5 use vars qw($VERSION @ISA @EXPORT);
8 @EXPORT = qw(pod2html htmlify);
17 use locale; # make \w work right in non-ASCII lands
21 Pod::Html - module to convert pod files to HTML
30 Converts files from pod format (see L<perlpod>) to HTML format. It
31 can automatically generate indexes and cross-references, and it keeps
32 a cache of things it knows how to cross-reference.
36 Pod::Html takes the following arguments:
42 --backlink="Back to Top"
44 Adds "Back to Top" links in front of every C<head1> heading (except for
45 the first). By default, no backlinks are generated.
51 Creates the item and directory caches in the given directory.
57 Specify the URL of a cascading style sheet. Also disables all HTML/CSS
58 C<style> attributes that are output by default (to avoid conflicts).
64 Flushes the item and directory caches.
71 Creates header and footer blocks containing the text of the C<NAME>
72 section. By default, no headers are generated.
78 Displays the usage message.
84 Sets the directory in which the resulting HTML file is placed. This
85 is used to generate relative links to other files. Not passing this
86 causes all links to be absolute, since this is the value that tells
87 Pod::Html the root of the documentation tree.
93 Sets the base URL for the HTML files. When cross-references are made,
94 the HTML root is prepended to the URL.
101 Generate an index at the top of the HTML file. This is the default
108 Specify the pod file to convert. Input is taken from STDIN if no
113 --libpods=name:...:name
115 List of page names (eg, "perlfunc") which contain linkable C<=item>s.
122 B<Deprecated>, has no effect. For backwards compatibility only.
128 Specify the HTML file to create. Output goes to STDOUT if no outfile
133 --podpath=name:...:name
135 Specify which subdirectories of the podroot contain pod files whose
136 HTML converted forms can be linked to in cross references.
142 Specify the base directory for finding library pods.
149 Don't display I<mostly harmless> warning messages. These messages
150 will be displayed by default. But this is not the same as C<verbose>
158 Recurse into subdirectories specified in podpath (default behaviour).
164 Specify the title of the resulting HTML file.
171 Display progress messages. By default, they won't be displayed.
178 "--podpath=lib:ext:pod:vms",
179 "--podroot=/usr/src/perl",
180 "--htmlroot=/perl/nmanual",
181 "--libpods=perlfunc:perlguts:perlvar:perlrun:perlop",
184 "--outfile=/perl/nmanual/foo.html");
188 Uses C<$Config{pod2html}> to setup default options.
192 Tom Christiansen, E<lt>tchrist@perl.comE<gt>.
200 This program is distributed under the Artistic License.
204 my $cachedir = "."; # The directory to which item and directory
205 # caches will be written.
206 my $cache_ext = $^O eq 'VMS' ? ".tmp" : ".x~~";
207 my $dircache = "pod2htmd$cache_ext";
208 my $itemcache = "pod2htmi$cache_ext";
210 my @begin_stack = (); # begin/end stack
212 my @libpods = (); # files to search for links from C<> directives
213 my $htmlroot = "/"; # http-server base directory from which all
214 # relative paths in $podpath stem.
215 my $htmldir = ""; # The directory to which the html pages
216 # will (eventually) be written.
217 my $htmlfile = ""; # write to stdout by default
218 my $htmlfileurl = "" ; # The url that other files would use to
219 # refer to this file. This is only used
220 # to make relative urls that point to
222 my $podfile = ""; # read from stdin by default
223 my @podpath = (); # list of directories containing library pods.
224 my $podroot = File::Spec->curdir; # filesystem base directory from which all
225 # relative paths in $podpath stem.
226 my $css = ''; # Cascading style sheet
227 my $recurse = 1; # recurse on subdirectories in $podpath.
228 my $quiet = 0; # not quiet by default
229 my $verbose = 0; # not verbose by default
230 my $doindex = 1; # non-zero if we should generate an index
231 my $backlink = ''; # text for "back to top" links
232 my $listlevel = 0; # current list depth
233 my @listend = (); # the text to use to end the list.
234 my $after_lpar = 0; # set to true after a par in an =item
235 my $ignore = 1; # whether or not to format text. we don't
236 # format text until we hit our first pod
239 my %items_named = (); # for the multiples of the same item in perlfunc
241 my $title; # title to give the pod(s)
242 my $header = 0; # produce block header/footer
243 my $top = 1; # true if we are at the top of the doc. used
244 # to prevent the first <hr /> directive.
245 my $paragraph; # which paragraph we're processing (used
246 # for error messages)
247 my $ptQuote = 0; # status of double-quote conversion
248 my %pages = (); # associative array used to find the location
249 # of pages referenced by L<> links.
250 my %sections = (); # sections within this page
251 my %items = (); # associative array used to find the location
252 # of =item directives referenced by C<> links
253 my %local_items = (); # local items - avoid destruction of %items
254 my $Is83; # is dos with short filenames (8.3)
257 $dircache = "pod2htmd$cache_ext";
258 $itemcache = "pod2htmi$cache_ext";
260 @begin_stack = (); # begin/end stack
262 @libpods = (); # files to search for links from C<> directives
263 $htmlroot = "/"; # http-server base directory from which all
264 # relative paths in $podpath stem.
265 $htmldir = ""; # The directory to which the html pages
266 # will (eventually) be written.
267 $htmlfile = ""; # write to stdout by default
268 $podfile = ""; # read from stdin by default
269 @podpath = (); # list of directories containing library pods.
270 $podroot = File::Spec->curdir; # filesystem base directory from which all
271 # relative paths in $podpath stem.
272 $css = ''; # Cascading style sheet
273 $recurse = 1; # recurse on subdirectories in $podpath.
274 $quiet = 0; # not quiet by default
275 $verbose = 0; # not verbose by default
276 $doindex = 1; # non-zero if we should generate an index
277 $backlink = ''; # text for "back to top" links
278 $listlevel = 0; # current list depth
279 @listend = (); # the text to use to end the list.
280 $after_lpar = 0; # set to true after a par in an =item
281 $ignore = 1; # whether or not to format text. we don't
282 # format text until we hit our first pod
287 $header = 0; # produce block header/footer
288 $title = ''; # title to give the pod(s)
289 $top = 1; # true if we are at the top of the doc. used
290 # to prevent the first <hr /> directive.
291 $paragraph = ''; # which paragraph we're processing (used
292 # for error messages)
293 %sections = (); # sections within this page
295 # These are not reinitialised here but are kept as a cache.
296 # See get_cache and related cache management code.
297 #%pages = (); # associative array used to find the location
298 # of pages referenced by L<> links.
299 #%items = (); # associative array used to find the location
300 # of =item directives referenced by C<> links
306 # clean_data: global clean-up of pod data
310 for my $i ( 0..$#{$dataref} ) {
311 ${$dataref}[$i] =~ s/\s+\Z//;
313 # have a look for all-space lines
314 if( ${$dataref}[$i] =~ /^\s+$/m and $dataref->[$i] !~ /^\s/ ){
315 my @chunks = split( /^\s+$/m, ${$dataref}[$i] );
316 splice( @$dataref, $i, 1, @chunks );
329 $Is83 = 0 if (defined (&Dos::UseLFN) && Dos::UseLFN());
331 # cache of %pages and %items from last time we ran pod2html
333 #undef $opt_help if defined $opt_help;
335 # parse the command-line parameters
336 parse_command_line();
338 # escape the backlink argument (same goes for title but is done later...)
339 $backlink = html_escape($backlink) if defined $backlink;
341 # set some variables to their default values if necessary
343 unless (@ARGV && $ARGV[0]) {
344 $podfile = "-" unless $podfile; # stdin
345 open(POD, "<$podfile")
346 || die "$0: cannot open $podfile file for input: $!\n";
348 $podfile = $ARGV[0]; # XXX: might be more filenames
351 $htmlfile = "-" unless $htmlfile; # stdout
352 $htmlroot = "" if $htmlroot eq "/"; # so we don't get a //
353 $htmldir =~ s#/\z## ; # so we don't get a //
355 && defined( $htmldir )
357 && substr( $htmlfile, 0, length( $htmldir ) ) eq $htmldir
360 # Set the 'base' url for this file, so that we can use it
361 # as the location from which to calculate relative links
362 # to other files. If this is '', then absolute links will
363 # be used throughout.
364 $htmlfileurl= "$htmldir/" . substr( $htmlfile, length( $htmldir ) + 1);
367 # read the pod a paragraph at a time
368 warn "Scanning for sections in input file(s)\n" if $verbose;
372 clean_data( \@poddata );
374 # scan the pod for =head[1-6] directives and build an index
375 my $index = scan_headings(\%sections, @poddata);
378 warn "No headings in $podfile\n" if $verbose;
381 # open the output file
382 open(HTML, ">$htmlfile")
383 || die "$0: cannot open $htmlfile file for output: $!\n";
385 # put a title in the HTML file if one wasn't specified
388 for (my $i = 0; $i < @poddata; $i++) {
389 if ($poddata[$i] =~ /^=head1\s*NAME\b/m) {
390 for my $para ( @poddata[$i, $i+1] ) {
392 if ($title) = $para =~ /(\S+\s+-+.*\S)/s;
399 if (!$title and $podfile =~ /\.pod\z/) {
400 # probably a split pod so take first =head[12] as title
401 for (my $i = 0; $i < @poddata; $i++) {
402 last if ($title) = $poddata[$i] =~ /^=head[12]\s*(.*)/;
404 warn "adopted '$title' as title for $podfile\n"
405 if $verbose and $title;
408 $title =~ s/\s*\(.*\)//;
410 warn "$0: no title for $podfile" unless $quiet;
411 $podfile =~ /^(.*)(\.[^.\/]+)?\z/s;
412 $title = ($podfile eq "-" ? 'No Title' : $1);
413 warn "using $title" if $verbose;
415 $title = html_escape($title);
418 my $bodystyle = ' style="background-color: white"';
419 my $tdstyle = ' style="background-color: #cccccc"';
422 $csslink = qq(\n<link rel="stylesheet" href="$css" type="text/css" />);
423 $csslink =~ s,\\,/,g;
424 $csslink =~ s,(/.):,$1|,;
429 my $block = $header ? <<END_OF_BLOCK : '';
430 <table border="0" width="100%" cellspacing="0" cellpadding="3">
431 <tr><td class="block"$tdstyle valign="middle">
432 <big><strong><span class="block"> $title</span></strong></big>
437 print HTML <<END_OF_HEAD;
438 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
439 <html xmlns="http://www.w3.org/1999/xhtml">
441 <title>$title</title>$csslink
442 <link rev="made" href="mailto:$Config{perladmin}" />
449 # load/reload/validate/cache %pages and %items
450 get_cache($dircache, $itemcache, \@podpath, $podroot, $recurse);
452 # scan the pod for =item directives
453 scan_items( \%local_items, "", @poddata);
455 # put an index at the top of the file. note, if $doindex is 0 we
456 # still generate an index, but surround it with an html comment.
457 # that way some other program can extract it if desired.
459 print HTML "<p><a name=\"__index__\"></a></p>\n";
460 print HTML "<!-- INDEX BEGIN -->\n";
461 print HTML "<!--\n" unless $doindex;
463 print HTML "-->\n" unless $doindex;
464 print HTML "<!-- INDEX END -->\n\n";
465 print HTML "<hr />\n" if $doindex and $index;
467 # now convert this file
468 my $after_item; # set to true after an =item
470 warn "Converting input file $podfile\n" if $verbose;
471 foreach my $i (0..$#poddata){
472 $ptQuote = 0; # status of quote conversion
476 if (/^(=.*)/s) { # is it a pod directive?
481 if (/^=begin\s+(\S+)\s*(.*)/si) {# =begin
482 process_begin($1, $2);
483 } elsif (/^=end\s+(\S+)\s*(.*)/si) {# =end
485 } elsif (/^=cut/) { # =cut
487 } elsif (/^=pod/) { # =pod
490 next if @begin_stack && $begin_stack[-1] ne 'html';
492 if (/^=(head[1-6])\s+(.*\S)/s) { # =head[1-6] heading
493 process_head( $1, $2, $doindex && $index );
494 } elsif (/^=item\s*(.*\S)?/sm) { # =item text
495 warn "$0: $podfile: =item without bullet, number or text"
496 . " in paragraph $paragraph.\n" if !defined($1) or $1 eq '';
497 $need_dd = process_item( $1 );
499 } elsif (/^=over\s*(.*)/) { # =over N
501 } elsif (/^=back/) { # =back
503 } elsif (/^=for\s+(\S+)\s*(.*)/si) {# =for
507 warn "$0: $podfile: unknown pod directive '$1' in "
508 . "paragraph $paragraph. ignoring.\n";
515 next if @begin_stack && $begin_stack[-1] ne 'html';
516 print HTML and next if @begin_stack && $begin_stack[-1] eq 'html';
517 print HTML "<dd>\n" if $need_dd;
519 if( $text =~ /\A\s+/ ){
520 process_pre( \$text );
521 print HTML "<pre>\n$text</pre>\n";
524 process_text( \$text );
526 # experimental: check for a paragraph where all lines
527 # have some ...\t...\t...\n pattern
529 my @lines = split( "\n", $text );
532 foreach my $line ( @lines ){
533 if( $line =~ /\S/ && $line !~ /\t/ ){
539 $text =~ s/\t+/<td>/g;
540 $text =~ s/^/<tr><td>/gm;
541 $text = '<table cellspacing="0" cellpadding="0">' .
546 ## end of experimental
549 print HTML "$text\n";
552 print HTML "<p>$text</p>\n";
555 print HTML "</dd>\n" if $need_dd;
560 # finish off any pending directives
564 print HTML "<p><a href=\"#__index__\"><small>$backlink</small></a></p>\n"
565 if $doindex and $index and $backlink;
567 print HTML <<END_OF_TAIL;
574 # close the html file
577 warn "Finished\n" if $verbose;
580 ##############################################################################
582 my $usage; # see below
585 warn "$0: $podfile: @_\n" if @_;
589 $usage =<<END_OF_USAGE;
590 Usage: $0 --help --htmlroot=<name> --infile=<name> --outfile=<name>
591 --podpath=<name>:...:<name> --podroot=<name>
592 --libpods=<name>:...:<name> --recurse --verbose --index
593 --netscape --norecurse --noindex --cachedir=<name>
595 --backlink - set text for "back to top" links (default: none).
596 --cachedir - directory for the item and directory cache files.
597 --css - stylesheet URL
598 --flush - flushes the item and directory caches.
599 --[no]header - produce block header/footer (default is no headers).
600 --help - prints this message.
601 --htmldir - directory for resulting HTML files.
602 --htmlroot - http-server base directory from which all relative paths
603 in podpath stem (default is /).
604 --[no]index - generate an index at the top of the resulting html
606 --infile - filename for the pod to convert (input taken from stdin
608 --libpods - colon-separated list of pages to search for =item pod
609 directives in as targets of C<> and implicit links (empty
610 by default). note, these are not filenames, but rather
611 page names like those that appear in L<> links.
612 --outfile - filename for the resulting html file (output sent to
614 --podpath - colon-separated list of directories containing library
615 pods (empty by default).
616 --podroot - filesystem base directory from which all relative paths
617 in podpath stem (default is .).
618 --[no]quiet - supress some benign warning messages (default is off).
619 --[no]recurse - recurse on those subdirectories listed in podpath
621 --title - title that will appear in resulting html file.
622 --[no]verbose - self-explanatory (off by default).
623 --[no]netscape - deprecated, has no effect. for backwards compatibility only.
627 sub parse_command_line {
628 my ($opt_backlink,$opt_cachedir,$opt_css,$opt_flush,$opt_header,$opt_help,
629 $opt_htmldir,$opt_htmlroot,$opt_index,$opt_infile,$opt_libpods,
630 $opt_netscape,$opt_outfile,$opt_podpath,$opt_podroot,$opt_quiet,
631 $opt_recurse,$opt_title,$opt_verbose);
633 unshift @ARGV, split ' ', $Config{pod2html} if $Config{pod2html};
634 my $result = GetOptions(
635 'backlink=s' => \$opt_backlink,
636 'cachedir=s' => \$opt_cachedir,
637 'css=s' => \$opt_css,
638 'flush' => \$opt_flush,
639 'header!' => \$opt_header,
640 'help' => \$opt_help,
641 'htmldir=s' => \$opt_htmldir,
642 'htmlroot=s' => \$opt_htmlroot,
643 'index!' => \$opt_index,
644 'infile=s' => \$opt_infile,
645 'libpods=s' => \$opt_libpods,
646 'netscape!' => \$opt_netscape,
647 'outfile=s' => \$opt_outfile,
648 'podpath=s' => \$opt_podpath,
649 'podroot=s' => \$opt_podroot,
650 'quiet!' => \$opt_quiet,
651 'recurse!' => \$opt_recurse,
652 'title=s' => \$opt_title,
653 'verbose!' => \$opt_verbose,
655 usage("-", "invalid parameters") if not $result;
657 usage("-") if defined $opt_help; # see if the user asked for help
658 $opt_help = ""; # just to make -w shut-up.
660 @podpath = split(":", $opt_podpath) if defined $opt_podpath;
661 @libpods = split(":", $opt_libpods) if defined $opt_libpods;
663 $backlink = $opt_backlink if defined $opt_backlink;
664 $cachedir = $opt_cachedir if defined $opt_cachedir;
665 $css = $opt_css if defined $opt_css;
666 $header = $opt_header if defined $opt_header;
667 $htmldir = $opt_htmldir if defined $opt_htmldir;
668 $htmlroot = $opt_htmlroot if defined $opt_htmlroot;
669 $doindex = $opt_index if defined $opt_index;
670 $podfile = $opt_infile if defined $opt_infile;
671 $htmlfile = $opt_outfile if defined $opt_outfile;
672 $podroot = $opt_podroot if defined $opt_podroot;
673 $quiet = $opt_quiet if defined $opt_quiet;
674 $recurse = $opt_recurse if defined $opt_recurse;
675 $title = $opt_title if defined $opt_title;
676 $verbose = $opt_verbose if defined $opt_verbose;
678 warn "Flushing item and directory caches\n"
679 if $opt_verbose && defined $opt_flush;
680 $dircache = "$cachedir/pod2htmd$cache_ext";
681 $itemcache = "$cachedir/pod2htmi$cache_ext";
682 unlink($dircache, $itemcache) if defined $opt_flush;
689 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
690 my @cache_key_args = @_;
692 # A first-level cache:
693 # Don't bother reading the cache files if they still apply
694 # and haven't changed since we last read them.
696 my $this_cache_key = cache_key(@cache_key_args);
698 return if $saved_cache_key and $this_cache_key eq $saved_cache_key;
700 # load the cache of %pages and %items if possible. $tests will be
701 # non-zero if successful.
703 if (-f $dircache && -f $itemcache) {
704 warn "scanning for item cache\n" if $verbose;
705 $tests = load_cache($dircache, $itemcache, $podpath, $podroot);
708 # if we didn't succeed in loading the cache then we must (re)build
711 warn "scanning directories in pod-path\n" if $verbose;
712 scan_podpath($podroot, $recurse, 0);
714 $saved_cache_key = cache_key(@cache_key_args);
718 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
719 return join('!', $dircache, $itemcache, $recurse,
720 @$podpath, $podroot, stat($dircache), stat($itemcache));
724 # load_cache - tries to find if the caches stored in $dircache and $itemcache
725 # are valid caches of %pages and %items. if they are valid then it loads
726 # them and returns a non-zero value.
729 my($dircache, $itemcache, $podpath, $podroot) = @_;
735 open(CACHE, "<$itemcache") ||
736 die "$0: error opening $itemcache for reading: $!\n";
739 # is it the same podpath?
742 $tests++ if (join(":", @$podpath) eq $_);
744 # is it the same podroot?
747 $tests++ if ($podroot eq $_);
749 # load the cache if its good
755 warn "loading item cache\n" if $verbose;
762 warn "scanning for directory cache\n" if $verbose;
763 open(CACHE, "<$dircache") ||
764 die "$0: error opening $dircache for reading: $!\n";
768 # is it the same podpath?
771 $tests++ if (join(":", @$podpath) eq $_);
773 # is it the same podroot?
776 $tests++ if ($podroot eq $_);
778 # load the cache if its good
784 warn "loading directory cache\n" if $verbose;
796 # scan_podpath - scans the directories specified in @podpath for directories,
797 # .pod files, and .pm files. it also scans the pod files specified in
798 # @libpods for =item directives.
801 my($podroot, $recurse, $append) = @_;
803 my($libpod, $dirname, $pod, @files, @poddata);
810 # scan each directory listed in @podpath
813 || die "$0: error changing to directory $podroot: $!\n";
814 foreach $dir (@podpath) {
815 scan_dir($dir, $recurse);
818 # scan the pods listed in @libpods for =item directives
819 foreach $libpod (@libpods) {
820 # if the page isn't defined then we won't know where to find it
822 next unless defined $pages{$libpod} && $pages{$libpod};
824 # if there is a directory then use the .pod and .pm files within it.
825 # NOTE: Only finds the first so-named directory in the tree.
826 # if ($pages{$libpod} =~ /([^:]*[^(\.pod|\.pm)]):/) {
827 if ($pages{$libpod} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
828 # find all the .pod and .pm files within the directory
830 opendir(DIR, $dirname) ||
831 die "$0: error opening directory $dirname: $!\n";
832 @files = grep(/(\.pod|\.pm)\z/ && ! -d $_, readdir(DIR));
835 # scan each .pod and .pm file for =item directives
836 foreach $pod (@files) {
837 open(POD, "<$dirname/$pod") ||
838 die "$0: error opening $dirname/$pod for input: $!\n";
841 clean_data( \@poddata );
843 scan_items( \%items, "$dirname/$pod", @poddata);
846 # use the names of files as =item directives too.
847 ### Don't think this should be done this way - confuses issues.(WL)
848 ### foreach $pod (@files) {
849 ### $pod =~ /^(.*)(\.pod|\.pm)$/;
850 ### $items{$1} = "$dirname/$1.html" if $1;
852 } elsif ($pages{$libpod} =~ /([^:]*\.pod):/ ||
853 $pages{$libpod} =~ /([^:]*\.pm):/) {
854 # scan the .pod or .pm file for =item directives
856 open(POD, "<$pod") ||
857 die "$0: error opening $pod for input: $!\n";
860 clean_data( \@poddata );
862 scan_items( \%items, "$pod", @poddata);
864 warn "$0: shouldn't be here (line ".__LINE__."\n";
867 @poddata = (); # clean-up a bit
870 || die "$0: error changing to directory $pwd: $!\n";
872 # cache the item list for later use
873 warn "caching items for later use\n" if $verbose;
874 open(CACHE, ">$itemcache") ||
875 die "$0: error open $itemcache for writing: $!\n";
877 print CACHE join(":", @podpath) . "\n$podroot\n";
878 foreach my $key (keys %items) {
879 print CACHE "$key $items{$key}\n";
884 # cache the directory list for later use
885 warn "caching directories for later use\n" if $verbose;
886 open(CACHE, ">$dircache") ||
887 die "$0: error open $dircache for writing: $!\n";
889 print CACHE join(":", @podpath) . "\n$podroot\n";
890 foreach my $key (keys %pages) {
891 print CACHE "$key $pages{$key}\n";
898 # scan_dir - scans the directory specified in $dir for subdirectories, .pod
899 # files, and .pm files. notes those that it finds. this information will
900 # be used later in order to figure out where the pages specified in L<>
901 # links are on the filesystem.
904 my($dir, $recurse) = @_;
905 my($t, @subdirs, @pods, $pod, $dirname, @dirs);
911 opendir(DIR, $dir) ||
912 die "$0: error opening directory $dir: $!\n";
913 while (defined($_ = readdir(DIR))) {
914 if (-d "$dir/$_" && $_ ne "." && $_ ne "..") { # directory
915 $pages{$_} = "" unless defined $pages{$_};
916 $pages{$_} .= "$dir/$_:";
918 } elsif (/\.pod\z/) { # .pod
920 $pages{$_} = "" unless defined $pages{$_};
921 $pages{$_} .= "$dir/$_.pod:";
922 push(@pods, "$dir/$_.pod");
923 } elsif (/\.html\z/) { # .html
925 $pages{$_} = "" unless defined $pages{$_};
926 $pages{$_} .= "$dir/$_.pod:";
927 } elsif (/\.pm\z/) { # .pm
929 $pages{$_} = "" unless defined $pages{$_};
930 $pages{$_} .= "$dir/$_.pm:";
931 push(@pods, "$dir/$_.pm");
936 # recurse on the subdirectories if necessary
938 foreach my $subdir (@subdirs) {
939 scan_dir("$dir/$subdir", $recurse);
945 # scan_headings - scan a pod file for head[1-6] tags, note the tags, and
949 my($sections, @data) = @_;
950 my($tag, $which_head, $otitle, $listdepth, $index);
952 # here we need local $ignore = 0;
953 # unfortunately, we can't have it, because $ignore is lexical
959 # scan for =head directives, note their name, and build an index
960 # pointing to each of them.
961 foreach my $line (@data) {
962 if ($line =~ /^=(head)([1-6])\s+(.*)/) {
963 ($tag, $which_head, $otitle) = ($1,$2,$3);
965 my $title = depod( $otitle );
966 my $name = htmlify( $title );
967 $$sections{$name} = 1;
968 $name =~ s/\s/_/g; # htmlify keeps spaces but we don't want them here..
969 $title = process_text( \$otitle );
971 while ($which_head != $listdepth) {
972 if ($which_head > $listdepth) {
973 $index .= "\n" . ("\t" x $listdepth) . "<ul>\n";
975 } elsif ($which_head < $listdepth) {
977 $index .= "\n" . ("\t" x $listdepth) . "</ul>\n";
981 $index .= "\n" . ("\t" x $listdepth) . "<li>" .
982 "<a href=\"#" . $name . "\">" .
983 $title . "</a></li>";
987 # finish off the lists
988 while ($listdepth--) {
989 $index .= "\n" . ("\t" x $listdepth) . "</ul>\n";
992 # get rid of bogus lists
993 $index =~ s,\t*<ul>\s*</ul>\n,,g;
995 $ignore = 1; # restore old value;
1001 # scan_items - scans the pod specified by $pod for =item directives. we
1002 # will use this information later on in resolving C<> links.
1005 my( $itemref, $pod, @poddata ) = @_;
1009 $pod =~ s/\.pod\z//;
1010 $pod .= ".html" if $pod;
1012 foreach $i (0..$#poddata) {
1013 my $txt = depod( $poddata[$i] );
1015 # figure out what kind of item it is.
1016 # Build string for referencing this item.
1017 if ( $txt =~ /\A=item\s+\*\s*(.*)\Z/s ) { # bullet
1020 } elsif( $txt =~ /\A=item\s+(?>\d+\.?)\s*(.*)\Z/s ) { # numbered list
1022 } elsif( $txt =~ /\A=item\s+(.*)\Z/s ) { # plain item
1027 my $fid = fragment_id( $item );
1028 $$itemref{$fid} = "$pod" if $fid;
1033 # process_head - convert a pod head[1-6] tag and convert it to HTML format.
1036 my($tag, $heading, $hasindex) = @_;
1038 # figure out the level of the =head
1039 $tag =~ /head([1-6])/;
1043 warn "$0: $podfile: unterminated list at =head in paragraph $paragraph. ignoring.\n";
1044 while( $listlevel ){
1050 if( $level == 1 && ! $top ){
1051 print HTML "<a href=\"#__index__\"><small>$backlink</small></a>\n"
1052 if $hasindex and $backlink;
1053 print HTML "</p>\n<hr />\n"
1055 print HTML "</p>\n";
1058 my $name = htmlify( depod( $heading ) );
1059 $name =~ s/\s/_/g; # htmlify keeps spaces but we don't want them here...
1060 my $convert = process_text( \$heading );
1061 print HTML "<h$level><a name=\"$name\">$convert</a></h$level>\n";
1066 # emit_item_tag - print an =item's text
1067 # Note: The global $EmittedItem is used for inhibiting self-references.
1071 sub emit_item_tag($$$){
1072 my( $otext, $text, $compact ) = @_;
1073 my $item = fragment_id( $text );
1075 $EmittedItem = $item;
1076 ### print STDERR "emit_item_tag=$item ($text)\n";
1078 print HTML '<strong>';
1079 if ($items_named{$item}++) {
1080 print HTML process_text( \$otext );
1082 my $name = 'item_' . $item;
1083 $name =~ s/\s/_/g; # we don't want spaces here...
1084 print HTML qq{<a name="$name">}, process_text( \$otext ), '</a>';
1086 print HTML "</strong><br />\n";
1087 undef( $EmittedItem );
1092 if( $items_seen[$listlevel]++ == 0 ){
1093 push( @listend, "</$tag>" );
1094 print HTML "<$tag>\n";
1096 my $emitted = $tag eq 'dl' ? 'dt' : 'li';
1097 print HTML "<$emitted>";
1102 # process_item - convert a pod item tag and convert it to HTML format.
1106 my $need_dd = 0; # set to 1 if we need a <dd></dd> after an item
1108 # lots of documents start a list without doing an =over. this is
1109 # bad! but, the proper thing to do seems to be to just assume
1110 # they did do an =over. so warn them once and then continue.
1111 if( $listlevel == 0 ){
1112 warn "$0: $podfile: unexpected =item directive in paragraph $paragraph. ignoring.\n";
1116 # formatting: insert a paragraph if preceding item has >1 paragraph
1118 print HTML "<p></p>\n";
1122 # remove formatting instructions from the text
1123 my $text = depod( $otext );
1125 my $emitted; # the tag actually emitted, used for closing
1127 # all the list variants:
1128 if( $text =~ /\A\*/ ){ # bullet
1129 $emitted = emit_li( 'ul' );
1130 if ($text =~ /\A\*\s+(.+)\Z/s ) { # with additional text
1132 $otext =~ s/\A\*\s+//;
1133 emit_item_tag( $otext, $tag, 1 );
1136 } elsif( $text =~ /\A\d+/ ){ # numbered list
1137 $emitted = emit_li( 'ol' );
1138 if ($text =~ /\A(?>\d+\.?)\s*(.+)\Z/s ) { # with additional text
1140 $otext =~ s/\A\d+\.?\s*//;
1141 emit_item_tag( $otext, $tag, 1 );
1144 } else { # definition list
1145 $emitted = emit_li( 'dl' );
1146 if ($text =~ /\A(.+)\Z/s ){ # should have text
1147 emit_item_tag( $otext, $text, 1 );
1151 print HTML "</$emitted>" if $emitted;
1157 # process_over - process a pod over tag and start a corresponding HTML list.
1162 push( @items_seen, 0 );
1167 # process_back - process a pod back tag and convert it to HTML format.
1170 if( $listlevel == 0 ){
1171 warn "$0: $podfile: unexpected =back directive in paragraph $paragraph. ignoring.\n";
1175 # close off the list. note, I check to see if $listend[$listlevel] is
1176 # defined because an =item directive may have never appeared and thus
1177 # $listend[$listlevel] may have never been initialized.
1179 if( defined $listend[$listlevel] ){
1180 print HTML '<p></p>' if $after_lpar;
1181 print HTML $listend[$listlevel];
1187 # clean up item count
1192 # process_cut - process a pod cut tag, thus start ignoring pod directives.
1199 # process_pod - process a pod tag, thus stop ignoring pod directives
1200 # until we see a corresponding cut.
1203 # no need to set $ignore to 0 cause the main loop did it
1207 # process_for - process a =for pod tag. if it's for html, spit
1208 # it out verbatim, if illustration, center it, otherwise ignore it.
1211 my($whom, $text) = @_;
1212 if ( $whom =~ /^(pod2)?html$/i) {
1214 } elsif ($whom =~ /^illustration$/i) {
1215 1 while chomp $text;
1216 for my $ext (qw[.png .gif .jpeg .jpg .tga .pcl .bmp]) {
1217 $text .= $ext, last if -r "$text$ext";
1219 print HTML qq{<p align="center"><img src="$text" alt="$text illustration" /></p>};
1224 # process_begin - process a =begin pod tag. this pushes
1225 # whom we're beginning on the begin stack. if there's a
1226 # begin stack, we only print if it us.
1229 my($whom, $text) = @_;
1231 push (@begin_stack, $whom);
1232 if ( $whom =~ /^(pod2)?html$/) {
1233 print HTML $text if $text;
1238 # process_end - process a =end pod tag. pop the
1239 # begin stack. die if we're mismatched.
1242 my($whom, $text) = @_;
1244 if ($begin_stack[-1] ne $whom ) {
1245 die "Unmatched begin/end at chunk $paragraph\n"
1247 pop( @begin_stack );
1251 # process_pre - indented paragraph, made into <pre></pre>
1260 # insert spaces in place of tabs
1263 1 while $line =~ s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e;
1267 # convert some special chars to HTML escapes
1268 $rest = html_escape($rest);
1270 # try and create links for all occurrences of perl.* within
1271 # the preformatted text.
1275 if ( defined $pages{$2} ){ # is a link
1276 qq($1<a href="$htmlroot/$pages{$2}">$2</a>);
1277 } elsif (defined $pages{dosify($2)}) { # is a link
1278 qq($1<a href="$htmlroot/$pages{dosify($2)}">$2</a>);
1284 (<a\ href="?) ([^>:]*:)? ([^>:]*) \.pod: ([^>:]*:)?
1287 if ( $htmlfileurl ne '' ){
1288 # Here, we take advantage of the knowledge
1289 # that $htmlfileurl ne '' implies $htmlroot eq ''.
1290 # Since $htmlroot eq '', we need to prepend $htmldir
1291 # on the fron of the link to get the absolute path
1292 # of the link's target. We check for a leading '/'
1293 # to avoid corrupting links that are #, file:, etc.
1295 $old_url = "$htmldir$old_url" if $old_url =~ m{^\/};
1296 $url = relativize_url( "$old_url.html", $htmlfileurl );
1303 # Look for embedded URLs and make them into links. We don't
1304 # relativize them since they are best left as the author intended.
1306 my $urls = '(' . join ('|', qw{
1319 my $gunk = '/#~:.?+=&%@!\-';
1320 my $punc = '.:!?\-;';
1321 my $any = "${ltrs}${gunk}${punc}";
1324 \b # start at word boundary
1326 $urls : # need resource and a colon
1327 (?!:) # Ignore File::, among others.
1328 [$any] +? # followed by one or more of any valid
1329 # character, but be conservative and
1330 # take only what you need to....
1332 (?= # look-ahead non-consumptive assertion
1333 [$punc]* # either 0 or more punctuation
1335 [^$any] # by a non-url char
1337 $ # end of the string
1340 $ # then end of the string
1342 }{<a href="$1">$1</a>}igox;
1344 # text should be as it is (verbatim)
1350 # pure text processing
1352 # pure_text/inIS_text: differ with respect to automatic C<> recognition.
1353 # we don't want this to happen within IS
1357 process_puretext( $text, \$ptQuote, 1 );
1362 process_puretext( $text, \$ptQuote, 0 );
1366 # process_puretext - process pure text (without pod-escapes) converting
1367 # double-quotes and handling implicit C<> links.
1369 sub process_puretext {
1370 my($text, $quote, $notinIS) = @_;
1372 ## Guessing at func() or [$@%&]*var references in plain text is destined
1373 ## to produce some strange looking ref's. uncomment to disable:
1376 my(@words, $lead, $trail);
1378 # convert double-quotes to single-quotes
1379 if( $$quote && $text =~ s/"/''/s ){
1382 while ($text =~ s/"([^"]*)"/``$1''/sg) {};
1383 $$quote = 1 if $text =~ s/"/``/s;
1385 # keep track of leading and trailing white-space
1386 $lead = ($text =~ s/\A(\s+)//s ? $1 : "");
1387 $trail = ($text =~ s/(\s+)\Z//s ? $1 : "");
1389 # split at space/non-space boundaries
1390 @words = split( /(?<=\s)(?=\S)|(?<=\S)(?=\s)/, $text );
1392 # process each word individually
1393 foreach my $word (@words) {
1395 next if $word =~ /^\s*$/;
1396 # see if we can infer a link
1397 if( $notinIS && $word =~ /^(\w+)\((.*)\)$/ ) {
1398 # has parenthesis so should have been a C<> ref
1399 ## try for a pagename (perlXXX(1))?
1400 my( $func, $args ) = ( $1, $2 );
1401 if( $args =~ /^\d+$/ ){
1402 my $url = page_sect( $word, '' );
1404 $word = "<a href=\"$url\">the $word manpage</a>";
1408 ## try function name for a link, append tt'ed argument list
1409 $word = emit_C( $func, '', "($args)");
1411 #### disabled. either all (including $\W, $\w+{.*} etc.) or nothing.
1412 ## } elsif( $notinIS && $word =~ /^[\$\@%&*]+\w+$/) {
1413 ## # perl variables, should be a C<> ref
1414 ## $word = emit_C( $word );
1416 } elsif ($word =~ m,^\w+://\w,) {
1418 # Don't relativize it: leave it as the author intended
1419 $word = qq(<a href="$word">$word</a>);
1420 } elsif ($word =~ /[\w.-]+\@[\w-]+\.\w/) {
1421 # looks like an e-mail address
1422 my ($w1, $w2, $w3) = ("", $word, "");
1423 ($w1, $w2, $w3) = ("(", $1, ")$2") if $word =~ /^\((.*?)\)(,?)/;
1424 ($w1, $w2, $w3) = ("<", $1, ">$2") if $word =~ /^<(.*?)>(,?)/;
1425 $word = qq($w1<a href="mailto:$w2">$w2</a>$w3);
1427 $word = html_escape($word) if $word =~ /["&<>]/;
1431 # put everything back together
1432 return $lead . join( '', @words ) . $trail;
1437 # process_text - handles plaintext that appears in the input pod file.
1438 # there may be pod commands embedded within the text so those must be
1439 # converted to html commands.
1442 sub process_text1($$;$$);
1443 sub pattern ($) { $_[0] ? '[^\S\n]+'.('>' x ($_[0] + 1)) : '>' }
1444 sub closing ($) { local($_) = shift; (defined && s/\s+$//) ? length : 0 }
1449 my $res = process_text1( 0, $tref );
1453 sub process_text1($$;$$){
1454 my( $lev, $rstr, $func, $closing ) = @_;
1457 unless (defined $func) {
1463 # B<text> - boldface
1464 $res = '<strong>' . process_text1( $lev, $rstr ) . '</strong>';
1466 } elsif( $func eq 'C' ){
1467 # C<code> - can be a ref or <code></code>
1468 # need to extract text
1469 my $par = go_ahead( $rstr, 'C', $closing );
1471 ## clean-up of the link target
1472 my $text = depod( $par );
1474 ### my $x = $par =~ /[BI]</ ? 'yes' : 'no' ;
1475 ### print STDERR "-->call emit_C($par) lev=$lev, par with BI=$x\n";
1477 $res = emit_C( $text, $lev > 1 || ($par =~ /[BI]</) );
1479 } elsif( $func eq 'E' ){
1480 # E<x> - convert to character
1481 $$rstr =~ s/^([^>]*)>//;
1483 $escape =~ s/^(\d+|X[\dA-F]+)$/#$1/i;
1486 } elsif( $func eq 'F' ){
1487 # F<filename> - italizice
1488 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1490 } elsif( $func eq 'I' ){
1491 # I<text> - italizice
1492 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1494 } elsif( $func eq 'L' ){
1496 ## L<text|cross-ref> => produce text, use cross-ref for linking
1497 ## L<cross-ref> => make text from cross-ref
1498 ## need to extract text
1499 my $par = go_ahead( $rstr, 'L', $closing );
1501 # some L<>'s that shouldn't be:
1502 # a) full-blown URL's are emitted as-is
1503 if( $par =~ m{^\w+://}s ){
1504 return make_URL_href( $par );
1506 # b) C<...> is stripped and treated as C<>
1507 if( $par =~ /^C<(.*)>$/ ){
1508 my $text = depod( $1 );
1509 return emit_C( $text, $lev > 1 || ($par =~ /[BI]</) );
1512 # analyze the contents
1513 $par =~ s/\n/ /g; # undo word-wrapped tags
1516 if( $par =~ s{^([^|]+)\|}{} ){
1520 # make sure sections start with a /
1523 my( $page, $section, $ident );
1525 # check for link patterns
1526 if( $par =~ m{^([^/]+?)/(?!")(.*?)$} ){ # name/ident
1527 # we've got a name/ident (no quotes)
1528 ( $page, $ident ) = ( $1, $2 );
1529 ### print STDERR "--> L<$par> to page $page, ident $ident\n";
1531 } elsif( $par =~ m{^(.*?)/"?(.*?)"?$} ){ # [name]/"section"
1532 # even though this should be a "section", we go for ident first
1533 ( $page, $ident ) = ( $1, $2 );
1534 ### print STDERR "--> L<$par> to page $page, section $section\n";
1536 } elsif( $par =~ /\s/ ){ # this must be a section with missing quotes
1537 ( $page, $section ) = ( '', $par );
1538 ### print STDERR "--> L<$par> to void page, section $section\n";
1541 ( $page, $section ) = ( $par, '' );
1542 ### print STDERR "--> L<$par> to page $par, void section\n";
1545 # now, either $section or $ident is defined. the convoluted logic
1546 # below tries to resolve L<> according to what the user specified.
1547 # failing this, we try to find the next best thing...
1548 my( $url, $ltext, $fid );
1551 if( defined $ident ){
1552 ## try to resolve $ident as an item
1553 ( $url, $fid ) = coderef( $page, $ident );
1555 if( ! defined( $linktext ) ){
1557 $linktext .= " in " if $ident && $page;
1558 $linktext .= "the $page manpage" if $page;
1560 ### print STDERR "got coderef url=$url\n";
1563 ## no luck: go for a section (auto-quoting!)
1566 ## now go for a section
1567 my $htmlsection = htmlify( $section );
1568 $url = page_sect( $page, $htmlsection );
1570 if( ! defined( $linktext ) ){
1571 $linktext = $section;
1572 $linktext .= " in " if $section && $page;
1573 $linktext .= "the $page manpage" if $page;
1575 ### print STDERR "got page/section url=$url\n";
1578 ## no luck: go for an ident
1585 ( $url, $fid ) = coderef( $page, $ident );
1587 if( ! defined( $linktext ) ){
1589 $linktext .= " in " if $ident && $page;
1590 $linktext .= "the $page manpage" if $page;
1592 ### print STDERR "got section=>coderef url=$url\n";
1596 # warning; show some text.
1597 $linktext = $opar unless defined $linktext;
1598 warn "$0: $podfile: cannot resolve L<$opar> in paragraph $paragraph.";
1601 # now we have a URL or just plain code
1602 $$rstr = $linktext . '>' . $$rstr;
1603 if( defined( $url ) ){
1604 $res = "<a href=\"$url\">" . process_text1( $lev, $rstr ) . '</a>';
1606 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1609 } elsif( $func eq 'S' ){
1610 # S<text> - non-breaking spaces
1611 $res = process_text1( $lev, $rstr );
1612 $res =~ s/ / /g;
1614 } elsif( $func eq 'X' ){
1616 $$rstr =~ s/^[^>]*>//;
1618 } elsif( $func eq 'Z' ){
1620 warn "$0: $podfile: invalid X<> in paragraph $paragraph."
1621 unless $$rstr =~ s/^>//;
1624 my $term = pattern $closing;
1625 while( $$rstr =~ s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|$term)//s ){
1626 # all others: either recurse into new function or
1627 # terminate at closing angle bracket(s)
1629 $pt .= $2 if !$3 && $lev == 1;
1630 $res .= $lev == 1 ? pure_text( $pt ) : inIS_text( $pt );
1631 return $res if !$3 && $lev > 1;
1633 $res .= process_text1( $lev, $rstr, $3, closing $4 );
1637 $res .= pure_text( $$rstr );
1639 warn "$0: $podfile: undelimited $func<> in paragraph $paragraph.";
1646 # go_ahead: extract text of an IS (can be nested)
1649 my( $rstr, $func, $closing ) = @_;
1651 my @closing = ($closing);
1653 s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|@{[pattern $closing[0]]})//s ){
1657 return $res unless @closing;
1659 unshift @closing, closing $4;
1663 warn "$0: $podfile: undelimited $func<> in paragraph $paragraph.";
1668 # emit_C - output result of C<text>
1669 # $text is the depod-ed text
1672 my( $text, $nocode, $args ) = @_;
1673 $args = '' unless defined $args;
1675 my( $url, $fid ) = coderef( undef(), $text );
1677 # need HTML-safe text
1678 my $linktext = html_escape( "$text$args" );
1680 if( defined( $url ) &&
1681 (!defined( $EmittedItem ) || $EmittedItem ne $fid ) ){
1682 $res = "<a href=\"$url\"><code>$linktext</code></a>";
1683 } elsif( 0 && $nocode ){
1686 $res = "<code>$linktext</code>";
1692 # html_escape: make text safe for HTML
1696 $rest =~ s/&/&/g;
1697 $rest =~ s/</</g;
1698 $rest =~ s/>/>/g;
1699 $rest =~ s/"/"/g;
1700 $rest =~ s/'/'/g;
1706 # dosify - convert filenames to 8.3
1710 return lc($str) if $^O eq 'VMS'; # VMS just needs casing
1713 $str =~ s/(\.\w+)/substr ($1,0,4)/ge;
1714 $str =~ s/(\w+)/substr ($1,0,8)/ge;
1720 # page_sect - make a URL from the text of a L<>
1723 my( $page, $section ) = @_;
1724 my( $linktext, $page83, $link); # work strings
1726 # check if we know that this is a section in this page
1727 if (!defined $pages{$page} && defined $sections{$page}) {
1730 ### print STDERR "reset page='', section=$section\n";
1733 $page83=dosify($page);
1734 $page=$page83 if (defined $pages{$page83});
1736 $link = "#" . htmlify( $section );
1737 } elsif ( $page =~ /::/ ) {
1739 # Search page cache for an entry keyed under the html page name,
1740 # then look to see what directory that page might be in. NOTE:
1741 # this will only find one page. A better solution might be to produce
1742 # an intermediate page that is an index to all such pages.
1743 my $page_name = $page ;
1744 $page_name =~ s,^.*/,,s ;
1745 if ( defined( $pages{ $page_name } ) &&
1746 $pages{ $page_name } =~ /([^:]*$page)\.(?:pod|pm):/
1751 # NOTE: This branch assumes that all A::B pages are located in
1752 # $htmlroot/A/B.html . This is often incorrect, since they are
1753 # often in $htmlroot/lib/A/B.html or such like. Perhaps we could
1754 # analyze the contents of %pages and figure out where any
1755 # cousins of A::B are, then assume that. So, if A::B isn't found,
1756 # but A::C is found in lib/A/C.pm, then A::B is assumed to be in
1757 # lib/A/B.pm. This is also limited, but it's an improvement.
1758 # Maybe a hints file so that the links point to the correct places
1762 $link = "$htmlroot/$page.html";
1763 $link .= "#" . htmlify( $section ) if ($section);
1764 } elsif (!defined $pages{$page}) {
1767 $section = htmlify( $section ) if $section ne "";
1768 ### print STDERR "...section=$section\n";
1770 # if there is a directory by the name of the page, then assume that an
1771 # appropriate section will exist in the subdirectory
1772 # if ($section ne "" && $pages{$page} =~ /([^:]*[^(\.pod|\.pm)]):/) {
1773 if ($section ne "" && $pages{$page} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
1774 $link = "$htmlroot/$1/$section.html";
1775 ### print STDERR "...link=$link\n";
1777 # since there is no directory by the name of the page, the section will
1778 # have to exist within a .html of the same name. thus, make sure there
1779 # is a .pod or .pm that might become that .html
1781 $section = "#$section" if $section;
1782 ### print STDERR "...section=$section\n";
1784 # check if there is a .pod with the page name
1785 if ($pages{$page} =~ /([^:]*)\.pod:/) {
1786 $link = "$htmlroot/$1.html$section";
1787 } elsif ($pages{$page} =~ /([^:]*)\.pm:/) {
1788 $link = "$htmlroot/$1.html$section";
1796 # Here, we take advantage of the knowledge that $htmlfileurl ne ''
1797 # implies $htmlroot eq ''. This means that the link in question
1798 # needs a prefix of $htmldir if it begins with '/'. The test for
1799 # the initial '/' is done to avoid '#'-only links, and to allow
1800 # for other kinds of links, like file:, ftp:, etc.
1802 if ( $htmlfileurl ne '' ) {
1803 $link = "$htmldir$link" if $link =~ m{^/}s;
1804 $url = relativize_url( $link, $htmlfileurl );
1805 # print( " b: [$link,$htmlfileurl,$url]\n" );
1818 # relativize_url - convert an absolute URL to one relative to a base URL.
1819 # Assumes both end in a filename.
1821 sub relativize_url {
1822 my ($dest,$source) = @_ ;
1824 my ($dest_volume,$dest_directory,$dest_file) =
1825 File::Spec::Unix->splitpath( $dest ) ;
1826 $dest = File::Spec::Unix->catpath( $dest_volume, $dest_directory, '' ) ;
1828 my ($source_volume,$source_directory,$source_file) =
1829 File::Spec::Unix->splitpath( $source ) ;
1830 $source = File::Spec::Unix->catpath( $source_volume, $source_directory, '' ) ;
1833 if ( $dest ne '' ) {
1834 $rel_path = File::Spec::Unix->abs2rel( $dest, $source ) ;
1837 if ( $rel_path ne '' &&
1838 substr( $rel_path, -1 ) ne '/' &&
1839 substr( $dest_file, 0, 1 ) ne '#'
1841 $rel_path .= "/$dest_file" ;
1844 $rel_path .= "$dest_file" ;
1852 # coderef - make URL from the text of a C<>
1855 my( $page, $item ) = @_;
1858 my $fid = fragment_id( $item );
1859 if( defined( $page ) ){
1860 # we have been given a $page...
1863 # Do we take it? Item could be a section!
1864 my $base = $items{$fid} || "";
1865 $base =~ s{[^/]*/}{};
1866 if( $base ne "$page.html" ){
1867 ### print STDERR "coderef( $page, $item ): items{$fid} = $items{$fid} = $base => discard page!\n";
1872 # no page - local items precede cached items
1873 if( defined( $fid ) ){
1874 if( exists $local_items{$fid} ){
1875 $page = $local_items{$fid};
1877 $page = $items{$fid};
1882 # if there was a pod file that we found earlier with an appropriate
1883 # =item directive, then create a link to that page.
1884 if( defined $page ){
1886 if( exists $pages{$page} and $pages{$page} =~ /([^:.]*)\.[^:]*:/){
1887 $page = $1 . '.html';
1889 my $link = "$htmlroot/$page#item_$fid";
1891 # Here, we take advantage of the knowledge that $htmlfileurl
1892 # ne '' implies $htmlroot eq ''.
1893 if ( $htmlfileurl ne '' ) {
1894 $link = "$htmldir$link" ;
1895 $url = relativize_url( $link, $htmlfileurl ) ;
1900 $url = "#item_" . $fid;
1903 confess "url has space: $url" if $url =~ /"[^"]*\s[^"]*"/;
1905 return( $url, $fid );
1911 # Adapted from Nick Ing-Simmons' PodToHtml package.
1913 my $source_file = shift ;
1914 my $destination_file = shift;
1916 my $source = URI::file->new_abs($source_file);
1917 my $uo = URI::file->new($destination_file,$source)->abs;
1918 return $uo->rel->as_string;
1923 # finish_list - finish off any pending HTML lists. this should be called
1924 # after the entire pod file has been read and converted.
1927 while ($listlevel > 0) {
1928 print HTML "</dl>\n";
1934 # htmlify - converts a pod section specification to a suitable section
1935 # specification for HTML. Note that we keep spaces and special characters
1936 # except ", ? (Netscape problem) and the hyphen (writer's problem...).
1940 $heading =~ s/(\s+)/ /g;
1941 $heading =~ s/\s+\Z//;
1942 $heading =~ s/\A\s+//;
1943 # The hyphen is a disgrace to the English language.
1944 $heading =~ s/[-"?]//g;
1945 $heading = lc( $heading );
1950 # depod - convert text by eliminating all interior sequences
1951 # Note: can be called with copy or modify semantics
1958 $E2c{amp} = '&'; # in Tk's pods
1966 ${$_[0]} = depod1( \$string );
1974 my( $rstr, $func, $closing ) = @_;
1976 return $res unless defined $$rstr;
1977 if( ! defined( $func ) ){
1978 # skip to next begin of an interior sequence
1979 while( $$rstr =~ s/\A(.*?)([BCEFILSXZ])<(<+[^\S\n]+)?// ){
1980 # recurse into its text
1981 $res .= $1 . depod1( $rstr, $2, closing $3);
1984 } elsif( $func eq 'E' ){
1985 # E<x> - convert to character
1986 $$rstr =~ s/^([^>]*)>//;
1987 $res .= $E2c{$1} || "";
1988 } elsif( $func eq 'X' ){
1990 $$rstr =~ s/^[^>]*>//;
1991 } elsif( $func eq 'Z' ){
1995 # all others: either recurse into new function or
1996 # terminate at closing angle bracket
1997 my $term = pattern $closing;
1998 while( $$rstr =~ s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|$term)// ){
2001 $res .= depod1( $rstr, $3, closing $4 );
2003 ## If we're here and $2 ne '>': undelimited interior sequence.
2004 ## Ignored, as this is called without proper indication of where we are.
2005 ## Rely on process_text to produce diagnostics.
2011 # fragment_id - construct a fragment identifier from:
2013 # b) contents of C<...>
2018 $text =~ s/\s+\Z//s;
2020 # a method or function?
2021 return $1 if $text =~ /(\w+)\s*\(/;
2022 return $1 if $text =~ /->\s*(\w+)\s*\(?/;
2025 return $1 if $text =~ /^([$@%*]\S+)/;
2027 # some pattern matching operator?
2028 return $1 if $text =~ m|^(\w+/).*/\w*$|;
2030 # fancy stuff... like "do { }"
2031 return $1 if $text =~ m|^(\w+)\s*{.*}$|;
2033 # honour the perlfunc manpage: func [PAR[,[ ]PAR]...]
2034 # and some funnies with ... Module ...
2035 return $1 if $text =~ m{^([a-z\d]+)(\s+[A-Z\d,/& ]+)?$};
2036 return $1 if $text =~ m{^([a-z\d]+)\s+Module(\s+[A-Z\d,/& ]+)?$};
2039 $text =~ s/\s+/_/sg;
2041 defined( $hc[ord($1)] ) ? $hc[ord($1)]
2042 : ( $hc[ord($1)] = sprintf( "%%%02X", ord($1) ) ) }gxe;
2043 $text = substr( $text, 0, 50 );
2050 # make_URL_href - generate HTML href from URL
2051 # Special treatment for CGI queries.
2053 sub make_URL_href($){
2056 s{^(http:[-\w/#~:.+=&%@!]+)(\?.*)$}{<a href="$1$2">$1</a>}i ){
2057 $url = "<a href=\"$url\">$url</a>";