
810 the CE<lt>interest rate> attribute...'").
814 Pod parsers, when processing a series of verbatim paragraphs one
815 after another, should consider them to be one large verbatim
816 paragraph that happens to contain blank lines. I.e., these two
817 lines, which have a blank line between them:
823 should be unified into one paragraph ("\tuse Foo;\n\n\tprint
824 Foo->VERSION") before being passed to the formatter or other
825 processor. Parsers may also allow an option for overriding this.
827 While this might be too cumbersome to implement in event-based Pod
828 parsers, it is straightforward for parsers that return parse trees.
832 Pod formatters, where feasible, are advised to avoid splitting short
833 verbatim paragraphs (under twelve lines, say) across pages.
837 Pod parsers must treat a line with only spaces and/or tabs on it as a
838 "blank line" such as separates paragraphs. (Some older parsers
839 recognized only two adjacent newlines as a "blank line" but would not
840 recognize a newline, a space, and a newline, as a blank line. This
841 is noncompliant behavior.)
845 Authors of Pod formatters/processors should make every effort to
846 avoid writing their own Pod parser. There are already several in
847 CPAN, with a wide range of interface styles -- and one of them,
848 Pod::Simple, comes with modern versions of Perl.
852 Characters in Pod documents may be conveyed either as literals, or by
853 number in EE<lt>n> codes, or by an equivalent mnemonic, as in
854 EE<lt>eacute> which is exactly equivalent to EE<lt>233>. The numbers
855 are the Latin1/Unicode values, even on EBCDIC platforms.
857 When referring to characters by using a EE<lt>n> numeric code, numbers
858 in the range 32-126 refer to those well known US-ASCII characters (also
859 defined there by Unicode, with the same meaning), which all Pod
860 formatters must render faithfully. Characters whose EE<lt>E<gt> numbers
861 are in the ranges 0-31 and 127-159 should not be used (neither as
863 nor as EE<lt>number> codes), except for the literal byte-sequences for
864 newline (ASCII 13, ASCII 13 10, or ASCII 10), and tab (ASCII 9).
866 Numbers in the range 160-255 refer to Latin-1 characters (also
867 defined there by Unicode, with the same meaning). Numbers above
868 255 should be understood to refer to Unicode characters.
873 that some formatters cannot reliably render characters outside 32-126;
874 and many are able to handle 32-126 and 160-255, but nothing above
879 Besides the well-known "EE<lt>lt>" and "EE<lt>gt>" codes for
880 less-than and greater-than, Pod parsers must understand "EE<lt>sol>"
881 for "/" (solidus, slash), and "EE<lt>verbar>" for "|" (vertical bar,
882 pipe). Pod parsers should also understand "EE<lt>lchevron>" and
883 "EE<lt>rchevron>" as legacy codes for characters 171 and 187, i.e.,
884 "left-pointing double angle quotation mark" = "left pointing
885 guillemet" and "right-pointing double angle quotation mark" = "right
886 pointing guillemet". (These look like little "<<" and ">>", and they
887 are now preferably expressed with the HTML/XHTML codes "EE<lt>laquo>"
892 Pod parsers should understand all "EE<lt>html>" codes as defined
893 in the entity declarations in the most recent XHTML specification at
894 C<www.W3.org>. Pod parsers must understand at least the entities
895 that define characters in the range 160-255 (Latin-1). Pod parsers,
896 when faced with some unknown "EE<lt>I<identifier>>" code,
897 shouldn't simply replace it with nullstring (by default, at least),
898 but may pass it through as a string consisting of the literal characters
899 E, less-than, I<identifier>, greater-than. Or Pod parsers may offer the
900 alternative option of processing such unknown
901 "EE<lt>I<identifier>>" codes by firing an event especially
902 for such codes, or by adding a special node-type to the in-memory
903 document tree. Such "EE<lt>I<identifier>>" may have special meaning
904 to some processors, or some processors may choose to add them to
905 a special error report.
909 Pod parsers must also support the XHTML codes "EE<lt>quot>" for
910 character 34 (doublequote, "), "EE<lt>amp>" for character 38
911 (ampersand, &), and "EE<lt>apos>" for character 39 (apostrophe, ').
915 Note that in all cases of "EE<lt>whateverE<gt>", I<whatever> (whether
916 an htmlname, or a number in any base) must consist only of
917 alphanumeric characters -- that is, I<whatever> must match
918 C<m/\A\w+\z/>. So S<"EE<lt> 0 1 2 3 E<gt>"> is invalid, because
919 it contains spaces, which aren't alphanumeric characters. This
920 presumably does not I<need> special treatment by a Pod processor;
921 S<" 0 1 2 3 "> doesn't look like a number in any base, so it would
922 presumably be looked up in the table of HTML-like names. Since
923 there isn't (and cannot be) an HTML-like entity called S<" 0 1 2 3 ">,
924 this will be treated as an error. However, Pod processors may
925 treat S<"EE<lt> 0 1 2 3 E<gt>"> or "EE<lt>e-acute>" as I<syntactically>
926 invalid, potentially earning a different error message than the
927 error message (or warning, or event) generated by a merely unknown
928 (but theoretically valid) htmlname, as in "EE<lt>qacute>"
929 [sic]. However, Pod parsers are not required to make this
934 Note that EE<lt>number> I<must not> be interpreted as simply
935 "codepoint I<number> in the current/native character set". It always
936 means only "the character represented by codepoint I<number> in
937 Unicode." (This is identical to the semantics of &#I<number>; in XML.)
939 This will likely require many formatters to have tables mapping from
940 treatable Unicode codepoints (such as the "\xE9" for the e-acute
941 character) to the escape sequences or codes necessary for conveying
942 such sequences in the target output format. A converter to *roff
943 would, for example know that "\xE9" (whether conveyed literally, or via
944 a EE<lt>...> sequence) is to be conveyed as "e\\*'".
945 Similarly, a program rendering Pod in a Mac OS application window, would
946 presumably need to know that "\xE9" maps to codepoint 142 in MacRoman
947 encoding that (at time of writing) is native for Mac OS. Such
948 Unicode2whatever mappings are presumably already widely available for
949 common output formats. (Such mappings may be incomplete! Implementers
950 are not expected to bend over backwards in an attempt to render
951 Cherokee syllabics, Etruscan runes, Byzantine musical symbols, or any
952 of the other weird things that Unicode can encode.) And
953 if a Pod document uses a character not found in such a mapping, the
954 formatter should consider it an unrenderable character.
958 If, surprisingly, the implementor of a Pod formatter can't find a
959 satisfactory pre-existing table mapping from Unicode characters to
960 escapes in the target format (e.g., a decent table of Unicode
961 characters to *roff escapes), it will be necessary to build such a
962 table. If you are in this circumstance, you should begin with the
963 characters in the range 0x00A0 - 0x00FF, which is mostly the heavily
964 used accented characters. Then proceed (as patience permits and
965 fastidiousness compels) through the characters that the (X)HTML
966 standards groups judged important enough to merit mnemonics
967 for. These are declared in the (X)HTML specifications at the
968 www.W3.org site. At time of writing (September 2001), the most recent
969 entity declaration files are:
971 http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent
972 http://www.w3.org/TR/xhtml1/DTD/xhtml-special.ent
973 http://www.w3.org/TR/xhtml1/DTD/xhtml-symbol.ent
975 Then you can progress through any remaining notable Unicode characters
976 in the range 0x2000-0x204D (consult the character tables at
977 www.unicode.org), and whatever else strikes your fancy. For example,
978 in F<xhtml-symbol.ent>, there is the entry:
980 <!ENTITY infin "∞"> <!-- infinity, U+221E ISOtech -->
982 While the mapping "infin" to the character "\x{221E}" will (hopefully)
983 have been already handled by the Pod parser, the presence of the
984 character in this file means that it's reasonably important enough to
985 include in a formatter's table that maps from notable Unicode characters
986 to the codes necessary for rendering them. So for a Unicode-to-*roff
987 mapping, for example, this would merit the entry:
989 "\x{221E}" => '\(in',
991 It is eagerly hoped that in the future, increasing numbers of formats
992 (and formatters) will support Unicode characters directly (as (X)HTML
993 does with C<∞>, C<∞>, or C<∞>), reducing the need
994 for idiosyncratic mappings of Unicode-to-I<my_escapes>.
998 It is up to individual Pod formatter to display good judgement when
999 confronted with an unrenderable character (which is distinct from an
1000 unknown EE<lt>thing> sequence that the parser couldn't resolve to
1001 anything, renderable or not). It is good practice to map Latin letters
1002 with diacritics (like "EE<lt>eacute>"/"EE<lt>233>") to the corresponding
1003 unaccented US-ASCII letters (like a simple character 101, "e"), but
1004 clearly this is often not feasible, and an unrenderable character may
1005 be represented as "?", or the like. In attempting a sane fallback
1006 (as from EE<lt>233> to "e"), Pod formatters may use the
1007 %Latin1Code_to_fallback table in L<Pod::Escapes|Pod::Escapes>, or
1008 L<Text::Unidecode|Text::Unidecode>, if available.
1010 For example, this Pod text:
1012 magic is enabled if you set C<$Currency> to 'E<euro>'.
1015 "magic is enabled if you set C<$Currency> to 'I<?>'" or as
1016 "magic is enabled if you set C<$Currency> to 'B<[euro]>'", or as
1017 "magic is enabled if you set C<$Currency> to '[x20AC]', etc.
1019 A Pod formatter may also note, in a comment or warning, a list of what
1020 unrenderable characters were encountered.
1024 EE<lt>...> may freely appear in any formatting code (other than
1025 in another EE<lt>...> or in an ZE<lt>>). That is, "XE<lt>The
1026 EE<lt>euro>1,000,000 Solution>" is valid, as is "LE<lt>The
1027 EE<lt>euro>1,000,000 Solution|Million::Euros>".
1031 Some Pod formatters output to formats that implement non-breaking
1032 spaces as an individual character (which I'll call "NBSP"), and
1033 others output to formats that implement non-breaking spaces just as
1034 spaces wrapped in a "don't break this across lines" code. Note that
1035 at the level of Pod, both sorts of codes can occur: Pod can contain a
1036 NBSP character (whether as a literal, or as a "EE<lt>160>" or
1037 "EE<lt>nbsp>" code); and Pod can contain "SE<lt>foo
1038 IE<lt>barE<gt> baz>" codes, where "mere spaces" (character 32) in
1039 such codes are taken to represent non-breaking spaces. Pod
1040 parsers should consider supporting the optional parsing of "SE<lt>foo
1041 IE<lt>barE<gt> baz>" as if it were
1042 "fooI<NBSP>IE<lt>barE<gt>I<NBSP>baz", and, going the other way, the
1043 optional parsing of groups of words joined by NBSP's as if each group
1044 were in a SE<lt>...> code, so that formatters may use the
1045 representation that maps best to what the output format demands.
1049 Some processors may find that the C<SE<lt>...E<gt>> code is easiest to
1050 implement by replacing each space in the parse tree under the content
1051 of the S, with an NBSP. But note: the replacement should apply I<not> to
1052 spaces in I<all> text, but I<only> to spaces in I<printable> text. (This
1053 distinction may or may not be evident in the particular tree/event
1054 model implemented by the Pod parser.) For example, consider this
1057 S<L</Autoloaded Functions>>
1059 This means that the space in the middle of the visible link text must
1060 not be broken across lines. In other words, it's the same as this:
1062 L<"AutoloadedE<160>Functions"/Autoloaded Functions>
1064 However, a misapplied space-to-NBSP replacement could (wrongly)
1065 produce something equivalent to this:
1067 L<"AutoloadedE<160>Functions"/AutoloadedE<160>Functions>
1069 ...which is almost definitely not going to work as a hyperlink (assuming
1070 this formatter outputs a format supporting hypertext).
1072 Formatters may choose to just not support the S format code,
1073 especially in cases where the output format simply has no NBSP
1074 character/code and no code for "don't break this stuff across lines".
1078 Besides the NBSP character discussed above, implementors are reminded
1079 of the existence of the other "special" character in Latin-1, the
1080 "soft hyphen" character, also known as "discretionary hyphen",
1081 i.e. C<EE<lt>173E<gt>> = C<EE<lt>0xADE<gt>> =
1082 C<EE<lt>shyE<gt>>). This character expresses an optional hyphenation
1083 point. That is, it normally renders as nothing, but may render as a
1084 "-" if a formatter breaks the word at that point. Pod formatters
1085 should, as appropriate, do one of the following: 1) render this with
1086 a code with the same meaning (e.g., "\-" in RTF), 2) pass it through
1087 in the expectation that the formatter understands this character as
1088 such, or 3) delete it.
1094 JarkE<shy>ko HieE<shy>taE<shy>nieE<shy>mi
1096 These signal to a formatter that if it is to hyphenate "sigaction"
1097 or "manuscript", then it should be done as
1098 "sig-I<[linebreak]>action" or "manu-I<[linebreak]>script"
1099 (and if it doesn't hyphenate it, then the C<EE<lt>shyE<gt>> doesn't
1100 show up at all). And if it is
1101 to hyphenate "Jarkko" and/or "Hietaniemi", it can do
1102 so only at the points where there is a C<EE<lt>shyE<gt>> code.
1104 In practice, it is anticipated that this character will not be used
1105 often, but formatters should either support it, or delete it.
1109 If you think that you want to add a new command to Pod (like, say, a
1110 "=biblio" command), consider whether you could get the same
1111 effect with a for or begin/end sequence: "=for biblio ..." or "=begin
1112 biblio" ... "=end biblio". Pod processors that don't understand
1113 "=for biblio", etc, will simply ignore it, whereas they may complain
1114 loudly if they see "=biblio".
1118 Throughout this document, "Pod" has been the preferred spelling for
1119 the name of the documentation format. One may also use "POD" or
1120 "pod". For the documentation that is (typically) in the Pod
1121 format, you may use "pod", or "Pod", or "POD". Understanding these
1122 distinctions is useful; but obsessing over how to spell them, usually
1131 =head1 About LE<lt>...E<gt> Codes
1133 As you can tell from a glance at L<perlpod|perlpod>, the LE<lt>...>
1134 code is the most complex of the Pod formatting codes. The points below
1135 will hopefully clarify what it means and how processors should deal
1142 In parsing an LE<lt>...> code, Pod parsers must distinguish at least
1149 The link-text. If there is none, this must be C<undef>. (E.g., in
1150 "LE<lt>Perl Functions|perlfunc>", the link-text is "Perl Functions".
1151 In "LE<lt>Time::HiRes>" and even "LE<lt>|Time::HiRes>", there is no
1152 link text. Note that link text may contain formatting.)
1156 The possibly inferred link-text; i.e., if there was no real link
1157 text, then this is the text that we'll infer in its place. (E.g., for
1158 "LE<lt>Getopt::Std>", the inferred link text is "Getopt::Std".)
1162 The name or URL, or C<undef> if none. (E.g., in "LE<lt>Perl
1163 Functions|perlfunc>", the name (also sometimes called the page)
1164 is "perlfunc". In "LE<lt>/CAVEATS>", the name is C<undef>.)
1168 The section (AKA "item" in older perlpods), or C<undef> if none. E.g.,
1169 in "LE<lt>Getopt::Std/DESCRIPTIONE<gt>", "DESCRIPTION" is the section. (Note
1170 that this is not the same as a manpage section like the "5" in "man 5
1171 crontab". "Section Foo" in the Pod sense means the part of the text
1172 that's introduced by the heading or item whose text is "Foo".)
1176 Pod parsers may also note additional attributes including:
1182 A flag for whether item 3 (if present) is a URL (like
1183 "http://lists.perl.org" is), in which case there should be no section
1184 attribute; a Pod name (like "perldoc" and "Getopt::Std" are); or
1185 possibly a man page name (like "crontab(5)" is).
1189 The raw original LE<lt>...> content, before text is split on
1190 "|", "/", etc, and before EE<lt>...> codes are expanded.
1194 (The above were numbered only for concise reference below. It is not
1195 a requirement that these be passed as an actual list or array.)
1200 => undef, # link text
1201 "Foo::Bar", # possibly inferred link text
1204 'pod', # what sort of link
1205 "Foo::Bar" # original content
1207 L<Perlport's section on NL's|perlport/Newlines>
1208 => "Perlport's section on NL's", # link text
1209 "Perlport's section on NL's", # possibly inferred link text
1211 "Newlines", # section
1212 'pod', # what sort of link
1213 "Perlport's section on NL's|perlport/Newlines"
1216 L<perlport/Newlines>
1217 => undef, # link text
1218 '"Newlines" in perlport', # possibly inferred link text
1220 "Newlines", # section
1221 'pod', # what sort of link
1222 "perlport/Newlines" # original content
1224 L<crontab(5)/"DESCRIPTION">
1225 => undef, # link text
1226 '"DESCRIPTION" in crontab(5)', # possibly inferred link text
1227 "crontab(5)", # name
1228 "DESCRIPTION", # section
1229 'man', # what sort of link
1230 'crontab(5)/"DESCRIPTION"' # original content
1232 L</Object Attributes>
1233 => undef, # link text
1234 '"Object Attributes"', # possibly inferred link text
1236 "Object Attributes", # section
1237 'pod', # what sort of link
1238 "/Object Attributes" # original content
1240 L<https://www.perl.org/>
1241 => undef, # link text
1242 "https://www.perl.org/", # possibly inferred link text
1243 "https://www.perl.org/", # name
1245 'url', # what sort of link
1246 "https://www.perl.org/" # original content
1248 L<Perl.org|https://www.perl.org/>
1249 => "Perl.org", # link text
1250 "https://www.perl.org/", # possibly inferred link text
1251 "https://www.perl.org/", # name
1253 'url', # what sort of link
1254 "Perl.org|https://www.perl.org/" # original content
1256 Note that you can distinguish URL-links from anything else by the
1257 fact that they match C<m/\A\w+:[^:\s]\S*\z/>. So
1258 C<LE<lt>http://www.perl.comE<gt>> is a URL, but
1259 C<LE<lt>HTTP::ResponseE<gt>> isn't.
1263 In case of LE<lt>...> codes with no "text|" part in them,
1264 older formatters have exhibited great variation in actually displaying
1265 the link or cross reference. For example, LE<lt>crontab(5)> would render
1266 as "the C<crontab(5)> manpage", or "in the C<crontab(5)> manpage"
1267 or just "C<crontab(5)>".
1269 Pod processors must now treat "text|"-less links as follows:
1271 L<name> => L<name|name>
1272 L</section> => L<"section"|/section>
1273 L<name/section> => L<"section" in name|name/section>
1277 Note that section names might contain markup. I.e., if a section
1280 =head2 About the C<-M> Operator
1284 =item About the C<-M> Operator
1286 then a link to it would look like this:
1288 L<somedoc/About the C<-M> Operator>
1290 Formatters may choose to ignore the markup for purposes of resolving
1291 the link and use only the renderable characters in the section name,
1294 <h1><a name="About_the_-M_Operator">About the <code>-M</code>
1299 <a href="somedoc#About_the_-M_Operator">About the <code>-M</code>
1300 Operator" in somedoc</a>
1304 Previous versions of perlpod distinguished C<LE<lt>name/"section"E<gt>>
1305 links from C<LE<lt>name/itemE<gt>> links (and their targets). These
1306 have been merged syntactically and semantically in the current
1307 specification, and I<section> can refer either to a "=headI<n> Heading
1308 Content" command or to a "=item Item Content" command. This
1309 specification does not specify what behavior should be in the case
1310 of a given document having several things all seeming to produce the
1311 same I<section> identifier (e.g., in HTML, several things all producing
1312 the same I<anchorname> in <a name="I<anchorname>">...</a>
1313 elements). Where Pod processors can control this behavior, they should
1314 use the first such anchor. That is, C<LE<lt>Foo/BarE<gt>> refers to the
1315 I<first> "Bar" section in Foo.
1317 But for some processors/formats this cannot be easily controlled; as
1318 with the HTML example, the behavior of multiple ambiguous
1319 <a name="I<anchorname>">...</a> is most easily just left up to
1324 In a C<LE<lt>text|...E<gt>> code, text may contain formatting codes
1325 for formatting or for EE<lt>...> escapes, as in:
1327 L<B<ummE<234>stuff>|...>
1329 For C<LE<lt>...E<gt>> codes without a "name|" part, only
1330 C<EE<lt>...E<gt>> and C<ZE<lt>E<gt>> codes may occur. That is,
1331 authors should not use "C<LE<lt>BE<lt>Foo::BarE<gt>E<gt>>".
1333 Note, however, that formatting codes and ZE<lt>>'s can occur in any
1334 and all parts of an LE<lt>...> (i.e., in I<name>, I<section>, I<text>,
1337 Authors must not nest LE<lt>...> codes. For example, "LE<lt>The
1338 LE<lt>Foo::Bar> man page>" should be treated as an error.
1342 Note that Pod authors may use formatting codes inside the "text"
1343 part of "LE<lt>text|name>" (and so on for LE<lt>text|/"sec">).
1345 In other words, this is valid:
1347 Go read L<the docs on C<$.>|perlvar/"$.">
1349 Some output formats that do allow rendering "LE<lt>...>" codes as
1350 hypertext, might not allow the link-text to be formatted; in
1351 that case, formatters will have to just ignore that formatting.
1355 At time of writing, C<LE<lt>nameE<gt>> values are of two types:
1356 either the name of a Pod page like C<LE<lt>Foo::BarE<gt>> (which
1357 might be a real Perl module or program in an @INC / PATH
1358 directory, or a .pod file in those places); or the name of a Unix
1359 man page, like C<LE<lt>crontab(5)E<gt>>. In theory, C<LE<lt>chmodE<gt>>
1360 is ambiguous between a Pod page called "chmod", or the Unix man page
1361 "chmod" (in whatever man-section). However, the presence of a string
1362 in parens, as in "crontab(5)", is sufficient to signal that what
1363 is being discussed is not a Pod page, and so is presumably a
1364 Unix man page. The distinction is of no importance to many
1365 Pod processors, but some processors that render to hypertext formats
1366 may need to distinguish them in order to know how to render a
1367 given C<LE<lt>fooE<gt>> code.
1371 Previous versions of perlpod allowed for a C<LE<lt>sectionE<gt>> syntax (as in
1372 C<LE<lt>Object AttributesE<gt>>), which was not easily distinguishable from
1373 C<LE<lt>nameE<gt>> syntax and for C<LE<lt>"section"E<gt>> which was only
1374 slightly less ambiguous. This syntax is no longer in the specification, and
1375 has been replaced by the C<LE<lt>/sectionE<gt>> syntax (where the slash was
1376 formerly optional). Pod parsers should tolerate the C<LE<lt>"section"E<gt>>
1377 syntax, for a while at least. The suggested heuristic for distinguishing
1378 C<LE<lt>sectionE<gt>> from C<LE<lt>nameE<gt>> is that if it contains any
1379 whitespace, it's a I<section>. Pod processors should warn about this being
1384 =head1 About =over...=back Regions
1386 "=over"..."=back" regions are used for various kinds of list-like
1387 structures. (I use the term "region" here simply as a collective
1388 term for everything from the "=over" to the matching "=back".)
1394 The non-zero numeric I<indentlevel> in "=over I<indentlevel>" ...
1395 "=back" is used for giving the formatter a clue as to how many
1396 "spaces" (ems, or roughly equivalent units) it should tab over,
1397 although many formatters will have to convert this to an absolute
1398 measurement that may not exactly match with the size of spaces (or M's)
1399 in the document's base font. Other formatters may have to completely
1400 ignore the number. The lack of any explicit I<indentlevel> parameter is
1401 equivalent to an I<indentlevel> value of 4. Pod processors may
1402 complain if I<indentlevel> is present but is not a positive number
1403 matching C<m/\A(\d*\.)?\d+\z/>.
1407 Authors of Pod formatters are reminded that "=over" ... "=back" may
1408 map to several different constructs in your output format. For
1409 example, in converting Pod to (X)HTML, it can map to any of
1410 <ul>...</ul>, <ol>...</ol>, <dl>...</dl>, or
1411 <blockquote>...</blockquote>. Similarly, "=item" can map to <li> or
1416 Each "=over" ... "=back" region should be one of the following:
1422 An "=over" ... "=back" region containing only "=item *" commands,
1423 each followed by some number of ordinary/verbatim paragraphs, other
1424 nested "=over" ... "=back" regions, "=for..." paragraphs, and
1425 "=begin"..."=end" regions.
1427 (Pod processors must tolerate a bare "=item" as if it were "=item
1428 *".) Whether "*" is rendered as a literal asterisk, an "o", or as
1429 some kind of real bullet character, is left up to the Pod formatter,
1430 and may depend on the level of nesting.
1434 An "=over" ... "=back" region containing only
1435 C<m/\A=item\s+\d+\.?\s*\z/> paragraphs, each one (or each group of them)
1436 followed by some number of ordinary/verbatim paragraphs, other nested
1437 "=over" ... "=back" regions, "=for..." paragraphs, and/or
1438 "=begin"..."=end" codes. Note that the numbers must start at 1
1439 in each section, and must proceed in order and without skipping
1442 (Pod processors must tolerate lines like "=item 1" as if they were
1443 "=item 1.", with the period.)
1447 An "=over" ... "=back" region containing only "=item [text]"
1448 commands, each one (or each group of them) followed by some number of
1449 ordinary/verbatim paragraphs, other nested "=over" ... "=back"
1450 regions, or "=for..." paragraphs, and "=begin"..."=end" regions.
1452 The "=item [text]" paragraph should not match
1453 C<m/\A=item\s+\d+\.?\s*\z/> or C<m/\A=item\s+\*\s*\z/>, nor should it
1454 match just C<m/\A=item\s*\z/>.
1458 An "=over" ... "=back" region containing no "=item" paragraphs at
1459 all, and containing only some number of
1460 ordinary/verbatim paragraphs, and possibly also some nested "=over"
1461 ... "=back" regions, "=for..." paragraphs, and "=begin"..."=end"
1462 regions. Such an itemless "=over" ... "=back" region in Pod is
1463 equivalent in meaning to a "<blockquote>...</blockquote>" element in
1468 Note that with all the above cases, you can determine which type of
1469 "=over" ... "=back" you have, by examining the first (non-"=cut",
1470 non-"=pod") Pod paragraph after the "=over" command.
1474 Pod formatters I<must> tolerate arbitrarily large amounts of text
1475 in the "=item I<text...>" paragraph. In practice, most such
1476 paragraphs are short, as in:
1478 =item For cutting off our trade with all parts of the world
1480 But they may be arbitrarily long:
1482 =item For transporting us beyond seas to be tried for pretended
1485 =item He is at this time transporting large armies of foreign
1486 mercenaries to complete the works of death, desolation and
1487 tyranny, already begun with circumstances of cruelty and perfidy
1488 scarcely paralleled in the most barbarous ages, and totally
1489 unworthy the head of a civilized nation.
1493 Pod processors should tolerate "=item *" / "=item I<number>" commands
1494 with no accompanying paragraph. The middle item is an example:
1500 Pick up dry cleaning.
1506 Stop by the store. Get Abba Zabas, Stoli, and cheap lawn chairs.
1512 No "=over" ... "=back" region can contain headings. Processors may
1513 treat such a heading as an error.
1517 Note that an "=over" ... "=back" region should have some
1518 content. That is, authors should not have an empty region like this:
1524 Pod processors seeing such a contentless "=over" ... "=back" region,
1525 may ignore it, or may report it as an error.
1529 Processors must tolerate an "=over" list that goes off the end of the
1530 document (i.e., which has no matching "=back"), but they may warn
1535 Authors of Pod formatters should note that this construct:
1543 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1544 velit, sed quia non numquam eius modi tempora incidunt ut
1545 labore et dolore magnam aliquam quaerat voluptatem.
1549 is semantically ambiguous, in a way that makes formatting decisions
1550 a bit difficult. On the one hand, it could be mention of an item
1551 "Neque", mention of another item "Porro", and mention of another
1552 item "Quisquam Est", with just the last one requiring the explanatory
1553 paragraph "Qui dolorem ipsum quia dolor..."; and then an item
1554 "Ut Enim". In that case, you'd want to format it like so:
1561 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1562 velit, sed quia non numquam eius modi tempora incidunt ut
1563 labore et dolore magnam aliquam quaerat voluptatem.
1567 But it could equally well be a discussion of three (related or equivalent)
1568 items, "Neque", "Porro", and "Quisquam Est", followed by a paragraph
1569 explaining them all, and then a new item "Ut Enim". In that case, you'd
1570 probably want to format it like so:
1575 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1576 velit, sed quia non numquam eius modi tempora incidunt ut
1577 labore et dolore magnam aliquam quaerat voluptatem.
1581 But (for the foreseeable future), Pod does not provide any way for Pod
1582 authors to distinguish which grouping is meant by the above
1583 "=item"-cluster structure. So formatters should format it like so:
1591 Qui dolorem ipsum quia dolor sit amet, consectetur, adipisci
1592 velit, sed quia non numquam eius modi tempora incidunt ut
1593 labore et dolore magnam aliquam quaerat voluptatem.
1597 That is, there should be (at least roughly) equal spacing between
1598 items as between paragraphs (although that spacing may well be less
1599 than the full height of a line of text). This leaves it to the reader
1600 to use (con)textual cues to figure out whether the "Qui dolorem
1601 ipsum..." paragraph applies to the "Quisquam Est" item or to all three
1602 items "Neque", "Porro", and "Quisquam Est". While not an ideal
1603 situation, this is preferable to providing formatting cues that may
1604 be actually contrary to the author's intent.
1610 =head1 About Data Paragraphs and "=begin/=end" Regions
1612 Data paragraphs are typically used for inlining non-Pod data that is
1613 to be used (typically passed through) when rendering the document to
1618 \par{\pard\qr\sa4500{\i Printed\~\chdate\~\chtime}\par}
1622 The exact same effect could, incidentally, be achieved with a single
1625 =for rtf \par{\pard\qr\sa4500{\i Printed\~\chdate\~\chtime}\par}
1627 (Although that is not formally a data paragraph, it has the same
1628 meaning as one, and Pod parsers may parse it as one.)
1630 Another example of a data paragraph:
1634 I like <em>PIE</em>!
1636 <hr>Especially pecan pie!
1640 If these were ordinary paragraphs, the Pod parser would try to
1641 expand the "EE<lt>/em>" (in the first paragraph) as a formatting
1642 code, just like "EE<lt>lt>" or "EE<lt>eacute>". But since this
1643 is in a "=begin I<identifier>"..."=end I<identifier>" region I<and>
1644 the identifier "html" doesn't begin have a ":" prefix, the contents
1645 of this region are stored as data paragraphs, instead of being
1646 processed as ordinary paragraphs (or if they began with a spaces
1647 and/or tabs, as verbatim paragraphs).
1649 As a further example: At time of writing, no "biblio" identifier is
1650 supported, but suppose some processor were written to recognize it as
1651 a way of (say) denoting a bibliographic reference (necessarily
1652 containing formatting codes in ordinary paragraphs). The fact that
1653 "biblio" paragraphs were meant for ordinary processing would be
1654 indicated by prefacing each "biblio" identifier with a colon:
1658 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1659 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1663 This would signal to the parser that paragraphs in this begin...end
1664 region are subject to normal handling as ordinary/verbatim paragraphs
1665 (while still tagged as meant only for processors that understand the
1666 "biblio" identifier). The same effect could be had with:
1669 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1670 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1672 The ":" on these identifiers means simply "process this stuff
1673 normally, even though the result will be for some special target".
1674 I suggest that parser APIs report "biblio" as the target identifier,
1675 but also report that it had a ":" prefix. (And similarly, with the
1676 above "html", report "html" as the target identifier, and note the
1677 I<lack> of a ":" prefix.)
1679 Note that a "=begin I<identifier>"..."=end I<identifier>" region where
1680 I<identifier> begins with a colon, I<can> contain commands. For example:
1684 Wirth's classic is available in several editions, including:
1687 hm, check abebooks.com for how much used copies cost.
1693 Wirth, Niklaus. 1975. I<Algorithmen und Datenstrukturen.>
1694 Teubner, Stuttgart. [Yes, it's in German.]
1698 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1699 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1705 Note, however, a "=begin I<identifier>"..."=end I<identifier>"
1706 region where I<identifier> does I<not> begin with a colon, should not
1707 directly contain "=head1" ... "=head4" commands, nor "=over", nor "=back",
1708 nor "=item". For example, this may be considered invalid:
1712 This is a data paragraph.
1714 =head1 Don't do this!
1716 This is a data paragraph too.
1720 A Pod processor may signal that the above (specifically the "=head1"
1721 paragraph) is an error. Note, however, that the following should
1722 I<not> be treated as an error:
1726 This is a data paragraph.
1730 # Yup, this isn't Pod anymore.
1731 sub excl { (rand() > .5) ? "hoo!" : "hah!" }
1735 This is a data paragraph too.
1739 And this too is valid:
1743 This is a data paragraph.
1745 And this is a data paragraph.
1747 =begin someotherformat
1749 This is a data paragraph too.
1751 And this is a data paragraph too.
1753 =begin :yetanotherformat
1755 =head2 This is a command paragraph!
1757 This is an ordinary paragraph!
1759 And this is a verbatim paragraph!
1761 =end :yetanotherformat
1763 =end someotherformat
1765 Another data paragraph!
1769 The contents of the above "=begin :yetanotherformat" ...
1770 "=end :yetanotherformat" region I<aren't> data paragraphs, because
1771 the immediately containing region's identifier (":yetanotherformat")
1772 begins with a colon. In practice, most regions that contain
1773 data paragraphs will contain I<only> data paragraphs; however,
1774 the above nesting is syntactically valid as Pod, even if it is
1775 rare. However, the handlers for some formats, like "html",
1776 will accept only data paragraphs, not nested regions; and they may
1777 complain if they see (targeted for them) nested regions, or commands,
1778 other than "=end", "=pod", and "=cut".
1780 Also consider this valid structure:
1784 Wirth's classic is available in several editions, including:
1790 Wirth, Niklaus. 1975. I<Algorithmen und Datenstrukturen.>
1791 Teubner, Stuttgart. [Yes, it's in German.]
1795 Wirth, Niklaus. 1976. I<Algorithms + Data Structures =
1796 Programs.> Prentice-Hall, Englewood Cliffs, NJ.
1804 <img src='wirth_spokesmodeling_book.png'>
1814 There, the "=begin html"..."=end html" region is nested inside
1815 the larger "=begin :biblio"..."=end :biblio" region. Note that the
1816 content of the "=begin html"..."=end html" region is data
1817 paragraph(s), because the immediately containing region's identifier
1818 ("html") I<doesn't> begin with a colon.
1820 Pod parsers, when processing a series of data paragraphs one
1821 after another (within a single region), should consider them to
1822 be one large data paragraph that happens to contain blank lines. So
1823 the content of the above "=begin html"..."=end html" I<may> be stored
1824 as two data paragraphs (one consisting of
1825 "<img src='wirth_spokesmodeling_book.png'>\n"
1826 and another consisting of "<hr>\n"), but I<should> be stored as
1827 a single data paragraph (consisting of
1828 "<img src='wirth_spokesmodeling_book.png'>\n\n<hr>\n").
1830 Pod processors should tolerate empty
1831 "=begin I<something>"..."=end I<something>" regions,
1832 empty "=begin :I<something>"..."=end :I<something>" regions, and
1833 contentless "=for I<something>" and "=for :I<something>"
1834 paragraphs. I.e., these should be tolerated:
1846 Incidentally, note that there's no easy way to express a data
1847 paragraph starting with something that looks like a command. Consider:
1855 There, "=shazbot" will be parsed as a Pod command "shazbot", not as a data
1856 paragraph "=shazbot\n". However, you can express a data paragraph consisting
1857 of "=shazbot\n" using this code:
1861 The situation where this is necessary, is presumably quite rare.
1863 Note that =end commands must match the currently open =begin command. That
1864 is, they must properly nest. For example, this is valid:
1880 while this is invalid:
1896 This latter is improper because when the "=end outer" command is seen, the
1897 currently open region has the formatname "inner", not "outer". (It just
1898 happens that "outer" is the format name of a higher-up region.) This is
1899 an error. Processors must by default report this as an error, and may halt
1900 processing the document containing that error. A corollary of this is that
1901 regions cannot "overlap". That is, the latter block above does not represent
1902 a region called "outer" which contains X and Y, overlapping a region called
1903 "inner" which contains Y and Z. But because it is invalid (as all
1904 apparently overlapping regions would be), it doesn't represent that, or
1907 Similarly, this is invalid:
1913 This is an error because the region is opened by "thing", and the "=end"
1914 tries to close "hting" [sic].
1916 This is also invalid:
1922 This is invalid because every "=end" command must have a formatname
1927 L<perlpod>, L<perlsyn/"PODs: Embedded Documentation">,