Commit | Line | Data |
---|---|---|
9741dab0 | 1 | # Pod::Man -- Convert POD data to formatted *roff input. |
a3e04946 | 2 | # $Id: Man.pm,v 1.4 2000/04/26 04:03:41 eagle Exp $ |
9741dab0 | 3 | # |
c9abbd5d | 4 | # Copyright 1999, 2000 by Russ Allbery <rra@stanford.edu> |
9741dab0 GS |
5 | # |
6 | # This program is free software; you can redistribute it and/or modify it | |
7 | # under the same terms as Perl itself. | |
8 | # | |
c9abbd5d GS |
9 | # This module is intended to be a replacement for the pod2man script |
10 | # distributed with versions of Perl prior to 5.6, and attempts to match its | |
11 | # output except for some specific circumstances where other decisions seemed | |
12 | # to produce better output. It uses Pod::Parser and is designed to be easy | |
13 | # to subclass. | |
14 | # | |
15 | # Perl core hackers, please note that this module is also separately | |
16 | # maintained outside of the Perl core as part of the podlators. Please send | |
17 | # me any patches at the address above in addition to sending them to the | |
18 | # standard Perl mailing lists. | |
9741dab0 GS |
19 | |
20 | ############################################################################ | |
21 | # Modules and declarations | |
22 | ############################################################################ | |
23 | ||
24 | package Pod::Man; | |
25 | ||
26 | require 5.004; | |
27 | ||
28 | use Carp qw(carp croak); | |
29 | use Pod::Parser (); | |
30 | ||
31 | use strict; | |
32 | use subs qw(makespace); | |
33 | use vars qw(@ISA %ESCAPES $PREAMBLE $VERSION); | |
34 | ||
35 | @ISA = qw(Pod::Parser); | |
36 | ||
c9abbd5d GS |
37 | # Don't use the CVS revision as the version, since this module is also in |
38 | # Perl core and too many things could munge CVS magic revision strings. | |
39 | # This number should ideally be the same as the CVS revision in podlators, | |
40 | # however. | |
a3e04946 | 41 | $VERSION = 1.04; |
9741dab0 GS |
42 | |
43 | ||
44 | ############################################################################ | |
45 | # Preamble and *roff output tables | |
46 | ############################################################################ | |
47 | ||
48 | # The following is the static preamble which starts all *roff output we | |
49 | # generate. It's completely static except for the font to use as a | |
50 | # fixed-width font, which is designed by @CFONT@. $PREAMBLE should | |
51 | # therefore be run through s/\@CFONT\@/<font>/g before output. | |
52 | $PREAMBLE = <<'----END OF PREAMBLE----'; | |
53 | .de Sh \" Subsection heading | |
54 | .br | |
55 | .if t .Sp | |
56 | .ne 5 | |
57 | .PP | |
58 | \fB\\$1\fR | |
59 | .PP | |
60 | .. | |
61 | .de Sp \" Vertical space (when we can't use .PP) | |
62 | .if t .sp .5v | |
63 | .if n .sp | |
64 | .. | |
65 | .de Ip \" List item | |
66 | .br | |
67 | .ie \\n(.$>=3 .ne \\$3 | |
68 | .el .ne 3 | |
69 | .IP "\\$1" \\$2 | |
70 | .. | |
71 | .de Vb \" Begin verbatim text | |
72 | .ft @CFONT@ | |
73 | .nf | |
74 | .ne \\$1 | |
75 | .. | |
76 | .de Ve \" End verbatim text | |
77 | .ft R | |
78 | ||
79 | .fi | |
80 | .. | |
81 | .\" Set up some character translations and predefined strings. \*(-- will | |
82 | .\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left | |
83 | .\" double quote, and \*(R" will give a right double quote. | will give a | |
84 | .\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used | |
85 | .\" to do unbreakable dashes and therefore won't be available. \*(C` and | |
86 | .\" \*(C' expand to `' in nroff, nothing in troff, for use with C<> | |
87 | .tr \(*W-|\(bv\*(Tr | |
88 | .ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p' | |
89 | .ie n \{\ | |
90 | . ds -- \(*W- | |
91 | . ds PI pi | |
92 | . if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch | |
93 | . if (\n(.H=4u)&(1m=20u) .ds -- \(*W\h'-12u'\(*W\h'-8u'-\" diablo 12 pitch | |
94 | . ds L" "" | |
95 | . ds R" "" | |
96 | . ds C` ` | |
97 | . ds C' ' | |
98 | 'br\} | |
99 | .el\{\ | |
100 | . ds -- \|\(em\| | |
101 | . ds PI \(*p | |
102 | . ds L" `` | |
103 | . ds R" '' | |
104 | 'br\} | |
105 | .\" | |
106 | .\" If the F register is turned on, we'll generate index entries on stderr | |
107 | .\" for titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and | |
108 | .\" index entries marked with X<> in POD. Of course, you'll have to process | |
109 | .\" the output yourself in some meaningful fashion. | |
110 | .if \nF \{\ | |
111 | . de IX | |
112 | . tm Index:\\$1\t\\n%\t"\\$2" | |
f3248e50 | 113 | .. |
9741dab0 GS |
114 | . nr % 0 |
115 | . rr F | |
116 | .\} | |
117 | .\" | |
118 | .\" For nroff, turn off justification. Always turn off hyphenation; it | |
119 | .\" makes way too many mistakes in technical documents. | |
120 | .hy 0 | |
121 | .if n .na | |
122 | .\" | |
123 | .\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2). | |
124 | .\" Fear. Run. Save yourself. No user-serviceable parts. | |
125 | .bd B 3 | |
126 | . \" fudge factors for nroff and troff | |
127 | .if n \{\ | |
128 | . ds #H 0 | |
129 | . ds #V .8m | |
130 | . ds #F .3m | |
131 | . ds #[ \f1 | |
132 | . ds #] \fP | |
133 | .\} | |
134 | .if t \{\ | |
135 | . ds #H ((1u-(\\\\n(.fu%2u))*.13m) | |
136 | . ds #V .6m | |
137 | . ds #F 0 | |
138 | . ds #[ \& | |
139 | . ds #] \& | |
140 | .\} | |
141 | . \" simple accents for nroff and troff | |
142 | .if n \{\ | |
143 | . ds ' \& | |
144 | . ds ` \& | |
145 | . ds ^ \& | |
146 | . ds , \& | |
147 | . ds ~ ~ | |
148 | . ds / | |
149 | .\} | |
150 | .if t \{\ | |
151 | . ds ' \\k:\h'-(\\n(.wu*8/10-\*(#H)'\'\h"|\\n:u" | |
152 | . ds ` \\k:\h'-(\\n(.wu*8/10-\*(#H)'\`\h'|\\n:u' | |
153 | . ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'^\h'|\\n:u' | |
154 | . ds , \\k:\h'-(\\n(.wu*8/10)',\h'|\\n:u' | |
155 | . ds ~ \\k:\h'-(\\n(.wu-\*(#H-.1m)'~\h'|\\n:u' | |
156 | . ds / \\k:\h'-(\\n(.wu*8/10-\*(#H)'\z\(sl\h'|\\n:u' | |
157 | .\} | |
158 | . \" troff and (daisy-wheel) nroff accents | |
159 | .ds : \\k:\h'-(\\n(.wu*8/10-\*(#H+.1m+\*(#F)'\v'-\*(#V'\z.\h'.2m+\*(#F'.\h'|\\n:u'\v'\*(#V' | |
160 | .ds 8 \h'\*(#H'\(*b\h'-\*(#H' | |
161 | .ds o \\k:\h'-(\\n(.wu+\w'\(de'u-\*(#H)/2u'\v'-.3n'\*(#[\z\(de\v'.3n'\h'|\\n:u'\*(#] | |
162 | .ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\(hy\fP\v'.25m'\h'-\*(#H' | |
163 | .ds D- D\\k:\h'-\w'D'u'\v'-.11m'\z\(hy\v'.11m'\h'|\\n:u' | |
164 | .ds th \*(#[\v'.3m'\s+1I\s-1\v'-.3m'\h'-(\w'I'u*2/3)'\s-1o\s+1\*(#] | |
165 | .ds Th \*(#[\s+2I\s-2\h'-\w'I'u*3/5'\v'-.3m'o\v'.3m'\*(#] | |
166 | .ds ae a\h'-(\w'a'u*4/10)'e | |
167 | .ds Ae A\h'-(\w'A'u*4/10)'E | |
168 | . \" corrections for vroff | |
169 | .if v .ds ~ \\k:\h'-(\\n(.wu*9/10-\*(#H)'\s-2\u~\d\s+2\h'|\\n:u' | |
170 | .if v .ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'\v'-.4m'^\v'.4m'\h'|\\n:u' | |
171 | . \" for low resolution devices (crt and lpr) | |
172 | .if \n(.H>23 .if \n(.V>19 \ | |
173 | \{\ | |
174 | . ds : e | |
175 | . ds 8 ss | |
176 | . ds o a | |
177 | . ds d- d\h'-1'\(ga | |
178 | . ds D- D\h'-1'\(hy | |
179 | . ds th \o'bp' | |
180 | . ds Th \o'LP' | |
181 | . ds ae ae | |
182 | . ds Ae AE | |
183 | .\} | |
184 | .rm #[ #] #H #V #F C | |
185 | ----END OF PREAMBLE---- | |
186 | ||
187 | # This table is taken nearly verbatim from Tom Christiansen's pod2man. It | |
188 | # assumes that the standard preamble has already been printed, since that's | |
189 | # what defines all of the accent marks. Note that some of these are quoted | |
190 | # with double quotes since they contain embedded single quotes, so use \\ | |
191 | # uniformly for backslash for readability. | |
192 | %ESCAPES = ( | |
193 | 'amp' => '&', # ampersand | |
194 | 'lt' => '<', # left chevron, less-than | |
195 | 'gt' => '>', # right chevron, greater-than | |
196 | 'quot' => '"', # double quote | |
be3174d2 GS |
197 | 'sol' => '/', # solidus |
198 | 'verbar' => '|', # vertical bar | |
9741dab0 GS |
199 | |
200 | 'Aacute' => "A\\*'", # capital A, acute accent | |
201 | 'aacute' => "a\\*'", # small a, acute accent | |
202 | 'Acirc' => 'A\\*^', # capital A, circumflex accent | |
203 | 'acirc' => 'a\\*^', # small a, circumflex accent | |
204 | 'AElig' => '\*(AE', # capital AE diphthong (ligature) | |
205 | 'aelig' => '\*(ae', # small ae diphthong (ligature) | |
206 | 'Agrave' => "A\\*`", # capital A, grave accent | |
207 | 'agrave' => "A\\*`", # small a, grave accent | |
208 | 'Aring' => 'A\\*o', # capital A, ring | |
209 | 'aring' => 'a\\*o', # small a, ring | |
210 | 'Atilde' => 'A\\*~', # capital A, tilde | |
211 | 'atilde' => 'a\\*~', # small a, tilde | |
212 | 'Auml' => 'A\\*:', # capital A, dieresis or umlaut mark | |
213 | 'auml' => 'a\\*:', # small a, dieresis or umlaut mark | |
214 | 'Ccedil' => 'C\\*,', # capital C, cedilla | |
215 | 'ccedil' => 'c\\*,', # small c, cedilla | |
216 | 'Eacute' => "E\\*'", # capital E, acute accent | |
217 | 'eacute' => "e\\*'", # small e, acute accent | |
218 | 'Ecirc' => 'E\\*^', # capital E, circumflex accent | |
219 | 'ecirc' => 'e\\*^', # small e, circumflex accent | |
220 | 'Egrave' => 'E\\*`', # capital E, grave accent | |
221 | 'egrave' => 'e\\*`', # small e, grave accent | |
222 | 'ETH' => '\\*(D-', # capital Eth, Icelandic | |
223 | 'eth' => '\\*(d-', # small eth, Icelandic | |
224 | 'Euml' => 'E\\*:', # capital E, dieresis or umlaut mark | |
225 | 'euml' => 'e\\*:', # small e, dieresis or umlaut mark | |
226 | 'Iacute' => "I\\*'", # capital I, acute accent | |
227 | 'iacute' => "i\\*'", # small i, acute accent | |
228 | 'Icirc' => 'I\\*^', # capital I, circumflex accent | |
229 | 'icirc' => 'i\\*^', # small i, circumflex accent | |
230 | 'Igrave' => 'I\\*`', # capital I, grave accent | |
231 | 'igrave' => 'i\\*`', # small i, grave accent | |
232 | 'Iuml' => 'I\\*:', # capital I, dieresis or umlaut mark | |
233 | 'iuml' => 'i\\*:', # small i, dieresis or umlaut mark | |
234 | 'Ntilde' => 'N\*~', # capital N, tilde | |
235 | 'ntilde' => 'n\*~', # small n, tilde | |
236 | 'Oacute' => "O\\*'", # capital O, acute accent | |
237 | 'oacute' => "o\\*'", # small o, acute accent | |
238 | 'Ocirc' => 'O\\*^', # capital O, circumflex accent | |
239 | 'ocirc' => 'o\\*^', # small o, circumflex accent | |
240 | 'Ograve' => 'O\\*`', # capital O, grave accent | |
241 | 'ograve' => 'o\\*`', # small o, grave accent | |
242 | 'Oslash' => 'O\\*/', # capital O, slash | |
243 | 'oslash' => 'o\\*/', # small o, slash | |
244 | 'Otilde' => 'O\\*~', # capital O, tilde | |
245 | 'otilde' => 'o\\*~', # small o, tilde | |
246 | 'Ouml' => 'O\\*:', # capital O, dieresis or umlaut mark | |
247 | 'ouml' => 'o\\*:', # small o, dieresis or umlaut mark | |
248 | 'szlig' => '\*8', # small sharp s, German (sz ligature) | |
249 | 'THORN' => '\\*(Th', # capital THORN, Icelandic | |
250 | 'thorn' => '\\*(th', # small thorn, Icelandic | |
251 | 'Uacute' => "U\\*'", # capital U, acute accent | |
252 | 'uacute' => "u\\*'", # small u, acute accent | |
253 | 'Ucirc' => 'U\\*^', # capital U, circumflex accent | |
254 | 'ucirc' => 'u\\*^', # small u, circumflex accent | |
255 | 'Ugrave' => 'U\\*`', # capital U, grave accent | |
256 | 'ugrave' => 'u\\*`', # small u, grave accent | |
257 | 'Uuml' => 'U\\*:', # capital U, dieresis or umlaut mark | |
258 | 'uuml' => 'u\\*:', # small u, dieresis or umlaut mark | |
259 | 'Yacute' => "Y\\*'", # capital Y, acute accent | |
260 | 'yacute' => "y\\*'", # small y, acute accent | |
261 | 'yuml' => 'y\\*:', # small y, dieresis or umlaut mark | |
262 | ); | |
263 | ||
264 | ||
265 | ############################################################################ | |
266 | # Static helper functions | |
267 | ############################################################################ | |
268 | ||
44464a02 GS |
269 | # Protect leading quotes and periods against interpretation as commands. |
270 | # Also protect anything starting with a backslash, since it could expand | |
271 | # or hide something that *roff would interpret as a command. This is | |
272 | # overkill, but it's much simpler than trying to parse *roff here. | |
c9abbd5d GS |
273 | sub protect { |
274 | local $_ = shift; | |
44464a02 | 275 | s/^([.\'\\])/\\&$1/mg; |
c9abbd5d GS |
276 | $_; |
277 | } | |
9741dab0 GS |
278 | |
279 | # Given a command and a single argument that may or may not contain double | |
280 | # quotes, handle double-quote formatting for it. If there are no double | |
281 | # quotes, just return the command followed by the argument in double quotes. | |
282 | # If there are double quotes, use an if statement to test for nroff, and for | |
283 | # nroff output the command followed by the argument in double quotes with | |
284 | # embedded double quotes doubled. For other formatters, remap paired double | |
285 | # quotes to `` and ''. | |
286 | sub switchquotes { | |
287 | my $command = shift; | |
288 | local $_ = shift; | |
289 | my $extra = shift; | |
290 | s/\\\*\([LR]\"/\"/g; | |
291 | if (/\"/) { | |
292 | s/\"/\"\"/g; | |
293 | my $troff = $_; | |
294 | $troff =~ s/\"\"([^\"]*)\"\"/\`\`$1\'\'/g; | |
295 | s/\"/\"\"/g if $extra; | |
296 | $troff =~ s/\"/\"\"/g if $extra; | |
297 | $_ = qq("$_") . ($extra ? " $extra" : ''); | |
298 | $troff = qq("$troff") . ($extra ? " $extra" : ''); | |
299 | return ".if n $command $_\n.el $command $troff\n"; | |
300 | } else { | |
301 | $_ = qq("$_") . ($extra ? " $extra" : ''); | |
302 | return "$command $_\n"; | |
303 | } | |
304 | } | |
305 | ||
306 | # Translate a font string into an escape. | |
307 | sub toescape { (length ($_[0]) > 1 ? '\f(' : '\f') . $_[0] } | |
308 | ||
309 | ||
310 | ############################################################################ | |
311 | # Initialization | |
312 | ############################################################################ | |
313 | ||
314 | # Initialize the object. Here, we also process any additional options | |
315 | # passed to the constructor or set up defaults if none were given. center | |
316 | # is the centered title, release is the version number, and date is the date | |
317 | # for the documentation. Note that we can't know what file name we're | |
318 | # processing due to the architecture of Pod::Parser, so that *has* to either | |
319 | # be passed to the constructor or set separately with Pod::Man::name(). | |
320 | sub initialize { | |
321 | my $self = shift; | |
322 | ||
323 | # Figure out the fixed-width font. If user-supplied, make sure that | |
324 | # they are the right length. | |
325 | for (qw/fixed fixedbold fixeditalic fixedbolditalic/) { | |
326 | if (defined $$self{$_}) { | |
327 | if (length ($$self{$_}) < 1 || length ($$self{$_}) > 2) { | |
328 | croak "roff font should be 1 or 2 chars, not `$$self{$_}'"; | |
329 | } | |
330 | } else { | |
331 | $$self{$_} = ''; | |
332 | } | |
333 | } | |
334 | ||
335 | # Set the default fonts. We can't be sure what fixed bold-italic is | |
336 | # going to be called, so default to just bold. | |
337 | $$self{fixed} ||= 'CW'; | |
338 | $$self{fixedbold} ||= 'CB'; | |
339 | $$self{fixeditalic} ||= 'CI'; | |
340 | $$self{fixedbolditalic} ||= 'CB'; | |
341 | ||
342 | # Set up a table of font escapes. First number is fixed-width, second | |
343 | # is bold, third is italic. | |
344 | $$self{FONTS} = { '000' => '\fR', '001' => '\fI', | |
345 | '010' => '\fB', '011' => '\f(BI', | |
346 | '100' => toescape ($$self{fixed}), | |
347 | '101' => toescape ($$self{fixeditalic}), | |
348 | '110' => toescape ($$self{fixedbold}), | |
349 | '111' => toescape ($$self{fixedbolditalic})}; | |
350 | ||
351 | # Extra stuff for page titles. | |
352 | $$self{center} = 'User Contributed Perl Documentation' | |
353 | unless defined $$self{center}; | |
354 | $$self{indent} = 4 unless defined $$self{indent}; | |
355 | ||
356 | # We used to try first to get the version number from a local binary, | |
357 | # but we shouldn't need that any more. Get the version from the running | |
c9abbd5d GS |
358 | # Perl. Work a little magic to handle subversions correctly under both |
359 | # the pre-5.6 and the post-5.6 version numbering schemes. | |
9741dab0 | 360 | if (!defined $$self{release}) { |
c9abbd5d GS |
361 | my @version = ($] =~ /^(\d+)\.(\d{3})(\d{0,3})$/); |
362 | $version[2] ||= 0; | |
363 | $version[2] *= 10 ** (3 - length $version[2]); | |
364 | for (@version) { $_ += 0 } | |
365 | $$self{release} = 'perl v' . join ('.', @version); | |
9741dab0 GS |
366 | } |
367 | ||
368 | # Double quotes in things that will be quoted. | |
c9abbd5d GS |
369 | for (qw/center date release/) { |
370 | $$self{$_} =~ s/\"/\"\"/g if $$self{$_}; | |
371 | } | |
9741dab0 GS |
372 | |
373 | $$self{INDENT} = 0; # Current indentation level. | |
374 | $$self{INDENTS} = []; # Stack of indentations. | |
375 | $$self{INDEX} = []; # Index keys waiting to be printed. | |
376 | ||
377 | $self->SUPER::initialize; | |
378 | } | |
379 | ||
380 | # For each document we process, output the preamble first. Note that the | |
381 | # fixed width font is a global default; once we interpolate it into the | |
382 | # PREAMBLE, it ain't ever changing. Maybe fix this later. | |
383 | sub begin_pod { | |
384 | my $self = shift; | |
385 | ||
386 | # Try to figure out the name and section from the file name. | |
387 | my $section = $$self{section} || 1; | |
388 | my $name = $$self{name}; | |
389 | if (!defined $name) { | |
390 | $name = $self->input_file; | |
fe6f1558 GS |
391 | $section = 3 if (!$$self{section} && $name =~ /\.pm\z/i); |
392 | $name =~ s/\.p(od|[lm])\z//i; | |
9741dab0 GS |
393 | if ($section =~ /^1/) { |
394 | require File::Basename; | |
395 | $name = uc File::Basename::basename ($name); | |
396 | } else { | |
397 | # Lose everything up to the first of | |
398 | # */lib/*perl* standard or site_perl module | |
399 | # */*perl*/lib from -D prefix=/opt/perl | |
400 | # */*perl*/ random module hierarchy | |
c88ded47 GS |
401 | # which works. Should be fixed to use File::Spec. Also handle |
402 | # a leading lib/ since that's what ExtUtils::MakeMaker creates. | |
9741dab0 GS |
403 | for ($name) { |
404 | s%//+%/%g; | |
c9abbd5d GS |
405 | if ( s%^.*?/lib/[^/]*perl[^/]*/%%si |
406 | or s%^.*?/[^/]*perl[^/]*/(?:lib/)?%%si) { | |
407 | s%^site(_perl)?/%%s; # site and site_perl | |
408 | s%^(.*-$^O|$^O-.*)/%%so; # arch | |
409 | s%^\d+\.\d+%%s; # version | |
9741dab0 | 410 | } |
c88ded47 | 411 | s%^lib/%%; |
9741dab0 GS |
412 | s%/%::%g; |
413 | } | |
414 | } | |
415 | } | |
416 | ||
417 | # Modification date header. Try to use the modification time of our | |
418 | # input. | |
419 | if (!defined $$self{date}) { | |
420 | my $time = (stat $self->input_file)[9] || time; | |
421 | my ($day, $month, $year) = (localtime $time)[3,4,5]; | |
422 | $month++; | |
423 | $year += 1900; | |
c9abbd5d | 424 | $$self{date} = sprintf ('%4d-%02d-%02d', $year, $month, $day); |
9741dab0 GS |
425 | } |
426 | ||
427 | # Now, print out the preamble and the title. | |
428 | $PREAMBLE =~ s/\@CFONT\@/$$self{fixed}/; | |
429 | chomp $PREAMBLE; | |
430 | print { $self->output_handle } <<"----END OF HEADER----"; | |
431 | .\\" Automatically generated by Pod::Man version $VERSION | |
432 | .\\" @{[ scalar localtime ]} | |
433 | .\\" | |
434 | .\\" Standard preamble: | |
435 | .\\" ====================================================================== | |
436 | $PREAMBLE | |
437 | .\\" ====================================================================== | |
438 | .\\" | |
439 | .IX Title "$name $section" | |
440 | .TH $name $section "$$self{release}" "$$self{date}" "$$self{center}" | |
441 | .UC | |
442 | ----END OF HEADER---- | |
443 | #"# for cperl-mode | |
444 | ||
445 | # Initialize a few per-file variables. | |
446 | $$self{INDENT} = 0; | |
447 | $$self{NEEDSPACE} = 0; | |
448 | } | |
449 | ||
450 | ||
451 | ############################################################################ | |
452 | # Core overrides | |
453 | ############################################################################ | |
454 | ||
455 | # Called for each command paragraph. Gets the command, the associated | |
456 | # paragraph, the line number, and a Pod::Paragraph object. Just dispatches | |
457 | # the command to a method named the same as the command. =cut is handled | |
458 | # internally by Pod::Parser. | |
459 | sub command { | |
460 | my $self = shift; | |
461 | my $command = shift; | |
462 | return if $command eq 'pod'; | |
463 | return if ($$self{EXCLUDE} && $command ne 'end'); | |
464 | $command = 'cmd_' . $command; | |
844b31e3 A |
465 | unless ($self -> can ($command)) { |
466 | my $com = substr $command => 4; | |
467 | my ($file, $line) = $_ [2] -> file_line; | |
468 | (my $text = $_ [0]) =~ s/\n+\z//g; | |
469 | $text = " $text" if $text =~ /^\S/; | |
470 | warn qq {$file: Unknown command paragraph "=$com${text}"}, | |
471 | qq { on line $line.\n}; | |
472 | return; | |
473 | } | |
474 | else { | |
475 | $self->$command (@_); | |
476 | } | |
9741dab0 GS |
477 | } |
478 | ||
479 | # Called for a verbatim paragraph. Gets the paragraph, the line number, and | |
480 | # a Pod::Paragraph object. Rofficate backslashes, untabify, put a | |
481 | # zero-width character at the beginning of each line to protect against | |
482 | # commands, and wrap in .Vb/.Ve. | |
483 | sub verbatim { | |
484 | my $self = shift; | |
485 | return if $$self{EXCLUDE}; | |
486 | local $_ = shift; | |
487 | return if /^\s+$/; | |
488 | s/\s+$/\n/; | |
489 | my $lines = tr/\n/\n/; | |
490 | 1 while s/^(.*?)(\t+)/$1 . ' ' x (length ($2) * 8 - length ($1) % 8)/me; | |
491 | s/\\/\\e/g; | |
492 | s/^(\s*\S)/'\&' . $1/gme; | |
493 | $self->makespace if $$self{NEEDSPACE}; | |
494 | $self->output (".Vb $lines\n$_.Ve\n"); | |
495 | $$self{NEEDSPACE} = 0; | |
496 | } | |
497 | ||
498 | # Called for a regular text block. Gets the paragraph, the line number, and | |
499 | # a Pod::Paragraph object. Perform interpolation and output the results. | |
500 | sub textblock { | |
501 | my $self = shift; | |
502 | return if $$self{EXCLUDE}; | |
503 | $self->output ($_[0]), return if $$self{VERBATIM}; | |
504 | ||
505 | # Perform a little magic to collapse multiple L<> references. We'll | |
506 | # just rewrite the whole thing into actual text at this part, bypassing | |
507 | # the whole internal sequence parsing thing. | |
c9abbd5d GS |
508 | my $text = shift; |
509 | $text =~ s{ | |
9741dab0 GS |
510 | (L< # A link of the form L</something>. |
511 | / | |
512 | ( | |
513 | [:\w]+ # The item has to be a simple word... | |
514 | (\(\))? # ...or simple function. | |
515 | ) | |
516 | > | |
517 | ( | |
518 | ,?\s+(and\s+)? # Allow lots of them, conjuncted. | |
519 | L< | |
520 | / | |
521 | ( [:\w]+ ( \(\) )? ) | |
522 | > | |
523 | )+ | |
524 | ) | |
525 | } { | |
526 | local $_ = $1; | |
c9abbd5d | 527 | s{ L< / ( [^>]+ ) > } {$1}xg; |
9741dab0 | 528 | my @items = split /(?:,?\s+(?:and\s+)?)/; |
c9abbd5d | 529 | my $string = 'the '; |
9741dab0 GS |
530 | my $i; |
531 | for ($i = 0; $i < @items; $i++) { | |
532 | $string .= $items[$i]; | |
c9abbd5d GS |
533 | $string .= ', ' if @items > 2 && $i != $#items; |
534 | $string .= ' ' if @items == 2 && $i == 2; | |
535 | $string .= 'and ' if ($i == $#items - 1); | |
9741dab0 | 536 | } |
c9abbd5d | 537 | $string .= ' entries elsewhere in this document'; |
9741dab0 GS |
538 | $string; |
539 | }gex; | |
540 | ||
541 | # Parse the tree and output it. collapse knows about references to | |
542 | # scalars as well as scalars and does the right thing with them. | |
c9abbd5d GS |
543 | $text = $self->parse ($text, @_); |
544 | $text =~ s/\n\s*$/\n/; | |
9741dab0 | 545 | $self->makespace if $$self{NEEDSPACE}; |
c9abbd5d | 546 | $self->output (protect $self->mapfonts ($text)); |
9741dab0 GS |
547 | $self->outindex; |
548 | $$self{NEEDSPACE} = 1; | |
549 | } | |
550 | ||
551 | # Called for an interior sequence. Takes a Pod::InteriorSequence object and | |
552 | # returns a reference to a scalar. This scalar is the final formatted text. | |
553 | # It's returned as a reference so that other interior sequences above us | |
554 | # know that the text has already been processed. | |
555 | sub sequence { | |
556 | my ($self, $seq) = @_; | |
557 | my $command = $seq->cmd_name; | |
558 | ||
559 | # Zero-width characters. | |
f5daac4a | 560 | if ($command eq 'Z') { |
c9abbd5d GS |
561 | # Workaround to generate a blessable reference, needed by 5.005. |
562 | my $tmp = '\&'; | |
563 | return bless \ "$tmp", 'Pod::Man::String'; | |
f5daac4a | 564 | } |
9741dab0 | 565 | |
a3e04946 GS |
566 | # C<>, L<>, X<>, and E<> don't apply guesswork to their contents. C<> |
567 | # needs some additional special handling. | |
568 | my $literal = ($command =~ /^[CELX]$/); | |
569 | $literal++ if $command eq 'C'; | |
570 | local $_ = $self->collapse ($seq->parse_tree, $literal); | |
9741dab0 GS |
571 | |
572 | # Handle E<> escapes. | |
573 | if ($command eq 'E') { | |
2e20e14f GS |
574 | if (/^\d+$/) { |
575 | return bless \ chr ($_), 'Pod::Man::String'; | |
576 | } elsif (exists $ESCAPES{$_}) { | |
9741dab0 GS |
577 | return bless \ "$ESCAPES{$_}", 'Pod::Man::String'; |
578 | } else { | |
579 | carp "Unknown escape E<$1>"; | |
580 | return bless \ "E<$_>", 'Pod::Man::String'; | |
581 | } | |
582 | } | |
583 | ||
584 | # For all the other sequences, empty content produces no output. | |
585 | return '' if $_ eq ''; | |
586 | ||
587 | # Handle formatting sequences. | |
588 | if ($command eq 'B') { | |
589 | return bless \ ('\f(BS' . $_ . '\f(BE'), 'Pod::Man::String'; | |
590 | } elsif ($command eq 'F') { | |
591 | return bless \ ('\f(IS' . $_ . '\f(IE'), 'Pod::Man::String'; | |
592 | } elsif ($command eq 'I') { | |
593 | return bless \ ('\f(IS' . $_ . '\f(IE'), 'Pod::Man::String'; | |
594 | } elsif ($command eq 'C') { | |
9741dab0 GS |
595 | return bless \ ('\f(FS\*(C`' . $_ . "\\*(C'\\f(FE"), |
596 | 'Pod::Man::String'; | |
597 | } | |
598 | ||
599 | # Handle links. | |
600 | if ($command eq 'L') { | |
c9abbd5d GS |
601 | # A bug in lvalue subs in 5.6 requires the temporary variable. |
602 | my $tmp = $self->buildlink ($_); | |
603 | return bless \ "$tmp", 'Pod::Man::String'; | |
9741dab0 GS |
604 | } |
605 | ||
606 | # Whitespace protection replaces whitespace with "\ ". | |
607 | if ($command eq 'S') { | |
608 | s/\s+/\\ /g; | |
609 | return bless \ "$_", 'Pod::Man::String'; | |
610 | } | |
611 | ||
612 | # Add an index entry to the list of ones waiting to be output. | |
613 | if ($command eq 'X') { push (@{ $$self{INDEX} }, $_); return '' } | |
614 | ||
615 | # Anything else is unknown. | |
616 | carp "Unknown sequence $command<$_>"; | |
617 | } | |
618 | ||
619 | ||
620 | ############################################################################ | |
621 | # Command paragraphs | |
622 | ############################################################################ | |
623 | ||
624 | # All command paragraphs take the paragraph and the line number. | |
625 | ||
626 | # First level heading. We can't output .IX in the NAME section due to a bug | |
627 | # in some versions of catman, so don't output a .IX for that section. .SH | |
628 | # already uses small caps, so remove any E<> sequences that would cause | |
629 | # them. | |
630 | sub cmd_head1 { | |
631 | my $self = shift; | |
632 | local $_ = $self->parse (@_); | |
633 | s/\s+$//; | |
634 | s/\\s-?\d//g; | |
635 | $self->output (switchquotes ('.SH', $self->mapfonts ($_))); | |
636 | $self->outindex (($_ eq 'NAME') ? () : ('Header', $_)); | |
637 | $$self{NEEDSPACE} = 0; | |
638 | } | |
639 | ||
640 | # Second level heading. | |
641 | sub cmd_head2 { | |
642 | my $self = shift; | |
643 | local $_ = $self->parse (@_); | |
644 | s/\s+$//; | |
645 | $self->output (switchquotes ('.Sh', $self->mapfonts ($_))); | |
646 | $self->outindex ('Subsection', $_); | |
647 | $$self{NEEDSPACE} = 0; | |
648 | } | |
649 | ||
650 | # Start a list. For indents after the first, wrap the outside indent in .RS | |
651 | # so that hanging paragraph tags will be correct. | |
652 | sub cmd_over { | |
653 | my $self = shift; | |
654 | local $_ = shift; | |
655 | unless (/^[-+]?\d+\s+$/) { $_ = $$self{indent} } | |
656 | if (@{ $$self{INDENTS} } > 0) { | |
657 | $self->output (".RS $$self{INDENT}\n"); | |
658 | } | |
659 | push (@{ $$self{INDENTS} }, $$self{INDENT}); | |
660 | $$self{INDENT} = ($_ + 0); | |
661 | } | |
662 | ||
663 | # End a list. If we've closed an embedded indent, we've mangled the hanging | |
664 | # paragraph indent, so temporarily replace it with .RS and set WEIRDINDENT. | |
665 | # We'll close that .RS at the next =back or =item. | |
666 | sub cmd_back { | |
667 | my $self = shift; | |
668 | $$self{INDENT} = pop @{ $$self{INDENTS} }; | |
669 | unless (defined $$self{INDENT}) { | |
670 | carp "Unmatched =back"; | |
671 | $$self{INDENT} = 0; | |
672 | } | |
673 | if ($$self{WEIRDINDENT}) { | |
674 | $self->output (".RE\n"); | |
675 | $$self{WEIRDINDENT} = 0; | |
676 | } | |
677 | if (@{ $$self{INDENTS} } > 0) { | |
678 | $self->output (".RE\n"); | |
679 | $self->output (".RS $$self{INDENT}\n"); | |
680 | $$self{WEIRDINDENT} = 1; | |
681 | } | |
682 | $$self{NEEDSPACE} = 1; | |
683 | } | |
684 | ||
685 | # An individual list item. Emit an index entry for anything that's | |
686 | # interesting, but don't emit index entries for things like bullets and | |
687 | # numbers. rofficate bullets too while we're at it (so for nice output, use | |
46bce7d0 GS |
688 | # * for your lists rather than o or . or - or some other thing). Newlines |
689 | # in an item title are turned into spaces since *roff can't handle them | |
690 | # embedded. | |
9741dab0 GS |
691 | sub cmd_item { |
692 | my $self = shift; | |
693 | local $_ = $self->parse (@_); | |
694 | s/\s+$//; | |
46bce7d0 | 695 | s/\s*\n\s*/ /g; |
9741dab0 GS |
696 | my $index; |
697 | if (/\w/ && !/^\w[.\)]\s*$/) { | |
698 | $index = $_; | |
fe2227f0 | 699 | $index =~ s/^\s*[-*+o.]?(?:\s+|\Z)//; |
9741dab0 GS |
700 | } |
701 | s/^\*(\s|\Z)/\\\(bu$1/; | |
702 | if ($$self{WEIRDINDENT}) { | |
703 | $self->output (".RE\n"); | |
704 | $$self{WEIRDINDENT} = 0; | |
705 | } | |
706 | $_ = $self->mapfonts ($_); | |
707 | $self->output (switchquotes ('.Ip', $_, $$self{INDENT})); | |
708 | $self->outindex ($index ? ('Item', $index) : ()); | |
709 | $$self{NEEDSPACE} = 0; | |
710 | } | |
711 | ||
712 | # Begin a block for a particular translator. Setting VERBATIM triggers | |
713 | # special handling in textblock(). | |
714 | sub cmd_begin { | |
715 | my $self = shift; | |
716 | local $_ = shift; | |
717 | my ($kind) = /^(\S+)/ or return; | |
718 | if ($kind eq 'man' || $kind eq 'roff') { | |
719 | $$self{VERBATIM} = 1; | |
720 | } else { | |
721 | $$self{EXCLUDE} = 1; | |
722 | } | |
723 | } | |
724 | ||
725 | # End a block for a particular translator. We assume that all =begin/=end | |
726 | # pairs are properly closed. | |
727 | sub cmd_end { | |
728 | my $self = shift; | |
729 | $$self{EXCLUDE} = 0; | |
730 | $$self{VERBATIM} = 0; | |
731 | } | |
732 | ||
733 | # One paragraph for a particular translator. Ignore it unless it's intended | |
734 | # for man or roff, in which case we output it verbatim. | |
735 | sub cmd_for { | |
736 | my $self = shift; | |
737 | local $_ = shift; | |
9741dab0 GS |
738 | return unless s/^(?:man|roff)\b[ \t]*\n?//; |
739 | $self->output ($_); | |
740 | } | |
741 | ||
742 | ||
743 | ############################################################################ | |
744 | # Link handling | |
745 | ############################################################################ | |
746 | ||
747 | # Handle links. We can't actually make real hyperlinks, so this is all to | |
748 | # figure out what text and formatting we print out. | |
749 | sub buildlink { | |
750 | my $self = shift; | |
751 | local $_ = shift; | |
752 | ||
753 | # Smash whitespace in case we were split across multiple lines. | |
754 | s/\s+/ /g; | |
755 | ||
756 | # If we were given any explicit text, just output it. | |
757 | if (m{ ^ ([^|]+) \| }x) { return $1 } | |
758 | ||
759 | # Okay, leading and trailing whitespace isn't important. | |
760 | s/^\s+//; | |
761 | s/\s+$//; | |
762 | ||
763 | # Default to using the whole content of the link entry as a section | |
764 | # name. Note that L<manpage/> forces a manpage interpretation, as does | |
765 | # something looking like L<manpage(section)>. Do the same thing to | |
766 | # L<manpage(section)> as we would to manpage(section) without the L<>; | |
767 | # see guesswork(). If we've added italics, don't add the "manpage" | |
768 | # text; markup is sufficient. | |
769 | my ($manpage, $section) = ('', $_); | |
770 | if (/^"\s*(.*?)\s*"$/) { | |
771 | $section = '"' . $1 . '"'; | |
772 | } elsif (m{ ^ [-:.\w]+ (?: \( \S+ \) )? $ }x) { | |
773 | ($manpage, $section) = ($_, ''); | |
774 | $manpage =~ s/^([^\(]+)\(/'\f(IS' . $1 . '\f(IE\|('/e; | |
775 | } elsif (m%/%) { | |
776 | ($manpage, $section) = split (/\s*\/\s*/, $_, 2); | |
777 | if ($manpage =~ /^[-:.\w]+(?:\(\S+\))?$/) { | |
778 | $manpage =~ s/^([^\(]+)\(/'\f(IS' . $1 . '\f(IE\|'/e; | |
779 | } | |
780 | $section =~ s/^\"\s*//; | |
781 | $section =~ s/\s*\"$//; | |
782 | } | |
783 | if ($manpage && $manpage !~ /\\f\(IS/) { | |
784 | $manpage = "the $manpage manpage"; | |
785 | } | |
786 | ||
787 | # Now build the actual output text. | |
788 | my $text = ''; | |
789 | if (!length ($section) && !length ($manpage)) { | |
790 | carp "Invalid link $_"; | |
791 | } elsif (!length ($section)) { | |
792 | $text = $manpage; | |
793 | } elsif ($section =~ /^[:\w]+(?:\(\))?/) { | |
794 | $text .= 'the ' . $section . ' entry'; | |
795 | $text .= (length $manpage) ? " in $manpage" | |
796 | : " elsewhere in this document"; | |
797 | } else { | |
2e20e14f GS |
798 | if ($section !~ /^".*"$/) { $section = '"' . $section . '"' } |
799 | $text .= 'the section on ' . $section; | |
9741dab0 GS |
800 | $text .= " in $manpage" if length $manpage; |
801 | } | |
802 | $text; | |
803 | } | |
804 | ||
805 | ||
806 | ############################################################################ | |
807 | # Escaping and fontification | |
808 | ############################################################################ | |
809 | ||
810 | # At this point, we'll have embedded font codes of the form \f(<font>[SE] | |
811 | # where <font> is one of B, I, or F. Turn those into the right font start | |
812 | # or end codes. B<someI<thing> else> should map to \fBsome\f(BIthing\fB | |
813 | # else\fR. The old pod2man didn't get this right; the second \fB was \fR, | |
814 | # so nested sequences didn't work right. We take care of this by using | |
815 | # variables as a combined pointer to our current font sequence, and set each | |
816 | # to the number of current nestings of start tags for that font. Use them | |
817 | # as a vector to look up what font sequence to use. | |
818 | sub mapfonts { | |
819 | my $self = shift; | |
820 | local $_ = shift; | |
821 | ||
822 | my ($fixed, $bold, $italic) = (0, 0, 0); | |
823 | my %magic = (F => \$fixed, B => \$bold, I => \$italic); | |
824 | s { \\f\((.)(.) } { | |
825 | ${ $magic{$1} } += ($2 eq 'S') ? 1 : -1; | |
826 | $$self{FONTS}{($fixed && 1) . ($bold && 1) . ($italic && 1)}; | |
827 | }gxe; | |
828 | $_; | |
829 | } | |
830 | ||
831 | ||
832 | ############################################################################ | |
833 | # *roff-specific parsing | |
834 | ############################################################################ | |
835 | ||
836 | # Called instead of parse_text, calls parse_text with the right flags. | |
837 | sub parse { | |
838 | my $self = shift; | |
839 | $self->parse_text ({ -expand_seq => 'sequence', | |
840 | -expand_ptree => 'collapse' }, @_); | |
841 | } | |
842 | ||
843 | # Takes a parse tree and a flag saying whether or not to treat it as literal | |
844 | # text (not call guesswork on it), and returns the concatenation of all of | |
845 | # the text strings in that parse tree. If the literal flag isn't true, | |
846 | # guesswork() will be called on all plain scalars in the parse tree. | |
a3e04946 GS |
847 | # Otherwise, just escape backslashes in the normal case. If collapse is |
848 | # being called on a C<> sequence, literal is set to 2, and we do some | |
849 | # additional cleanup. Assumes that everything in the parse tree is either a | |
850 | # scalar or a reference to a scalar. | |
9741dab0 GS |
851 | sub collapse { |
852 | my ($self, $ptree, $literal) = @_; | |
853 | if ($literal) { | |
854 | return join ('', map { | |
855 | if (ref $_) { | |
856 | $$_; | |
857 | } else { | |
858 | s/\\/\\e/g; | |
a3e04946 GS |
859 | s/-/\\-/g if $literal > 1; |
860 | s/__/_\\|_/g if $literal > 1; | |
9741dab0 GS |
861 | $_; |
862 | } | |
863 | } $ptree->children); | |
864 | } else { | |
865 | return join ('', map { | |
866 | ref ($_) ? $$_ : $self->guesswork ($_) | |
867 | } $ptree->children); | |
868 | } | |
869 | } | |
870 | ||
871 | # Takes a text block to perform guesswork on; this is guaranteed not to | |
872 | # contain any interior sequences. Returns the text block with remapping | |
873 | # done. | |
874 | sub guesswork { | |
875 | my $self = shift; | |
876 | local $_ = shift; | |
877 | ||
878 | # rofficate backslashes. | |
879 | s/\\/\\e/g; | |
880 | ||
881 | # Ensure double underbars have a tiny space between them. | |
882 | s/__/_\\|_/g; | |
883 | ||
884 | # Make all caps a little smaller. Be careful here, since we don't want | |
885 | # to make @ARGV into small caps, nor do we want to fix the MIME in | |
886 | # MIME-Version, since it looks weird with the full-height V. | |
887 | s{ | |
888 | ( ^ | [\s\(\"\'\`\[\{<>] ) | |
889 | ( [A-Z] [A-Z] [/A-Z+:\d_\$&-]* ) | |
890 | (?: (?= [\s>\}\]\)\'\".?!,;:] | -- ) | $ ) | |
c9abbd5d | 891 | } { $1 . '\s-1' . $2 . '\s0' }egx; |
9741dab0 GS |
892 | |
893 | # Turn PI into a pretty pi. | |
894 | s{ (?: \\s-1 | \b ) PI (?: \\s0 | \b ) } {\\*\(PI}gx; | |
895 | ||
896 | # Italize functions in the form func(). | |
897 | s{ | |
898 | \b | |
899 | ( | |
900 | [:\w]+ (?:\\s-1)? \(\) | |
901 | ) | |
902 | } { '\f(IS' . $1 . '\f(IE' }egx; | |
903 | ||
904 | # func(n) is a reference to a manual page. Make it \fIfunc\fR\|(n). | |
905 | s{ | |
906 | \b | |
907 | (\w[-:.\w]+ (?:\\s-1)?) | |
908 | ( | |
909 | \( [^\)] \) | |
910 | ) | |
911 | } { '\f(IS' . $1 . '\f(IE\|' . $2 }egx; | |
912 | ||
913 | # Convert simple Perl variable references to a fixed-width font. | |
914 | s{ | |
915 | ( \s+ ) | |
916 | ( [\$\@%] [\w:]+ ) | |
917 | (?! \( ) | |
918 | } { $1 . '\f(FS' . $2 . '\f(FE'}egx; | |
919 | ||
920 | # Translate -- into a real em dash if it's used like one and fix up | |
921 | # dashes, but keep hyphens hyphens. | |
922 | s{ (\G|^|.) (-+) (\b|.) } { | |
923 | my ($pre, $dash, $post) = ($1, $2, $3); | |
924 | if (length ($dash) == 1) { | |
925 | ($pre =~ /[a-zA-Z]/) ? "$pre-$post" : "$pre\\-$post"; | |
926 | } elsif (length ($dash) == 2 | |
927 | && ((!$pre && !$post) | |
928 | || ($pre =~ /\w/ && !$post) | |
929 | || ($pre eq ' ' && $post eq ' ') | |
930 | || ($pre eq '=' && $post ne '=') | |
931 | || ($pre ne '=' && $post eq '='))) { | |
932 | "$pre\\*(--$post"; | |
933 | } else { | |
934 | $pre . ('\-' x length $dash) . $post; | |
935 | } | |
936 | }egxs; | |
937 | ||
938 | # Fix up double quotes. | |
939 | s{ \" ([^\"]+) \" } { '\*(L"' . $1 . '\*(R"' }egx; | |
940 | ||
941 | # Make C++ into \*(C+, which is a squinched version. | |
942 | s{ \b C\+\+ } {\\*\(C+}gx; | |
943 | ||
944 | # All done. | |
945 | $_; | |
946 | } | |
947 | ||
948 | ||
949 | ############################################################################ | |
950 | # Output formatting | |
951 | ############################################################################ | |
952 | ||
953 | # Make vertical whitespace. | |
954 | sub makespace { | |
955 | my $self = shift; | |
956 | $self->output ($$self{INDENT} > 0 ? ".Sp\n" : ".PP\n"); | |
957 | } | |
958 | ||
959 | # Output any pending index entries, and optionally an index entry given as | |
960 | # an argument. Support multiple index entries in X<> separated by slashes, | |
961 | # and strip special escapes from index entries. | |
962 | sub outindex { | |
963 | my ($self, $section, $index) = @_; | |
964 | my @entries = map { split m%\s*/\s*% } @{ $$self{INDEX} }; | |
965 | return unless ($section || @entries); | |
966 | $$self{INDEX} = []; | |
967 | my $output; | |
968 | if (@entries) { | |
969 | my $output = '.IX Xref "' | |
970 | . join (' ', map { s/\"/\"\"/; $_ } @entries) | |
971 | . '"' . "\n"; | |
972 | } | |
973 | if ($section) { | |
974 | $index =~ s/\"/\"\"/; | |
975 | $index =~ s/\\-/-/g; | |
976 | $index =~ s/\\(?:s-?\d|.\(..|.)//g; | |
977 | $output .= ".IX $section " . '"' . $index . '"' . "\n"; | |
978 | } | |
979 | $self->output ($output); | |
980 | } | |
981 | ||
982 | # Output text to the output device. | |
983 | sub output { print { $_[0]->output_handle } $_[1] } | |
984 | ||
985 | __END__ | |
986 | ||
987 | .\" These are some extra bits of roff that I don't want to lose track of | |
988 | .\" but that have been removed from the preamble to make it a bit shorter | |
989 | .\" since they're not currently being used. They're accents and special | |
990 | .\" characters we don't currently have escapes for. | |
991 | .if n \{\ | |
992 | . ds ? ? | |
993 | . ds ! ! | |
994 | . ds q | |
995 | .\} | |
996 | .if t \{\ | |
997 | . ds ? \s-2c\h'-\w'c'u*7/10'\u\h'\*(#H'\zi\d\s+2\h'\w'c'u*8/10' | |
998 | . ds ! \s-2\(or\s+2\h'-\w'\(or'u'\v'-.8m'.\v'.8m' | |
999 | . ds q o\h'-\w'o'u*8/10'\s-4\v'.4m'\z\(*i\v'-.4m'\s+4\h'\w'o'u*8/10' | |
1000 | .\} | |
1001 | .ds v \\k:\h'-(\\n(.wu*9/10-\*(#H)'\v'-\*(#V'\*(#[\s-4v\s0\v'\*(#V'\h'|\\n:u'\*(#] | |
1002 | .ds _ \\k:\h'-(\\n(.wu*9/10-\*(#H+(\*(#F*2/3))'\v'-.4m'\z\(hy\v'.4m'\h'|\\n:u' | |
1003 | .ds . \\k:\h'-(\\n(.wu*8/10)'\v'\*(#V*4/10'\z.\v'-\*(#V*4/10'\h'|\\n:u' | |
1004 | .ds 3 \*(#[\v'.2m'\s-2\&3\s0\v'-.2m'\*(#] | |
1005 | .ds oe o\h'-(\w'o'u*4/10)'e | |
1006 | .ds Oe O\h'-(\w'O'u*4/10)'E | |
1007 | .if \n(.H>23 .if \n(.V>19 \ | |
1008 | \{\ | |
1009 | . ds v \h'-1'\o'\(aa\(ga' | |
1010 | . ds _ \h'-1'^ | |
1011 | . ds . \h'-1'. | |
1012 | . ds 3 3 | |
1013 | . ds oe oe | |
1014 | . ds Oe OE | |
1015 | .\} | |
1016 | ||
1017 | ############################################################################ | |
1018 | # Documentation | |
1019 | ############################################################################ | |
1020 | ||
1021 | =head1 NAME | |
1022 | ||
1023 | Pod::Man - Convert POD data to formatted *roff input | |
1024 | ||
1025 | =head1 SYNOPSIS | |
1026 | ||
1027 | use Pod::Man; | |
1028 | my $parser = Pod::Man->new (release => $VERSION, section => 8); | |
1029 | ||
1030 | # Read POD from STDIN and write to STDOUT. | |
1031 | $parser->parse_from_filehandle; | |
1032 | ||
1033 | # Read POD from file.pod and write to file.1. | |
1034 | $parser->parse_from_file ('file.pod', 'file.1'); | |
1035 | ||
1036 | =head1 DESCRIPTION | |
1037 | ||
1038 | Pod::Man is a module to convert documentation in the POD format (the | |
1039 | preferred language for documenting Perl) into *roff input using the man | |
1040 | macro set. The resulting *roff code is suitable for display on a terminal | |
1041 | using nroff(1), normally via man(1), or printing using troff(1). It is | |
9e107c59 | 1042 | conventionally invoked using the driver script B<pod2man>, but it can also |
9741dab0 GS |
1043 | be used directly. |
1044 | ||
1045 | As a derived class from Pod::Parser, Pod::Man supports the same methods and | |
1046 | interfaces. See L<Pod::Parser> for all the details; briefly, one creates a | |
1047 | new parser with C<Pod::Man-E<gt>new()> and then calls either | |
1048 | parse_from_filehandle() or parse_from_file(). | |
1049 | ||
1050 | new() can take options, in the form of key/value pairs that control the | |
1051 | behavior of the parser. See below for details. | |
1052 | ||
1053 | If no options are given, Pod::Man uses the name of the input file with any | |
1054 | trailing C<.pod>, C<.pm>, or C<.pl> stripped as the man page title, to | |
1055 | section 1 unless the file ended in C<.pm> in which case it defaults to | |
1056 | section 3, to a centered title of "User Contributed Perl Documentation", to | |
1057 | a centered footer of the Perl version it is run with, and to a left-hand | |
1058 | footer of the modification date of its input (or the current date if given | |
1059 | STDIN for input). | |
1060 | ||
1061 | Pod::Man assumes that your *roff formatters have a fixed-width font named | |
1062 | CW. If yours is called something else (like CR), use the C<fixed> option to | |
1063 | specify it. This generally only matters for troff output for printing. | |
1064 | Similarly, you can set the fonts used for bold, italic, and bold italic | |
1065 | fixed-width output. | |
1066 | ||
1067 | Besides the obvious pod conversions, Pod::Man also takes care of formatting | |
1068 | func(), func(n), and simple variable references like $foo or @bar so you | |
1069 | don't have to use code escapes for them; complex expressions like | |
1070 | C<$fred{'stuff'}> will still need to be escaped, though. It also translates | |
1071 | dashes that aren't used as hyphens into en dashes, makes long dashes--like | |
1072 | this--into proper em dashes, fixes "paired quotes," makes C++ and PI look | |
1073 | right, puts a little space between double underbars, makes ALLCAPS a teeny | |
1074 | bit smaller in troff(1), and escapes stuff that *roff treats as special so | |
1075 | that you don't have to. | |
1076 | ||
1077 | The recognized options to new() are as follows. All options take a single | |
1078 | argument. | |
1079 | ||
1080 | =over 4 | |
1081 | ||
1082 | =item center | |
1083 | ||
1084 | Sets the centered page header to use instead of "User Contributed Perl | |
1085 | Documentation". | |
1086 | ||
1087 | =item date | |
1088 | ||
1089 | Sets the left-hand footer. By default, the modification date of the input | |
1090 | file will be used, or the current date if stat() can't find that file (the | |
1091 | case if the input is from STDIN), and the date will be formatted as | |
1092 | YYYY-MM-DD. | |
1093 | ||
1094 | =item fixed | |
1095 | ||
1096 | The fixed-width font to use for vertabim text and code. Defaults to CW. | |
1097 | Some systems may want CR instead. Only matters for troff(1) output. | |
1098 | ||
1099 | =item fixedbold | |
1100 | ||
1101 | Bold version of the fixed-width font. Defaults to CB. Only matters for | |
1102 | troff(1) output. | |
1103 | ||
1104 | =item fixeditalic | |
1105 | ||
1106 | Italic version of the fixed-width font (actually, something of a misnomer, | |
1107 | since most fixed-width fonts only have an oblique version, not an italic | |
1108 | version). Defaults to CI. Only matters for troff(1) output. | |
1109 | ||
1110 | =item fixedbolditalic | |
1111 | ||
1112 | Bold italic (probably actually oblique) version of the fixed-width font. | |
1113 | Pod::Man doesn't assume you have this, and defaults to CB. Some systems | |
1114 | (such as Solaris) have this font available as CX. Only matters for troff(1) | |
1115 | output. | |
1116 | ||
1117 | =item release | |
1118 | ||
1119 | Set the centered footer. By default, this is the version of Perl you run | |
1120 | Pod::Man under. Note that some system an macro sets assume that the | |
1121 | centered footer will be a modification date and will prepend something like | |
1122 | "Last modified: "; if this is the case, you may want to set C<release> to | |
1123 | the last modified date and C<date> to the version number. | |
1124 | ||
1125 | =item section | |
1126 | ||
1127 | Set the section for the C<.TH> macro. The standard section numbering | |
1128 | convention is to use 1 for user commands, 2 for system calls, 3 for | |
1129 | functions, 4 for devices, 5 for file formats, 6 for games, 7 for | |
1130 | miscellaneous information, and 8 for administrator commands. There is a lot | |
1131 | of variation here, however; some systems (like Solaris) use 4 for file | |
1132 | formats, 5 for miscellaneous information, and 7 for devices. Still others | |
1133 | use 1m instead of 8, or some mix of both. About the only section numbers | |
1134 | that are reliably consistent are 1, 2, and 3. | |
1135 | ||
1136 | By default, section 1 will be used unless the file ends in .pm in which case | |
1137 | section 3 will be selected. | |
1138 | ||
1139 | =back | |
1140 | ||
1141 | The standard Pod::Parser method parse_from_filehandle() takes up to two | |
1142 | arguments, the first being the file handle to read POD from and the second | |
1143 | being the file handle to write the formatted output to. The first defaults | |
1144 | to STDIN if not given, and the second defaults to STDOUT. The method | |
1145 | parse_from_file() is almost identical, except that its two arguments are the | |
1146 | input and output disk files instead. See L<Pod::Parser> for the specific | |
1147 | details. | |
1148 | ||
1149 | =head1 DIAGNOSTICS | |
1150 | ||
1151 | =over 4 | |
1152 | ||
1153 | =item roff font should be 1 or 2 chars, not `%s' | |
1154 | ||
1155 | (F) You specified a *roff font (using C<fixed>, C<fixedbold>, etc.) that | |
1156 | wasn't either one or two characters. Pod::Man doesn't support *roff fonts | |
1157 | longer than two characters, although some *roff extensions do (the canonical | |
1158 | versions of nroff(1) and troff(1) don't either). | |
1159 | ||
1160 | =item Invalid link %s | |
1161 | ||
1162 | (W) The POD source contained a C<LE<lt>E<gt>> sequence that Pod::Man was | |
1163 | unable to parse. You should never see this error message; it probably | |
1164 | indicates a bug in Pod::Man. | |
1165 | ||
1166 | =item Unknown escape EE<lt>%sE<gt> | |
1167 | ||
1168 | (W) The POD source contained an C<EE<lt>E<gt>> escape that Pod::Man didn't | |
1169 | know about. C<EE<lt>%sE<gt>> was printed verbatim in the output. | |
1170 | ||
1171 | =item Unknown sequence %s | |
1172 | ||
1173 | (W) The POD source contained a non-standard interior sequence (something of | |
1174 | the form C<XE<lt>E<gt>>) that Pod::Man didn't know about. It was ignored. | |
1175 | ||
844b31e3 A |
1176 | =item %s: Unknown command paragraph "%s" on line %d. |
1177 | ||
1178 | (W) The POD source contained a non-standard command paragraph (something of | |
1179 | the form C<=command args>) that Pod::Man didn't know about. It was ignored. | |
1180 | ||
9741dab0 GS |
1181 | =item Unmatched =back |
1182 | ||
1183 | (W) Pod::Man encountered a C<=back> command that didn't correspond to an | |
1184 | C<=over> command. | |
1185 | ||
1186 | =back | |
1187 | ||
1188 | =head1 BUGS | |
1189 | ||
1190 | The lint-like features and strict POD format checking done by B<pod2man> are | |
1191 | not yet implemented and should be, along with the corresponding C<lax> | |
1192 | option. | |
1193 | ||
1194 | The NAME section should be recognized specially and index entries emitted | |
1195 | for everything in that section. This would have to be deferred until the | |
1196 | next section, since extraneous things in NAME tends to confuse various man | |
1197 | page processors. | |
1198 | ||
1199 | The handling of hyphens, en dashes, and em dashes is somewhat fragile, and | |
1200 | one may get the wrong one under some circumstances. This should only matter | |
1201 | for troff(1) output. | |
1202 | ||
1203 | When and whether to use small caps is somewhat tricky, and Pod::Man doesn't | |
1204 | necessarily get it right. | |
1205 | ||
1206 | Pod::Man doesn't handle font names longer than two characters. Neither do | |
1207 | most troff(1) implementations, but GNU troff does as an extension. It would | |
1208 | be nice to support as an option for those who want to use it. | |
1209 | ||
1210 | The preamble added to each output file is rather verbose, and most of it is | |
1211 | only necessary in the presence of EE<lt>E<gt> escapes for non-ASCII | |
1212 | characters. It would ideally be nice if all of those definitions were only | |
1213 | output if needed, perhaps on the fly as the characters are used. | |
1214 | ||
1215 | Some of the automagic applied to file names assumes Unix directory | |
1216 | separators. | |
1217 | ||
1218 | Pod::Man is excessively slow. | |
1219 | ||
9741dab0 GS |
1220 | =head1 SEE ALSO |
1221 | ||
9e107c59 | 1222 | L<Pod::Parser|Pod::Parser>, perlpod(1), pod2man(1), nroff(1), troff(1), |
9741dab0 GS |
1223 | man(1), man(7) |
1224 | ||
1225 | Ossanna, Joseph F., and Brian W. Kernighan. "Troff User's Manual," | |
1226 | Computing Science Technical Report No. 54, AT&T Bell Laboratories. This is | |
1227 | the best documentation of standard nroff(1) and troff(1). At the time of | |
1228 | this writing, it's available at http://www.cs.bell-labs.com/cm/cs/cstr.html. | |
1229 | ||
1230 | The man page documenting the man macro set may be man(5) instead of man(7) | |
9e107c59 | 1231 | on your system. Also, please see pod2man(1) for extensive documentation on |
9741dab0 GS |
1232 | writing manual pages if you've not done it before and aren't familiar with |
1233 | the conventions. | |
1234 | ||
1235 | =head1 AUTHOR | |
1236 | ||
1237 | Russ Allbery E<lt>rra@stanford.eduE<gt>, based I<very> heavily on the | |
1238 | original B<pod2man> by Tom Christiansen E<lt>tchrist@mox.perl.comE<gt>. | |
1239 | ||
1240 | =cut |