Commit | Line | Data |
54310121 |
1 | package Pod::Html; |
2 | |
3 | use Pod::Functions; |
4 | use Getopt::Long; # package for handling command-line parameters |
5 | require Exporter; |
6 | @ISA = Exporter; |
7 | @EXPORT = qw(pod2html htmlify); |
8 | use Cwd; |
9 | |
10 | use Carp; |
11 | |
12 | use strict; |
13 | |
14 | =head1 NAME |
15 | |
16 | Pod::HTML - module to convert pod files to HTML |
17 | |
18 | =head1 SYNOPSIS |
19 | |
20 | use Pod::HTML; |
21 | pod2html([options]); |
22 | |
23 | =head1 DESCRIPTION |
24 | |
25 | Converts files from pod format (see L<perlpod>) to HTML format. It |
26 | can automatically generate indexes and cross-references, and it keeps |
27 | a cache of things it knows how to cross-reference. |
28 | |
29 | =head1 ARGUMENTS |
30 | |
31 | Pod::Html takes the following arguments: |
32 | |
33 | =over 4 |
34 | |
35 | =item help |
36 | |
37 | --help |
38 | |
39 | Displays the usage message. |
40 | |
41 | =item htmlroot |
42 | |
43 | --htmlroot=name |
44 | |
45 | Sets the base URL for the HTML files. When cross-references are made, |
46 | the HTML root is prepended to the URL. |
47 | |
48 | =item infile |
49 | |
50 | --infile=name |
51 | |
52 | Specify the pod file to convert. Input is taken from STDIN if no |
53 | infile is specified. |
54 | |
55 | =item outfile |
56 | |
57 | --outfile=name |
58 | |
59 | Specify the HTML file to create. Output goes to STDOUT if no outfile |
60 | is specified. |
61 | |
62 | =item podroot |
63 | |
64 | --podroot=name |
65 | |
66 | Specify the base directory for finding library pods. |
67 | |
68 | =item podpath |
69 | |
70 | --podpath=name:...:name |
71 | |
72 | Specify which subdirectories of the podroot contain pod files whose |
73 | HTML converted forms can be linked-to in cross-references. |
74 | |
75 | =item libpods |
76 | |
77 | --libpods=name:...:name |
78 | |
79 | List of page names (eg, "perlfunc") which contain linkable C<=item>s. |
80 | |
81 | =item netscape |
82 | |
83 | --netscape |
84 | |
85 | Use Netscape HTML directives when applicable. |
86 | |
87 | =item nonetscape |
88 | |
89 | --nonetscape |
90 | |
91 | Do not use Netscape HTML directives (default). |
92 | |
93 | =item index |
94 | |
95 | --index |
96 | |
97 | Generate an index at the top of the HTML file (default behaviour). |
98 | |
99 | =item noindex |
100 | |
101 | --noindex |
102 | |
103 | Do not generate an index at the top of the HTML file. |
104 | |
105 | |
106 | =item recurse |
107 | |
108 | --recurse |
109 | |
110 | Recurse into subdirectories specified in podpath (default behaviour). |
111 | |
112 | =item norecurse |
113 | |
114 | --norecurse |
115 | |
116 | Do not recurse into subdirectories specified in podpath. |
117 | |
118 | =item title |
119 | |
120 | --title=title |
121 | |
122 | Specify the title of the resulting HTML file. |
123 | |
124 | =item verbose |
125 | |
126 | --verbose |
127 | |
128 | Display progress messages. |
129 | |
130 | =back |
131 | |
132 | =head1 EXAMPLE |
133 | |
134 | pod2html("pod2html", |
135 | "--podpath=lib:ext:pod:vms", |
136 | "--podroot=/usr/src/perl", |
137 | "--htmlroot=/perl/nmanual", |
138 | "--libpods=perlfunc:perlguts:perlvar:perlrun:perlop", |
139 | "--recurse", |
140 | "--infile=foo.pod", |
141 | "--outfile=/perl/nmanual/foo.html"); |
142 | |
143 | =head1 AUTHOR |
144 | |
145 | Tom Christiansen, E<lt>tchrist@perl.comE<gt>. |
146 | |
147 | =head1 BUGS |
148 | |
149 | Has trouble with C<> etc in = commands. |
150 | |
151 | =head1 SEE ALSO |
152 | |
153 | L<perlpod> |
154 | |
155 | =head1 COPYRIGHT |
156 | |
157 | This program is distributed under the Artistic License. |
158 | |
159 | =cut |
160 | |
161 | my $dircache = "pod2html-dircache"; |
162 | my $itemcache = "pod2html-itemcache"; |
163 | |
164 | my @begin_stack = (); # begin/end stack |
165 | |
166 | my @libpods = (); # files to search for links from C<> directives |
167 | my $htmlroot = "/"; # http-server base directory from which all |
168 | # relative paths in $podpath stem. |
169 | my $htmlfile = ""; # write to stdout by default |
170 | my $podfile = ""; # read from stdin by default |
171 | my @podpath = (); # list of directories containing library pods. |
172 | my $podroot = "."; # filesystem base directory from which all |
173 | # relative paths in $podpath stem. |
174 | my $recurse = 1; # recurse on subdirectories in $podpath. |
175 | my $verbose = 0; # not verbose by default |
176 | my $doindex = 1; # non-zero if we should generate an index |
177 | my $listlevel = 0; # current list depth |
178 | my @listitem = (); # stack of HTML commands to use when a =item is |
179 | # encountered. the top of the stack is the |
180 | # current list. |
181 | my @listdata = (); # similar to @listitem, but for the text after |
182 | # an =item |
183 | my @listend = (); # similar to @listitem, but the text to use to |
184 | # end the list. |
185 | my $ignore = 1; # whether or not to format text. we don't |
186 | # format text until we hit our first pod |
187 | # directive. |
188 | |
189 | my %items_named = (); # for the multiples of the same item in perlfunc |
190 | my @items_seen = (); |
191 | my $netscape = 0; # whether or not to use netscape directives. |
192 | my $title; # title to give the pod(s) |
193 | my $top = 1; # true if we are at the top of the doc. used |
194 | # to prevent the first <HR> directive. |
195 | my $paragraph; # which paragraph we're processing (used |
196 | # for error messages) |
197 | my %pages = (); # associative array used to find the location |
198 | # of pages referenced by L<> links. |
199 | my %sections = (); # sections within this page |
200 | my %items = (); # associative array used to find the location |
201 | # of =item directives referenced by C<> links |
39e571d4 |
202 | my $Is83; # is dos with short filenames (8.3) |
203 | |
54310121 |
204 | sub init_globals { |
205 | $dircache = "pod2html-dircache"; |
206 | $itemcache = "pod2html-itemcache"; |
207 | |
208 | @begin_stack = (); # begin/end stack |
209 | |
210 | @libpods = (); # files to search for links from C<> directives |
211 | $htmlroot = "/"; # http-server base directory from which all |
212 | # relative paths in $podpath stem. |
213 | $htmlfile = ""; # write to stdout by default |
214 | $podfile = ""; # read from stdin by default |
215 | @podpath = (); # list of directories containing library pods. |
216 | $podroot = "."; # filesystem base directory from which all |
217 | # relative paths in $podpath stem. |
218 | $recurse = 1; # recurse on subdirectories in $podpath. |
219 | $verbose = 0; # not verbose by default |
220 | $doindex = 1; # non-zero if we should generate an index |
221 | $listlevel = 0; # current list depth |
222 | @listitem = (); # stack of HTML commands to use when a =item is |
223 | # encountered. the top of the stack is the |
224 | # current list. |
225 | @listdata = (); # similar to @listitem, but for the text after |
226 | # an =item |
227 | @listend = (); # similar to @listitem, but the text to use to |
228 | # end the list. |
229 | $ignore = 1; # whether or not to format text. we don't |
230 | # format text until we hit our first pod |
231 | # directive. |
232 | |
233 | @items_seen = (); |
234 | %items_named = (); |
235 | $netscape = 0; # whether or not to use netscape directives. |
236 | $title = ''; # title to give the pod(s) |
237 | $top = 1; # true if we are at the top of the doc. used |
238 | # to prevent the first <HR> directive. |
239 | $paragraph = ''; # which paragraph we're processing (used |
240 | # for error messages) |
54310121 |
241 | %sections = (); # sections within this page |
3e3baf6d |
242 | |
243 | # These are not reinitialised here but are kept as a cache. |
244 | # See get_cache and related cache management code. |
245 | #%pages = (); # associative array used to find the location |
246 | # of pages referenced by L<> links. |
247 | #%items = (); # associative array used to find the location |
54310121 |
248 | # of =item directives referenced by C<> links |
39e571d4 |
249 | $Is83=$^O eq 'dos'; |
54310121 |
250 | } |
251 | |
252 | sub pod2html { |
253 | local(@ARGV) = @_; |
254 | local($/); |
255 | local $_; |
256 | |
257 | init_globals(); |
258 | |
39e571d4 |
259 | $Is83 = 0 if (defined (&Dos::UseLFN) && Dos::UseLFN()); |
260 | |
54310121 |
261 | # cache of %pages and %items from last time we ran pod2html |
54310121 |
262 | |
263 | #undef $opt_help if defined $opt_help; |
264 | |
265 | # parse the command-line parameters |
266 | parse_command_line(); |
267 | |
268 | # set some variables to their default values if necessary |
269 | local *POD; |
270 | unless (@ARGV && $ARGV[0]) { |
271 | $podfile = "-" unless $podfile; # stdin |
272 | open(POD, "<$podfile") |
273 | || die "$0: cannot open $podfile file for input: $!\n"; |
274 | } else { |
275 | $podfile = $ARGV[0]; # XXX: might be more filenames |
276 | *POD = *ARGV; |
277 | } |
278 | $htmlfile = "-" unless $htmlfile; # stdout |
279 | $htmlroot = "" if $htmlroot eq "/"; # so we don't get a // |
280 | |
281 | # read the pod a paragraph at a time |
282 | warn "Scanning for sections in input file(s)\n" if $verbose; |
283 | $/ = ""; |
284 | my @poddata = <POD>; |
285 | close(POD); |
286 | |
287 | # scan the pod for =head[1-6] directives and build an index |
288 | my $index = scan_headings(\%sections, @poddata); |
289 | |
3e3baf6d |
290 | unless($index) { |
291 | warn "No pod in $podfile\n" if $verbose; |
292 | return; |
293 | } |
294 | |
54310121 |
295 | # open the output file |
296 | open(HTML, ">$htmlfile") |
297 | || die "$0: cannot open $htmlfile file for output: $!\n"; |
298 | |
299 | # put a title in the HTML file |
300 | $title = ''; |
301 | TITLE_SEARCH: { |
302 | for (my $i = 0; $i < @poddata; $i++) { |
303 | if ($poddata[$i] =~ /^=head1\s*NAME\b/m) { |
304 | for my $para ( @poddata[$i, $i+1] ) { |
305 | last TITLE_SEARCH if ($title) = $para =~ /(\S+\s+-+\s*.*)/s; |
306 | } |
307 | } |
308 | |
309 | } |
310 | } |
3e3baf6d |
311 | if (!$title and $podfile =~ /\.pod$/) { |
312 | # probably a split pod so take first =head[12] as title |
313 | for (my $i = 0; $i < @poddata; $i++) { |
314 | last if ($title) = $poddata[$i] =~ /^=head[12]\s*(.*)/; |
315 | } |
316 | warn "adopted '$title' as title for $podfile\n" |
317 | if $verbose and $title; |
318 | } |
54310121 |
319 | unless ($title) { |
3e3baf6d |
320 | warn "$0: no title for $podfile"; |
54310121 |
321 | $podfile =~ /^(.*)(\.[^.\/]+)?$/; |
322 | $title = ($podfile eq "-" ? 'No Title' : $1); |
3e3baf6d |
323 | warn "using $title" if $verbose; |
54310121 |
324 | } |
325 | print HTML <<END_OF_HEAD; |
326 | <HTML> |
327 | <HEAD> |
328 | <TITLE>$title</TITLE> |
329 | </HEAD> |
330 | |
331 | <BODY> |
332 | |
333 | END_OF_HEAD |
334 | |
3e3baf6d |
335 | # load/reload/validate/cache %pages and %items |
336 | get_cache($dircache, $itemcache, \@podpath, $podroot, $recurse); |
54310121 |
337 | |
338 | # scan the pod for =item directives |
339 | scan_items("", \%items, @poddata); |
340 | |
341 | # put an index at the top of the file. note, if $doindex is 0 we |
342 | # still generate an index, but surround it with an html comment. |
343 | # that way some other program can extract it if desired. |
344 | $index =~ s/--+/-/g; |
345 | print HTML "<!-- INDEX BEGIN -->\n"; |
346 | print HTML "<!--\n" unless $doindex; |
347 | print HTML $index; |
348 | print HTML "-->\n" unless $doindex; |
349 | print HTML "<!-- INDEX END -->\n\n"; |
350 | print HTML "<HR>\n" if $doindex; |
351 | |
352 | # now convert this file |
353 | warn "Converting input file\n" if $verbose; |
354 | foreach my $i (0..$#poddata) { |
355 | $_ = $poddata[$i]; |
356 | $paragraph = $i+1; |
357 | if (/^(=.*)/s) { # is it a pod directive? |
358 | $ignore = 0; |
359 | $_ = $1; |
360 | if (/^=begin\s+(\S+)\s*(.*)/si) {# =begin |
361 | process_begin($1, $2); |
362 | } elsif (/^=end\s+(\S+)\s*(.*)/si) {# =end |
363 | process_end($1, $2); |
364 | } elsif (/^=cut/) { # =cut |
365 | process_cut(); |
366 | } elsif (/^=pod/) { # =pod |
367 | process_pod(); |
368 | } else { |
369 | next if @begin_stack && $begin_stack[-1] ne 'html'; |
370 | |
371 | if (/^=(head[1-6])\s+(.*)/s) { # =head[1-6] heading |
372 | process_head($1, $2); |
373 | } elsif (/^=item\s*(.*)/sm) { # =item text |
374 | process_item($1); |
375 | } elsif (/^=over\s*(.*)/) { # =over N |
376 | process_over(); |
377 | } elsif (/^=back/) { # =back |
378 | process_back(); |
379 | } elsif (/^=for\s+(\S+)\s+(.*)/si) {# =for |
380 | process_for($1,$2); |
381 | } else { |
382 | /^=(\S*)\s*/; |
383 | warn "$0: $podfile: unknown pod directive '$1' in " |
384 | . "paragraph $paragraph. ignoring.\n"; |
385 | } |
386 | } |
387 | $top = 0; |
388 | } |
389 | else { |
390 | next if $ignore; |
391 | next if @begin_stack && $begin_stack[-1] ne 'html'; |
392 | my $text = $_; |
393 | process_text(\$text, 1); |
394 | print HTML "$text\n<P>\n\n"; |
395 | } |
396 | } |
397 | |
398 | # finish off any pending directives |
399 | finish_list(); |
400 | print HTML <<END_OF_TAIL; |
401 | </BODY> |
402 | |
403 | </HTML> |
404 | END_OF_TAIL |
405 | |
406 | # close the html file |
407 | close(HTML); |
408 | |
409 | warn "Finished\n" if $verbose; |
410 | } |
411 | |
412 | ############################################################################## |
413 | |
414 | my $usage; # see below |
415 | sub usage { |
416 | my $podfile = shift; |
417 | warn "$0: $podfile: @_\n" if @_; |
418 | die $usage; |
419 | } |
420 | |
421 | $usage =<<END_OF_USAGE; |
422 | Usage: $0 --help --htmlroot=<name> --infile=<name> --outfile=<name> |
423 | --podpath=<name>:...:<name> --podroot=<name> |
424 | --libpods=<name>:...:<name> --recurse --verbose --index |
425 | --netscape --norecurse --noindex |
426 | |
427 | --flush - flushes the item and directory caches. |
428 | --help - prints this message. |
429 | --htmlroot - http-server base directory from which all relative paths |
430 | in podpath stem (default is /). |
431 | --index - generate an index at the top of the resulting html |
432 | (default). |
433 | --infile - filename for the pod to convert (input taken from stdin |
434 | by default). |
435 | --libpods - colon-separated list of pages to search for =item pod |
436 | directives in as targets of C<> and implicit links (empty |
437 | by default). note, these are not filenames, but rather |
438 | page names like those that appear in L<> links. |
439 | --netscape - will use netscape html directives when applicable. |
440 | --nonetscape - will not use netscape directives (default). |
441 | --outfile - filename for the resulting html file (output sent to |
442 | stdout by default). |
443 | --podpath - colon-separated list of directories containing library |
444 | pods. empty by default. |
445 | --podroot - filesystem base directory from which all relative paths |
446 | in podpath stem (default is .). |
447 | --noindex - don't generate an index at the top of the resulting html. |
448 | --norecurse - don't recurse on those subdirectories listed in podpath. |
449 | --recurse - recurse on those subdirectories listed in podpath |
450 | (default behavior). |
451 | --title - title that will appear in resulting html file. |
452 | --verbose - self-explanatory |
453 | |
454 | END_OF_USAGE |
455 | |
456 | sub parse_command_line { |
457 | my ($opt_flush,$opt_help,$opt_htmlroot,$opt_index,$opt_infile,$opt_libpods,$opt_netscape,$opt_outfile,$opt_podpath,$opt_podroot,$opt_norecurse,$opt_recurse,$opt_title,$opt_verbose); |
458 | my $result = GetOptions( |
459 | 'flush' => \$opt_flush, |
460 | 'help' => \$opt_help, |
461 | 'htmlroot=s' => \$opt_htmlroot, |
462 | 'index!' => \$opt_index, |
463 | 'infile=s' => \$opt_infile, |
464 | 'libpods=s' => \$opt_libpods, |
465 | 'netscape!' => \$opt_netscape, |
466 | 'outfile=s' => \$opt_outfile, |
467 | 'podpath=s' => \$opt_podpath, |
468 | 'podroot=s' => \$opt_podroot, |
469 | 'norecurse' => \$opt_norecurse, |
470 | 'recurse!' => \$opt_recurse, |
471 | 'title=s' => \$opt_title, |
472 | 'verbose' => \$opt_verbose, |
473 | ); |
474 | usage("-", "invalid parameters") if not $result; |
475 | |
476 | usage("-") if defined $opt_help; # see if the user asked for help |
477 | $opt_help = ""; # just to make -w shut-up. |
478 | |
479 | $podfile = $opt_infile if defined $opt_infile; |
480 | $htmlfile = $opt_outfile if defined $opt_outfile; |
481 | |
482 | @podpath = split(":", $opt_podpath) if defined $opt_podpath; |
483 | @libpods = split(":", $opt_libpods) if defined $opt_libpods; |
484 | |
485 | warn "Flushing item and directory caches\n" |
486 | if $opt_verbose && defined $opt_flush; |
487 | unlink($dircache, $itemcache) if defined $opt_flush; |
488 | |
489 | $htmlroot = $opt_htmlroot if defined $opt_htmlroot; |
490 | $podroot = $opt_podroot if defined $opt_podroot; |
491 | |
492 | $doindex = $opt_index if defined $opt_index; |
493 | $recurse = $opt_recurse if defined $opt_recurse; |
494 | $title = $opt_title if defined $opt_title; |
495 | $verbose = defined $opt_verbose ? 1 : 0; |
496 | $netscape = $opt_netscape if defined $opt_netscape; |
497 | } |
498 | |
3e3baf6d |
499 | |
500 | my $saved_cache_key; |
501 | |
502 | sub get_cache { |
503 | my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_; |
504 | my @cache_key_args = @_; |
505 | |
506 | # A first-level cache: |
507 | # Don't bother reading the cache files if they still apply |
508 | # and haven't changed since we last read them. |
509 | |
510 | my $this_cache_key = cache_key(@cache_key_args); |
511 | |
512 | return if $saved_cache_key and $this_cache_key eq $saved_cache_key; |
513 | |
514 | # load the cache of %pages and %items if possible. $tests will be |
515 | # non-zero if successful. |
516 | my $tests = 0; |
517 | if (-f $dircache && -f $itemcache) { |
518 | warn "scanning for item cache\n" if $verbose; |
519 | $tests = load_cache($dircache, $itemcache, $podpath, $podroot); |
520 | } |
521 | |
522 | # if we didn't succeed in loading the cache then we must (re)build |
523 | # %pages and %items. |
524 | if (!$tests) { |
525 | warn "scanning directories in pod-path\n" if $verbose; |
526 | scan_podpath($podroot, $recurse, 0); |
527 | } |
528 | $saved_cache_key = cache_key(@cache_key_args); |
529 | } |
530 | |
531 | sub cache_key { |
532 | my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_; |
533 | return join('!', $dircache, $itemcache, $recurse, |
534 | @$podpath, $podroot, stat($dircache), stat($itemcache)); |
535 | } |
536 | |
54310121 |
537 | # |
3e3baf6d |
538 | # load_cache - tries to find if the caches stored in $dircache and $itemcache |
54310121 |
539 | # are valid caches of %pages and %items. if they are valid then it loads |
540 | # them and returns a non-zero value. |
541 | # |
3e3baf6d |
542 | |
543 | sub load_cache { |
54310121 |
544 | my($dircache, $itemcache, $podpath, $podroot) = @_; |
545 | my($tests); |
546 | local $_; |
547 | |
548 | $tests = 0; |
549 | |
550 | open(CACHE, "<$itemcache") || |
551 | die "$0: error opening $itemcache for reading: $!\n"; |
552 | $/ = "\n"; |
553 | |
554 | # is it the same podpath? |
555 | $_ = <CACHE>; |
556 | chomp($_); |
3e3baf6d |
557 | $tests++ if (join(":", @$podpath) eq $_); |
54310121 |
558 | |
559 | # is it the same podroot? |
560 | $_ = <CACHE>; |
561 | chomp($_); |
562 | $tests++ if ($podroot eq $_); |
563 | |
564 | # load the cache if its good |
565 | if ($tests != 2) { |
566 | close(CACHE); |
54310121 |
567 | return 0; |
568 | } |
569 | |
570 | warn "loading item cache\n" if $verbose; |
571 | while (<CACHE>) { |
572 | /(.*?) (.*)$/; |
573 | $items{$1} = $2; |
574 | } |
575 | close(CACHE); |
576 | |
577 | warn "scanning for directory cache\n" if $verbose; |
578 | open(CACHE, "<$dircache") || |
579 | die "$0: error opening $dircache for reading: $!\n"; |
580 | $/ = "\n"; |
581 | $tests = 0; |
582 | |
583 | # is it the same podpath? |
584 | $_ = <CACHE>; |
585 | chomp($_); |
3e3baf6d |
586 | $tests++ if (join(":", @$podpath) eq $_); |
54310121 |
587 | |
588 | # is it the same podroot? |
589 | $_ = <CACHE>; |
590 | chomp($_); |
591 | $tests++ if ($podroot eq $_); |
592 | |
593 | # load the cache if its good |
594 | if ($tests != 2) { |
595 | close(CACHE); |
54310121 |
596 | return 0; |
597 | } |
598 | |
599 | warn "loading directory cache\n" if $verbose; |
600 | while (<CACHE>) { |
601 | /(.*?) (.*)$/; |
602 | $pages{$1} = $2; |
603 | } |
604 | |
605 | close(CACHE); |
606 | |
607 | return 1; |
608 | } |
609 | |
610 | # |
611 | # scan_podpath - scans the directories specified in @podpath for directories, |
612 | # .pod files, and .pm files. it also scans the pod files specified in |
613 | # @libpods for =item directives. |
614 | # |
615 | sub scan_podpath { |
3e3baf6d |
616 | my($podroot, $recurse, $append) = @_; |
54310121 |
617 | my($pwd, $dir); |
618 | my($libpod, $dirname, $pod, @files, @poddata); |
619 | |
3e3baf6d |
620 | unless($append) { |
621 | %items = (); |
622 | %pages = (); |
623 | } |
624 | |
54310121 |
625 | # scan each directory listed in @podpath |
626 | $pwd = getcwd(); |
627 | chdir($podroot) |
628 | || die "$0: error changing to directory $podroot: $!\n"; |
629 | foreach $dir (@podpath) { |
630 | scan_dir($dir, $recurse); |
631 | } |
632 | |
633 | # scan the pods listed in @libpods for =item directives |
634 | foreach $libpod (@libpods) { |
635 | # if the page isn't defined then we won't know where to find it |
636 | # on the system. |
637 | next unless defined $pages{$libpod} && $pages{$libpod}; |
638 | |
639 | # if there is a directory then use the .pod and .pm files within it. |
640 | if ($pages{$libpod} =~ /([^:]*[^(\.pod|\.pm)]):/) { |
641 | # find all the .pod and .pm files within the directory |
642 | $dirname = $1; |
643 | opendir(DIR, $dirname) || |
644 | die "$0: error opening directory $dirname: $!\n"; |
645 | @files = grep(/(\.pod|\.pm)$/ && ! -d $_, readdir(DIR)); |
646 | closedir(DIR); |
647 | |
648 | # scan each .pod and .pm file for =item directives |
649 | foreach $pod (@files) { |
650 | open(POD, "<$dirname/$pod") || |
651 | die "$0: error opening $dirname/$pod for input: $!\n"; |
652 | @poddata = <POD>; |
653 | close(POD); |
654 | |
655 | scan_items("$dirname/$pod", @poddata); |
656 | } |
657 | |
658 | # use the names of files as =item directives too. |
659 | foreach $pod (@files) { |
660 | $pod =~ /^(.*)(\.pod|\.pm)$/; |
661 | $items{$1} = "$dirname/$1.html" if $1; |
662 | } |
663 | } elsif ($pages{$libpod} =~ /([^:]*\.pod):/ || |
664 | $pages{$libpod} =~ /([^:]*\.pm):/) { |
665 | # scan the .pod or .pm file for =item directives |
666 | $pod = $1; |
667 | open(POD, "<$pod") || |
668 | die "$0: error opening $pod for input: $!\n"; |
669 | @poddata = <POD>; |
670 | close(POD); |
671 | |
672 | scan_items("$pod", @poddata); |
673 | } else { |
674 | warn "$0: shouldn't be here (line ".__LINE__."\n"; |
675 | } |
676 | } |
677 | @poddata = (); # clean-up a bit |
678 | |
679 | chdir($pwd) |
680 | || die "$0: error changing to directory $pwd: $!\n"; |
681 | |
682 | # cache the item list for later use |
683 | warn "caching items for later use\n" if $verbose; |
684 | open(CACHE, ">$itemcache") || |
685 | die "$0: error open $itemcache for writing: $!\n"; |
686 | |
687 | print CACHE join(":", @podpath) . "\n$podroot\n"; |
688 | foreach my $key (keys %items) { |
689 | print CACHE "$key $items{$key}\n"; |
690 | } |
691 | |
692 | close(CACHE); |
693 | |
694 | # cache the directory list for later use |
695 | warn "caching directories for later use\n" if $verbose; |
696 | open(CACHE, ">$dircache") || |
697 | die "$0: error open $dircache for writing: $!\n"; |
698 | |
699 | print CACHE join(":", @podpath) . "\n$podroot\n"; |
700 | foreach my $key (keys %pages) { |
701 | print CACHE "$key $pages{$key}\n"; |
702 | } |
703 | |
704 | close(CACHE); |
705 | } |
706 | |
707 | # |
708 | # scan_dir - scans the directory specified in $dir for subdirectories, .pod |
709 | # files, and .pm files. notes those that it finds. this information will |
710 | # be used later in order to figure out where the pages specified in L<> |
711 | # links are on the filesystem. |
712 | # |
713 | sub scan_dir { |
714 | my($dir, $recurse) = @_; |
715 | my($t, @subdirs, @pods, $pod, $dirname, @dirs); |
716 | local $_; |
717 | |
718 | @subdirs = (); |
719 | @pods = (); |
720 | |
721 | opendir(DIR, $dir) || |
722 | die "$0: error opening directory $dir: $!\n"; |
723 | while (defined($_ = readdir(DIR))) { |
724 | if (-d "$dir/$_" && $_ ne "." && $_ ne "..") { # directory |
725 | $pages{$_} = "" unless defined $pages{$_}; |
726 | $pages{$_} .= "$dir/$_:"; |
727 | push(@subdirs, $_); |
728 | } elsif (/\.pod$/) { # .pod |
729 | s/\.pod$//; |
730 | $pages{$_} = "" unless defined $pages{$_}; |
731 | $pages{$_} .= "$dir/$_.pod:"; |
732 | push(@pods, "$dir/$_.pod"); |
733 | } elsif (/\.pm$/) { # .pm |
734 | s/\.pm$//; |
735 | $pages{$_} = "" unless defined $pages{$_}; |
736 | $pages{$_} .= "$dir/$_.pm:"; |
737 | push(@pods, "$dir/$_.pm"); |
738 | } |
739 | } |
740 | closedir(DIR); |
741 | |
742 | # recurse on the subdirectories if necessary |
743 | if ($recurse) { |
744 | foreach my $subdir (@subdirs) { |
745 | scan_dir("$dir/$subdir", $recurse); |
746 | } |
747 | } |
748 | } |
749 | |
750 | # |
751 | # scan_headings - scan a pod file for head[1-6] tags, note the tags, and |
752 | # build an index. |
753 | # |
754 | sub scan_headings { |
755 | my($sections, @data) = @_; |
756 | my($tag, $which_head, $title, $listdepth, $index); |
757 | |
be173d55 |
758 | # here we need local $ignore = 0; |
759 | # unfortunately, we can't have it, because $ignore is lexical |
760 | $ignore = 0; |
761 | |
54310121 |
762 | $listdepth = 0; |
763 | $index = ""; |
764 | |
765 | # scan for =head directives, note their name, and build an index |
766 | # pointing to each of them. |
767 | foreach my $line (@data) { |
bb9460ed |
768 | if ($line =~ /^=(head)([1-6])\s+(.*)/) { |
54310121 |
769 | ($tag,$which_head, $title) = ($1,$2,$3); |
770 | chomp($title); |
771 | $$sections{htmlify(0,$title)} = 1; |
772 | |
773 | if ($which_head > $listdepth) { |
774 | $index .= "\n" . ("\t" x $listdepth) . "<UL>\n"; |
775 | } elsif ($which_head < $listdepth) { |
776 | $listdepth--; |
777 | $index .= "\n" . ("\t" x $listdepth) . "</UL>\n"; |
778 | } |
779 | $listdepth = $which_head; |
780 | |
781 | $index .= "\n" . ("\t" x $listdepth) . "<LI>" . |
be173d55 |
782 | "<A HREF=\"#" . htmlify(0,$title) . "\">" . |
783 | process_text(\$title, 0) . "</A>"; |
54310121 |
784 | } |
785 | } |
786 | |
787 | # finish off the lists |
788 | while ($listdepth--) { |
789 | $index .= "\n" . ("\t" x $listdepth) . "</UL>\n"; |
790 | } |
791 | |
792 | # get rid of bogus lists |
793 | $index =~ s,\t*<UL>\s*</UL>\n,,g; |
794 | |
bb9460ed |
795 | $ignore = 1; # restore old value; |
be173d55 |
796 | |
54310121 |
797 | return $index; |
798 | } |
799 | |
800 | # |
801 | # scan_items - scans the pod specified by $pod for =item directives. we |
802 | # will use this information later on in resolving C<> links. |
803 | # |
804 | sub scan_items { |
805 | my($pod, @poddata) = @_; |
806 | my($i, $item); |
807 | local $_; |
808 | |
809 | $pod =~ s/\.pod$//; |
810 | $pod .= ".html" if $pod; |
811 | |
812 | foreach $i (0..$#poddata) { |
813 | $_ = $poddata[$i]; |
814 | |
815 | # remove any formatting instructions |
816 | s,[A-Z]<([^<>]*)>,$1,g; |
817 | |
818 | # figure out what kind of item it is and get the first word of |
819 | # it's name. |
820 | if (/^=item\s+(\w*)\s*.*$/s) { |
821 | if ($1 eq "*") { # bullet list |
822 | /\A=item\s+\*\s*(.*?)\s*\Z/s; |
823 | $item = $1; |
824 | } elsif ($1 =~ /^[0-9]+/) { # numbered list |
825 | /\A=item\s+[0-9]+\.?(.*?)\s*\Z/s; |
826 | $item = $1; |
827 | } else { |
828 | # /\A=item\s+(.*?)\s*\Z/s; |
829 | /\A=item\s+(\w*)/s; |
830 | $item = $1; |
831 | } |
832 | |
833 | $items{$item} = "$pod" if $item; |
834 | } |
835 | } |
836 | } |
837 | |
838 | # |
839 | # process_head - convert a pod head[1-6] tag and convert it to HTML format. |
840 | # |
841 | sub process_head { |
842 | my($tag, $heading) = @_; |
843 | my $firstword; |
844 | |
845 | # figure out the level of the =head |
846 | $tag =~ /head([1-6])/; |
847 | my $level = $1; |
848 | |
849 | # can't have a heading full of spaces and speechmarks and so on |
850 | $firstword = $heading; $firstword =~ s/\s*(\w+)\s.*/$1/; |
851 | |
852 | print HTML "<P>\n" unless $listlevel; |
853 | print HTML "<HR>\n" unless $listlevel || $top; |
854 | print HTML "<H$level>"; # unless $listlevel; |
855 | #print HTML "<H$level>" unless $listlevel; |
be173d55 |
856 | my $convert = $heading; process_text(\$convert, 0); |
54310121 |
857 | print HTML '<A NAME="' . htmlify(0,$heading) . "\">$convert</A>"; |
858 | print HTML "</H$level>"; # unless $listlevel; |
859 | print HTML "\n"; |
860 | } |
861 | |
862 | # |
863 | # process_item - convert a pod item tag and convert it to HTML format. |
864 | # |
865 | sub process_item { |
866 | my $text = $_[0]; |
867 | my($i, $quote, $name); |
868 | |
869 | my $need_preamble = 0; |
870 | my $this_entry; |
871 | |
872 | |
873 | # lots of documents start a list without doing an =over. this is |
874 | # bad! but, the proper thing to do seems to be to just assume |
875 | # they did do an =over. so warn them once and then continue. |
876 | warn "$0: $podfile: unexpected =item directive in paragraph $paragraph. ignoring.\n" |
877 | unless $listlevel; |
878 | process_over() unless $listlevel; |
879 | |
880 | return unless $listlevel; |
881 | |
882 | # remove formatting instructions from the text |
883 | 1 while $text =~ s/[A-Z]<([^<>]*)>/$1/g; |
884 | pre_escape(\$text); |
885 | |
886 | $need_preamble = $items_seen[$listlevel]++ == 0; |
887 | |
888 | # check if this is the first =item after an =over |
889 | $i = $listlevel - 1; |
890 | my $need_new = $listlevel >= @listitem; |
891 | |
892 | if ($text =~ /\A\*/) { # bullet |
893 | |
894 | if ($need_preamble) { |
895 | push(@listend, "</UL>"); |
896 | print HTML "<UL>\n"; |
897 | } |
898 | |
899 | print HTML "<LI><STRONG>"; |
900 | $text =~ /\A\*\s*(.*)\Z/s; |
901 | print HTML "<A NAME=\"item_" . htmlify(1,$1) . "\">" if $1 && !$items_named{$1}++; |
902 | $quote = 1; |
903 | #print HTML process_puretext($1, \$quote); |
904 | print HTML $1; |
905 | print HTML "</A>" if $1; |
906 | print HTML "</STRONG>"; |
907 | |
908 | } elsif ($text =~ /\A[0-9#]+/) { # numbered list |
909 | |
910 | if ($need_preamble) { |
911 | push(@listend, "</OL>"); |
912 | print HTML "<OL>\n"; |
913 | } |
914 | |
915 | print HTML "<LI><STRONG>"; |
916 | $text =~ /\A[0-9]+\.?(.*)\Z/s; |
917 | print HTML "<A NAME=\"item_" . htmlify(0,$1) . "\">" if $1; |
918 | $quote = 1; |
919 | #print HTML process_puretext($1, \$quote); |
920 | print HTML $1 if $1; |
921 | print HTML "</A>" if $1; |
922 | print HTML "</STRONG>"; |
923 | |
924 | } else { # all others |
925 | |
926 | if ($need_preamble) { |
927 | push(@listend, '</DL>'); |
928 | print HTML "<DL>\n"; |
929 | } |
930 | |
931 | print HTML "<DT><STRONG>"; |
932 | print HTML "<A NAME=\"item_" . htmlify(1,$text) . "\">" |
933 | if $text && !$items_named{($text =~ /(\S+)/)[0]}++; |
934 | # preceding craziness so that the duplicate leading bits in |
935 | # perlfunc work to find just the first one. otherwise |
936 | # open etc would have many names |
937 | $quote = 1; |
938 | #print HTML process_puretext($text, \$quote); |
939 | print HTML $text; |
940 | print HTML "</A>" if $text; |
941 | print HTML "</STRONG>"; |
942 | |
943 | print HTML '<DD>'; |
944 | } |
945 | |
946 | print HTML "\n"; |
947 | } |
948 | |
949 | # |
950 | # process_over - process a pod over tag and start a corresponding HTML |
951 | # list. |
952 | # |
953 | sub process_over { |
954 | # start a new list |
955 | $listlevel++; |
956 | } |
957 | |
958 | # |
959 | # process_back - process a pod back tag and convert it to HTML format. |
960 | # |
961 | sub process_back { |
2ceaccd7 |
962 | warn "$0: $podfile: unexpected =back directive in paragraph $paragraph. ignoring.\n" |
54310121 |
963 | unless $listlevel; |
964 | return unless $listlevel; |
965 | |
966 | # close off the list. note, I check to see if $listend[$listlevel] is |
967 | # defined because an =item directive may have never appeared and thus |
968 | # $listend[$listlevel] may have never been initialized. |
969 | $listlevel--; |
970 | print HTML $listend[$listlevel] if defined $listend[$listlevel]; |
971 | print HTML "\n"; |
972 | |
973 | # don't need the corresponding perl code anymore |
974 | pop(@listitem); |
975 | pop(@listdata); |
976 | pop(@listend); |
977 | |
978 | pop(@items_seen); |
979 | } |
980 | |
981 | # |
982 | # process_cut - process a pod cut tag, thus stop ignoring pod directives. |
983 | # |
984 | sub process_cut { |
985 | $ignore = 1; |
986 | } |
987 | |
988 | # |
989 | # process_pod - process a pod pod tag, thus ignore pod directives until we see a |
990 | # corresponding cut. |
991 | # |
992 | sub process_pod { |
993 | # no need to set $ignore to 0 cause the main loop did it |
994 | } |
995 | |
996 | # |
997 | # process_for - process a =for pod tag. if it's for html, split |
998 | # it out verbatim, otherwise ignore it. |
999 | # |
1000 | sub process_for { |
1001 | my($whom, $text) = @_; |
1002 | if ( $whom =~ /^(pod2)?html$/i) { |
1003 | print HTML $text; |
1004 | } |
1005 | } |
1006 | |
1007 | # |
1008 | # process_begin - process a =begin pod tag. this pushes |
1009 | # whom we're beginning on the begin stack. if there's a |
1010 | # begin stack, we only print if it us. |
1011 | # |
1012 | sub process_begin { |
1013 | my($whom, $text) = @_; |
1014 | $whom = lc($whom); |
1015 | push (@begin_stack, $whom); |
1016 | if ( $whom =~ /^(pod2)?html$/) { |
1017 | print HTML $text if $text; |
1018 | } |
1019 | } |
1020 | |
1021 | # |
1022 | # process_end - process a =end pod tag. pop the |
1023 | # begin stack. die if we're mismatched. |
1024 | # |
1025 | sub process_end { |
1026 | my($whom, $text) = @_; |
1027 | $whom = lc($whom); |
1028 | if ($begin_stack[-1] ne $whom ) { |
1029 | die "Unmatched begin/end at chunk $paragraph\n" |
1030 | } |
1031 | pop @begin_stack; |
1032 | } |
1033 | |
1034 | # |
1035 | # process_text - handles plaintext that appears in the input pod file. |
1036 | # there may be pod commands embedded within the text so those must be |
1037 | # converted to html commands. |
1038 | # |
1039 | sub process_text { |
1040 | my($text, $escapeQuotes) = @_; |
1041 | my($result, $rest, $s1, $s2, $s3, $s4, $match, $bf); |
1042 | my($podcommand, $params, $tag, $quote); |
1043 | |
1044 | return if $ignore; |
1045 | |
1046 | $quote = 0; # status of double-quote conversion |
1047 | $result = ""; |
1048 | $rest = $$text; |
1049 | |
1050 | if ($rest =~ /^\s+/) { # preformatted text, no pod directives |
be173d55 |
1051 | $rest =~ s/\n+\Z//; |
1052 | $rest =~ s#.*# |
1053 | my $line = $&; |
1054 | 1 while $line =~ s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e; |
1055 | $line; |
1056 | #eg; |
54310121 |
1057 | |
1058 | $rest =~ s/&/&/g; |
1059 | $rest =~ s/</</g; |
1060 | $rest =~ s/>/>/g; |
1061 | $rest =~ s/"/"/g; |
1062 | |
1063 | # try and create links for all occurrences of perl.* within |
1064 | # the preformatted text. |
1065 | $rest =~ s{ |
1066 | (\s*)(perl\w+) |
1067 | }{ |
1068 | if (defined $pages{$2}) { # is a link |
1069 | qq($1<A HREF="$htmlroot/$pages{$2}">$2</A>); |
39e571d4 |
1070 | } elsif (defined $pages{dosify($2)}) { # is a link |
1071 | qq($1<A HREF="$htmlroot/$pages{dosify($2)}">$2</A>); |
54310121 |
1072 | } else { |
1073 | "$1$2"; |
1074 | } |
1075 | }xeg; |
1076 | $rest =~ s/(<A HREF=)([^>:]*:)?([^>:]*)\.pod:([^>:]*:)?/$1$3.html/g; |
1077 | |
1078 | my $urls = '(' . join ('|', qw{ |
1079 | http |
1080 | telnet |
1081 | mailto |
1082 | news |
1083 | gopher |
1084 | file |
1085 | wais |
1086 | ftp |
1087 | } ) |
1088 | . ')'; |
1089 | |
1090 | my $ltrs = '\w'; |
1091 | my $gunk = '/#~:.?+=&%@!\-'; |
1092 | my $punc = '.:?\-'; |
1093 | my $any = "${ltrs}${gunk}${punc}"; |
1094 | |
1095 | $rest =~ s{ |
1096 | \b # start at word boundary |
1097 | ( # begin $1 { |
1098 | $urls : # need resource and a colon |
1099 | [$any] +? # followed by on or more |
1100 | # of any valid character, but |
1101 | # be conservative and take only |
1102 | # what you need to.... |
1103 | ) # end $1 } |
1104 | (?= # look-ahead non-consumptive assertion |
1105 | [$punc]* # either 0 or more puntuation |
1106 | [^$any] # followed by a non-url char |
1107 | | # or else |
1108 | $ # then end of the string |
1109 | ) |
1110 | }{<A HREF="$1">$1</A>}igox; |
1111 | |
1112 | $result = "<PRE>" # text should be as it is (verbatim) |
1113 | . "$rest\n" |
1114 | . "</PRE>\n"; |
1115 | } else { # formatted text |
1116 | # parse through the string, stopping each time we find a |
1117 | # pod-escape. once the string has been throughly processed |
1118 | # we can output it. |
1119 | while ($rest) { |
1120 | # check to see if there are any possible pod directives in |
1121 | # the remaining part of the text. |
1122 | if ($rest =~ m/[BCEIFLSZ]</) { |
1123 | warn "\$rest\t= $rest\n" unless |
1124 | $rest =~ /\A |
1125 | ([^<]*?) |
1126 | ([BCEIFLSZ]?) |
1127 | < |
1128 | (.*)\Z/xs; |
1129 | |
1130 | $s1 = $1; # pure text |
1131 | $s2 = $2; # the type of pod-escape that follows |
1132 | $s3 = '<'; # '<' |
1133 | $s4 = $3; # the rest of the string |
1134 | } else { |
1135 | $s1 = $rest; |
1136 | $s2 = ""; |
1137 | $s3 = ""; |
1138 | $s4 = ""; |
1139 | } |
1140 | |
1141 | if ($s3 eq '<' && $s2) { # a pod-escape |
1142 | $result .= ($escapeQuotes ? process_puretext($s1, \$quote) : $s1); |
1143 | $podcommand = "$s2<"; |
1144 | $rest = $s4; |
1145 | |
1146 | # find the matching '>' |
1147 | $match = 1; |
1148 | $bf = 0; |
1149 | while ($match && !$bf) { |
1150 | $bf = 1; |
1151 | if ($rest =~ /\A([^<>]*[BCEIFLSZ]<)(.*)\Z/s) { |
1152 | $bf = 0; |
1153 | $match++; |
1154 | $podcommand .= $1; |
1155 | $rest = $2; |
1156 | } elsif ($rest =~ /\A([^>]*>)(.*)\Z/s) { |
1157 | $bf = 0; |
1158 | $match--; |
1159 | $podcommand .= $1; |
1160 | $rest = $2; |
1161 | } |
1162 | } |
1163 | |
1164 | if ($match != 0) { |
1165 | warn <<WARN; |
1166 | $0: $podfile: cannot find matching > for $s2 in paragraph $paragraph. |
1167 | WARN |
1168 | $result .= substr $podcommand, 0, 2; |
1169 | $rest = substr($podcommand, 2) . $rest; |
1170 | next; |
1171 | } |
1172 | |
1173 | # pull out the parameters to the pod-escape |
1174 | $podcommand =~ /^([BCFEILSZ]?)<(.*)>$/s; |
1175 | $tag = $1; |
1176 | $params = $2; |
1177 | |
1178 | # process the text within the pod-escape so that any escapes |
1179 | # which must occur do. |
1180 | process_text(\$params, 0) unless $tag eq 'L'; |
1181 | |
1182 | $s1 = $params; |
1183 | if (!$tag || $tag eq " ") { # <> : no tag |
1184 | $s1 = "<$params>"; |
1185 | } elsif ($tag eq "L") { # L<> : link |
1186 | $s1 = process_L($params); |
1187 | } elsif ($tag eq "I" || # I<> : italicize text |
1188 | $tag eq "B" || # B<> : bold text |
1189 | $tag eq "F") { # F<> : file specification |
1190 | $s1 = process_BFI($tag, $params); |
1191 | } elsif ($tag eq "C") { # C<> : literal code |
1192 | $s1 = process_C($params, 1); |
1193 | } elsif ($tag eq "E") { # E<> : escape |
1194 | $s1 = process_E($params); |
1195 | } elsif ($tag eq "Z") { # Z<> : zero-width character |
1196 | $s1 = process_Z($params); |
1197 | } elsif ($tag eq "S") { # S<> : non-breaking space |
1198 | $s1 = process_S($params); |
1199 | } elsif ($tag eq "X") { # S<> : non-breaking space |
1200 | $s1 = process_X($params); |
1201 | } else { |
1202 | warn "$0: $podfile: unhandled tag '$tag' in paragraph $paragraph\n"; |
1203 | } |
1204 | |
1205 | $result .= "$s1"; |
1206 | } else { |
1207 | # for pure text we must deal with implicit links and |
1208 | # double-quotes among other things. |
1209 | $result .= ($escapeQuotes ? process_puretext("$s1$s2$s3", \$quote) : "$s1$s2$s3"); |
1210 | $rest = $s4; |
1211 | } |
1212 | } |
1213 | } |
1214 | $$text = $result; |
1215 | } |
1216 | |
1217 | sub html_escape { |
1218 | my $rest = $_[0]; |
1219 | $rest =~ s/&/&/g; |
1220 | $rest =~ s/</</g; |
1221 | $rest =~ s/>/>/g; |
1222 | $rest =~ s/"/"/g; |
1223 | return $rest; |
1224 | } |
1225 | |
1226 | # |
1227 | # process_puretext - process pure text (without pod-escapes) converting |
1228 | # double-quotes and handling implicit C<> links. |
1229 | # |
1230 | sub process_puretext { |
1231 | my($text, $quote) = @_; |
1232 | my(@words, $result, $rest, $lead, $trail); |
1233 | |
1234 | # convert double-quotes to single-quotes |
1235 | $text =~ s/\A([^"]*)"/$1''/s if $$quote; |
1236 | while ($text =~ s/\A([^"]*)["]([^"]*)["]/$1``$2''/sg) {} |
1237 | |
1238 | $$quote = ($text =~ m/"/ ? 1 : 0); |
1239 | $text =~ s/\A([^"]*)"/$1``/s if $$quote; |
1240 | |
1241 | # keep track of leading and trailing white-space |
1242 | $lead = ($text =~ /\A(\s*)/s ? $1 : ""); |
1243 | $trail = ($text =~ /(\s*)\Z/s ? $1 : ""); |
1244 | |
1245 | # collapse all white space into a single space |
1246 | $text =~ s/\s+/ /g; |
1247 | @words = split(" ", $text); |
1248 | |
1249 | # process each word individually |
1250 | foreach my $word (@words) { |
1251 | # see if we can infer a link |
1252 | if ($word =~ /^\w+\(/) { |
1253 | # has parenthesis so should have been a C<> ref |
1254 | $word = process_C($word); |
1255 | # $word =~ /^[^()]*]\(/; |
1256 | # if (defined $items{$1} && $items{$1}) { |
1257 | # $word = "\n<CODE><A HREF=\"$htmlroot/$items{$1}#item_" |
1258 | # . htmlify(0,$word) |
1259 | # . "\">$word</A></CODE>"; |
1260 | # } elsif (defined $items{$word} && $items{$word}) { |
1261 | # $word = "\n<CODE><A HREF=\"$htmlroot/$items{$word}#item_" |
1262 | # . htmlify(0,$word) |
1263 | # . "\">$word</A></CODE>"; |
1264 | # } else { |
1265 | # $word = "\n<CODE><A HREF=\"#item_" |
1266 | # . htmlify(0,$word) |
1267 | # . "\">$word</A></CODE>"; |
1268 | # } |
1269 | } elsif ($word =~ /^[\$\@%&*]+\w+$/) { |
1270 | # perl variables, should be a C<> ref |
1271 | $word = process_C($word, 1); |
1272 | } elsif ($word =~ m,^\w+://\w,) { |
1273 | # looks like a URL |
1274 | $word = qq(<A HREF="$word">$word</A>); |
1275 | } elsif ($word =~ /[\w.-]+\@\w+\.\w/) { |
1276 | # looks like an e-mail address |
1277 | $word = qq(<A HREF="MAILTO:$word">$word</A>); |
1278 | } elsif ($word !~ /[a-z]/ && $word =~ /[A-Z]/) { # all uppercase? |
1279 | $word = html_escape($word) if $word =~ /[&<>]/; |
1280 | $word = "\n<FONT SIZE=-1>$word</FONT>" if $netscape; |
1281 | } else { |
1282 | $word = html_escape($word) if $word =~ /[&<>]/; |
1283 | } |
1284 | } |
1285 | |
1286 | # build a new string based upon our conversion |
1287 | $result = ""; |
1288 | $rest = join(" ", @words); |
1289 | while (length($rest) > 75) { |
1290 | if ( $rest =~ m/^(.{0,75})\s(.*?)$/o || |
1291 | $rest =~ m/^(\S*)\s(.*?)$/o) { |
1292 | |
1293 | $result .= "$1\n"; |
1294 | $rest = $2; |
1295 | } else { |
1296 | $result .= "$rest\n"; |
1297 | $rest = ""; |
1298 | } |
1299 | } |
1300 | $result .= $rest if $rest; |
1301 | |
1302 | # restore the leading and trailing white-space |
1303 | $result = "$lead$result$trail"; |
1304 | |
1305 | return $result; |
1306 | } |
1307 | |
1308 | # |
1309 | # pre_escape - convert & in text to $amp; |
1310 | # |
1311 | sub pre_escape { |
1312 | my($str) = @_; |
1313 | |
1314 | $$str =~ s,&,&,g; |
1315 | } |
1316 | |
1317 | # |
39e571d4 |
1318 | # dosify - convert filenames to 8.3 |
1319 | # |
1320 | sub dosify { |
1321 | my($str) = @_; |
1322 | if ($Is83) { |
1323 | $str = lc $str; |
1324 | $str =~ s/(\.\w+)/substr ($1,0,4)/ge; |
1325 | $str =~ s/(\w+)/substr ($1,0,8)/ge; |
1326 | } |
1327 | return $str; |
1328 | } |
1329 | |
1330 | # |
54310121 |
1331 | # process_L - convert a pod L<> directive to a corresponding HTML link. |
1332 | # most of the links made are inferred rather than known about directly |
1333 | # (i.e it's not known whether the =head\d section exists in the target file, |
1334 | # or whether a .pod file exists in the case of split files). however, the |
1335 | # guessing usually works. |
1336 | # |
1337 | # Unlike the other directives, this should be called with an unprocessed |
1338 | # string, else tags in the link won't be matched. |
1339 | # |
1340 | sub process_L { |
1341 | my($str) = @_; |
39e571d4 |
1342 | my($s1, $s2, $linktext, $page, $page83, $section, $link); # work strings |
54310121 |
1343 | |
1344 | $str =~ s/\n/ /g; # undo word-wrapped tags |
1345 | $s1 = $str; |
1346 | for ($s1) { |
b74bceb9 |
1347 | # LREF: a la HREF L<show this text|man/section> |
1348 | $linktext = $1 if s:^([^|]+)\|::; |
1349 | |
54310121 |
1350 | # a :: acts like a / |
1351 | s,::,/,; |
1352 | |
1353 | # make sure sections start with a / |
1354 | s,^",/",g; |
1355 | s,^,/,g if (!m,/, && / /); |
1356 | |
1357 | # check if there's a section specified |
1358 | if (m,^(.*?)/"?(.*?)"?$,) { # yes |
1359 | ($page, $section) = ($1, $2); |
1360 | } else { # no |
1361 | ($page, $section) = ($str, ""); |
1362 | } |
1363 | |
1364 | # check if we know that this is a section in this page |
1365 | if (!defined $pages{$page} && defined $sections{$page}) { |
1366 | $section = $page; |
1367 | $page = ""; |
1368 | } |
1369 | } |
1370 | |
39e571d4 |
1371 | $page83=dosify($page); |
1372 | $page=$page83 if (defined $pages{$page83}); |
54310121 |
1373 | if ($page eq "") { |
1374 | $link = "#" . htmlify(0,$section); |
b74bceb9 |
1375 | $linktext = $section unless defined($linktext); |
54310121 |
1376 | } elsif (!defined $pages{$page}) { |
1377 | warn "$0: $podfile: cannot resolve L<$str> in paragraph $paragraph: no such page '$page'\n"; |
1378 | $link = ""; |
b74bceb9 |
1379 | $linktext = $page unless defined($linktext); |
54310121 |
1380 | } else { |
b74bceb9 |
1381 | $linktext = ($section ? "$section" : "the $page manpage") unless defined($linktext); |
54310121 |
1382 | $section = htmlify(0,$section) if $section ne ""; |
1383 | |
1384 | # if there is a directory by the name of the page, then assume that an |
1385 | # appropriate section will exist in the subdirectory |
1386 | if ($section ne "" && $pages{$page} =~ /([^:]*[^(\.pod|\.pm)]):/) { |
1387 | $link = "$htmlroot/$1/$section.html"; |
1388 | |
1389 | # since there is no directory by the name of the page, the section will |
1390 | # have to exist within a .html of the same name. thus, make sure there |
1391 | # is a .pod or .pm that might become that .html |
1392 | } else { |
1393 | $section = "#$section"; |
1394 | # check if there is a .pod with the page name |
1395 | if ($pages{$page} =~ /([^:]*)\.pod:/) { |
1396 | $link = "$htmlroot/$1.html$section"; |
1397 | } elsif ($pages{$page} =~ /([^:]*)\.pm:/) { |
1398 | $link = "$htmlroot/$1.html$section"; |
1399 | } else { |
1400 | warn "$0: $podfile: cannot resolve L$str in paragraph $paragraph: ". |
1401 | "no .pod or .pm found\n"; |
1402 | $link = ""; |
b74bceb9 |
1403 | $linktext = $section unless defined($linktext); |
54310121 |
1404 | } |
1405 | } |
1406 | } |
1407 | |
1408 | process_text(\$linktext, 0); |
1409 | if ($link) { |
1410 | $s1 = "<A HREF=\"$link\">$linktext</A>"; |
1411 | } else { |
1412 | $s1 = "<EM>$linktext</EM>"; |
1413 | } |
1414 | return $s1; |
1415 | } |
1416 | |
1417 | # |
1418 | # process_BFI - process any of the B<>, F<>, or I<> pod-escapes and |
1419 | # convert them to corresponding HTML directives. |
1420 | # |
1421 | sub process_BFI { |
1422 | my($tag, $str) = @_; |
1423 | my($s1); # work string |
1424 | my(%repltext) = ( 'B' => 'STRONG', |
1425 | 'F' => 'EM', |
1426 | 'I' => 'EM'); |
1427 | |
1428 | # extract the modified text and convert to HTML |
1429 | $s1 = "<$repltext{$tag}>$str</$repltext{$tag}>"; |
1430 | return $s1; |
1431 | } |
1432 | |
1433 | # |
1434 | # process_C - process the C<> pod-escape. |
1435 | # |
1436 | sub process_C { |
1437 | my($str, $doref) = @_; |
1438 | my($s1, $s2); |
1439 | |
1440 | $s1 = $str; |
1441 | $s1 =~ s/\([^()]*\)//g; # delete parentheses |
be173d55 |
1442 | $s2 = $s1; |
54310121 |
1443 | $s1 =~ s/\W//g; # delete bogus characters |
1444 | |
1445 | # if there was a pod file that we found earlier with an appropriate |
1446 | # =item directive, then create a link to that page. |
1447 | if ($doref && defined $items{$s1}) { |
1448 | $s1 = ($items{$s1} ? |
1449 | "<A HREF=\"$htmlroot/$items{$s1}#item_" . htmlify(0,$s2) . "\">$str</A>" : |
1450 | "<A HREF=\"#item_" . htmlify(0,$s2) . "\">$str</A>"); |
1451 | $s1 =~ s,(perl\w+/(\S+)\.html)#item_\2\b,$1,; |
1452 | confess "s1 has space: $s1" if $s1 =~ /HREF="[^"]*\s[^"]*"/; |
1453 | } else { |
1454 | $s1 = "<CODE>$str</CODE>"; |
1455 | # warn "$0: $podfile: cannot resolve C<$str> in paragraph $paragraph\n" if $verbose |
1456 | } |
1457 | |
1458 | |
1459 | return $s1; |
1460 | } |
1461 | |
1462 | # |
1463 | # process_E - process the E<> pod directive which seems to escape a character. |
1464 | # |
1465 | sub process_E { |
1466 | my($str) = @_; |
1467 | |
1468 | for ($str) { |
1469 | s,([^/].*),\&$1\;,g; |
1470 | } |
1471 | |
1472 | return $str; |
1473 | } |
1474 | |
1475 | # |
1476 | # process_Z - process the Z<> pod directive which really just amounts to |
1477 | # ignoring it. this allows someone to start a paragraph with an = |
1478 | # |
1479 | sub process_Z { |
1480 | my($str) = @_; |
1481 | |
1482 | # there is no equivalent in HTML for this so just ignore it. |
1483 | $str = ""; |
1484 | return $str; |
1485 | } |
1486 | |
1487 | # |
1488 | # process_S - process the S<> pod directive which means to convert all |
1489 | # spaces in the string to non-breaking spaces (in HTML-eze). |
1490 | # |
1491 | sub process_S { |
1492 | my($str) = @_; |
1493 | |
1494 | # convert all spaces in the text to non-breaking spaces in HTML. |
1495 | $str =~ s/ / /g; |
1496 | return $str; |
1497 | } |
1498 | |
1499 | # |
1500 | # process_X - this is supposed to make an index entry. we'll just |
1501 | # ignore it. |
1502 | # |
1503 | sub process_X { |
1504 | return ''; |
1505 | } |
1506 | |
1507 | |
1508 | # |
1509 | # finish_list - finish off any pending HTML lists. this should be called |
1510 | # after the entire pod file has been read and converted. |
1511 | # |
1512 | sub finish_list { |
1513 | while ($listlevel >= 0) { |
1514 | print HTML "</DL>\n"; |
1515 | $listlevel--; |
1516 | } |
1517 | } |
1518 | |
1519 | # |
1520 | # htmlify - converts a pod section specification to a suitable section |
1521 | # specification for HTML. if first arg is 1, only takes 1st word. |
1522 | # |
1523 | sub htmlify { |
1524 | my($compact, $heading) = @_; |
1525 | |
1526 | if ($compact) { |
1527 | $heading =~ /^(\w+)/; |
1528 | $heading = $1; |
1529 | } |
1530 | |
1531 | # $heading = lc($heading); |
1532 | $heading =~ s/[^\w\s]/_/g; |
1533 | $heading =~ s/(\s+)/ /g; |
1534 | $heading =~ s/^\s*(.*?)\s*$/$1/s; |
1535 | $heading =~ s/ /_/g; |
1536 | $heading =~ s/\A(.{32}).*\Z/$1/s; |
1537 | $heading =~ s/\s+\Z//; |
1538 | $heading =~ s/_{2,}/_/g; |
1539 | |
1540 | return $heading; |
1541 | } |
1542 | |
1543 | BEGIN { |
1544 | } |
1545 | |
1546 | 1; |
1547 | |