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 $verbose;
377 @poddata = map { s/\r\n/\n/g;
379 map { "$_\n\n" } split /\n\n/ :
382 @poddata = map { s/\r/\n/g;
384 map { "$_\n\n" } split /\n\n/ :
391 clean_data( \@poddata );
393 # scan the pod for =head[1-6] directives and build an index
394 my $index = scan_headings(\%sections, @poddata);
397 warn "No headings in $podfile\n" if $verbose;
400 # open the output file
401 open(HTML, ">$htmlfile")
402 || die "$0: cannot open $htmlfile file for output: $!\n";
404 # put a title in the HTML file if one wasn't specified
407 for (my $i = 0; $i < @poddata; $i++) {
408 if ($poddata[$i] =~ /^=head1\s*NAME\b/m) {
409 for my $para ( @poddata[$i, $i+1] ) {
411 if ($title) = $para =~ /(\S+\s+-+.*\S)/s;
418 if (!$title and $podfile =~ /\.pod\z/) {
419 # probably a split pod so take first =head[12] as title
420 for (my $i = 0; $i < @poddata; $i++) {
421 last if ($title) = $poddata[$i] =~ /^=head[12]\s*(.*)/;
423 warn "adopted '$title' as title for $podfile\n"
424 if $verbose and $title;
427 $title =~ s/\s*\(.*\)//;
429 warn "$0: no title for $podfile.\n" unless $quiet;
430 $podfile =~ /^(.*)(\.[^.\/]+)?\z/s;
431 $title = ($podfile eq "-" ? 'No Title' : $1);
432 warn "using $title" if $verbose;
434 $title = html_escape($title);
437 my $bodystyle = ' style="background-color: white"';
438 my $tdstyle = ' style="background-color: #cccccc"';
441 $csslink = qq(\n<link rel="stylesheet" href="$css" type="text/css" />);
442 $csslink =~ s,\\,/,g;
443 $csslink =~ s,(/.):,$1|,;
448 my $block = $header ? <<END_OF_BLOCK : '';
449 <table border="0" width="100%" cellspacing="0" cellpadding="3">
450 <tr><td class="block"$tdstyle valign="middle">
451 <big><strong><span class="block"> $title</span></strong></big>
456 print HTML <<END_OF_HEAD;
457 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
458 <html xmlns="http://www.w3.org/1999/xhtml">
460 <title>$title</title>$csslink
461 <link rev="made" href="mailto:$Config{perladmin}" />
468 # load/reload/validate/cache %pages and %items
469 get_cache($dircache, $itemcache, \@podpath, $podroot, $recurse);
471 # scan the pod for =item directives
472 scan_items( \%local_items, "", @poddata);
474 # put an index at the top of the file. note, if $doindex is 0 we
475 # still generate an index, but surround it with an html comment.
476 # that way some other program can extract it if desired.
478 print HTML "<p><a name=\"__index__\"></a></p>\n";
479 print HTML "<!-- INDEX BEGIN -->\n";
480 print HTML "<!--\n" unless $doindex;
482 print HTML "-->\n" unless $doindex;
483 print HTML "<!-- INDEX END -->\n\n";
484 print HTML "<hr />\n" if $doindex and $index;
486 # now convert this file
487 my $after_item; # set to true after an =item
489 warn "Converting input file $podfile\n" if $verbose;
490 foreach my $i (0..$#poddata){
491 $ptQuote = 0; # status of quote conversion
495 if (/^(=.*)/s) { # is it a pod directive?
500 if (/^=begin\s+(\S+)\s*(.*)/si) {# =begin
501 process_begin($1, $2);
502 } elsif (/^=end\s+(\S+)\s*(.*)/si) {# =end
504 } elsif (/^=cut/) { # =cut
506 } elsif (/^=pod/) { # =pod
509 next if @begin_stack && $begin_stack[-1] ne 'html';
511 if (/^=(head[1-6])\s+(.*\S)/s) { # =head[1-6] heading
512 process_head( $1, $2, $doindex && $index );
513 } elsif (/^=item\s*(.*\S)?/sm) { # =item text
514 $need_dd = process_item( $1 );
516 } elsif (/^=over\s*(.*)/) { # =over N
518 } elsif (/^=back/) { # =back
520 } elsif (/^=for\s+(\S+)\s*(.*)/si) {# =for
524 warn "$0: $podfile: unknown pod directive '$1' in "
525 . "paragraph $paragraph. ignoring.\n";
532 next if @begin_stack && $begin_stack[-1] ne 'html';
533 print HTML and next if @begin_stack && $begin_stack[-1] eq 'html';
534 print HTML "<dd>\n" if $need_dd;
536 if( $text =~ /\A\s+/ ){
537 process_pre( \$text );
538 print HTML "<pre>\n$text</pre>\n";
541 process_text( \$text );
543 # experimental: check for a paragraph where all lines
544 # have some ...\t...\t...\n pattern
546 my @lines = split( "\n", $text );
549 foreach my $line ( @lines ){
550 if( $line =~ /\S/ && $line !~ /\t/ ){
556 $text =~ s/\t+/<td>/g;
557 $text =~ s/^/<tr><td>/gm;
558 $text = '<table cellspacing="0" cellpadding="0">' .
563 ## end of experimental
566 print HTML "$text\n";
569 print HTML "<p>$text</p>\n";
572 print HTML "</dd>\n" if $need_dd;
577 # finish off any pending directives
581 print HTML "<p><a href=\"#__index__\"><small>$backlink</small></a></p>\n"
582 if $doindex and $index and $backlink;
584 print HTML <<END_OF_TAIL;
591 # close the html file
594 warn "Finished\n" if $verbose;
597 ##############################################################################
599 my $usage; # see below
602 warn "$0: $podfile: @_\n" if @_;
606 $usage =<<END_OF_USAGE;
607 Usage: $0 --help --htmlroot=<name> --infile=<name> --outfile=<name>
608 --podpath=<name>:...:<name> --podroot=<name>
609 --libpods=<name>:...:<name> --recurse --verbose --index
610 --netscape --norecurse --noindex --cachedir=<name>
612 --backlink - set text for "back to top" links (default: none).
613 --cachedir - directory for the item and directory cache files.
614 --css - stylesheet URL
615 --flush - flushes the item and directory caches.
616 --[no]header - produce block header/footer (default is no headers).
617 --help - prints this message.
618 --htmldir - directory for resulting HTML files.
619 --htmlroot - http-server base directory from which all relative paths
620 in podpath stem (default is /).
621 --[no]index - generate an index at the top of the resulting html
623 --infile - filename for the pod to convert (input taken from stdin
625 --libpods - colon-separated list of pages to search for =item pod
626 directives in as targets of C<> and implicit links (empty
627 by default). note, these are not filenames, but rather
628 page names like those that appear in L<> links.
629 --outfile - filename for the resulting html file (output sent to
631 --podpath - colon-separated list of directories containing library
632 pods (empty by default).
633 --podroot - filesystem base directory from which all relative paths
634 in podpath stem (default is .).
635 --[no]quiet - supress some benign warning messages (default is off).
636 --[no]recurse - recurse on those subdirectories listed in podpath
638 --title - title that will appear in resulting html file.
639 --[no]verbose - self-explanatory (off by default).
640 --[no]netscape - deprecated, has no effect. for backwards compatibility only.
644 sub parse_command_line {
645 my ($opt_backlink,$opt_cachedir,$opt_css,$opt_flush,$opt_header,$opt_help,
646 $opt_htmldir,$opt_htmlroot,$opt_index,$opt_infile,$opt_libpods,
647 $opt_netscape,$opt_outfile,$opt_podpath,$opt_podroot,$opt_quiet,
648 $opt_recurse,$opt_title,$opt_verbose);
650 unshift @ARGV, split ' ', $Config{pod2html} if $Config{pod2html};
651 my $result = GetOptions(
652 'backlink=s' => \$opt_backlink,
653 'cachedir=s' => \$opt_cachedir,
654 'css=s' => \$opt_css,
655 'flush' => \$opt_flush,
656 'header!' => \$opt_header,
657 'help' => \$opt_help,
658 'htmldir=s' => \$opt_htmldir,
659 'htmlroot=s' => \$opt_htmlroot,
660 'index!' => \$opt_index,
661 'infile=s' => \$opt_infile,
662 'libpods=s' => \$opt_libpods,
663 'netscape!' => \$opt_netscape,
664 'outfile=s' => \$opt_outfile,
665 'podpath=s' => \$opt_podpath,
666 'podroot=s' => \$opt_podroot,
667 'quiet!' => \$opt_quiet,
668 'recurse!' => \$opt_recurse,
669 'title=s' => \$opt_title,
670 'verbose!' => \$opt_verbose,
672 usage("-", "invalid parameters") if not $result;
674 usage("-") if defined $opt_help; # see if the user asked for help
675 $opt_help = ""; # just to make -w shut-up.
677 @podpath = split(":", $opt_podpath) if defined $opt_podpath;
678 @libpods = split(":", $opt_libpods) if defined $opt_libpods;
680 $backlink = $opt_backlink if defined $opt_backlink;
681 $cachedir = $opt_cachedir if defined $opt_cachedir;
682 $css = $opt_css if defined $opt_css;
683 $header = $opt_header if defined $opt_header;
684 $htmldir = $opt_htmldir if defined $opt_htmldir;
685 $htmlroot = $opt_htmlroot if defined $opt_htmlroot;
686 $doindex = $opt_index if defined $opt_index;
687 $podfile = $opt_infile if defined $opt_infile;
688 $htmlfile = $opt_outfile if defined $opt_outfile;
689 $podroot = $opt_podroot if defined $opt_podroot;
690 $quiet = $opt_quiet if defined $opt_quiet;
691 $recurse = $opt_recurse if defined $opt_recurse;
692 $title = $opt_title if defined $opt_title;
693 $verbose = $opt_verbose if defined $opt_verbose;
695 warn "Flushing item and directory caches\n"
696 if $opt_verbose && defined $opt_flush;
697 $dircache = "$cachedir/pod2htmd$cache_ext";
698 $itemcache = "$cachedir/pod2htmi$cache_ext";
699 unlink($dircache, $itemcache) if defined $opt_flush;
706 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
707 my @cache_key_args = @_;
709 # A first-level cache:
710 # Don't bother reading the cache files if they still apply
711 # and haven't changed since we last read them.
713 my $this_cache_key = cache_key(@cache_key_args);
715 return if $saved_cache_key and $this_cache_key eq $saved_cache_key;
717 # load the cache of %pages and %items if possible. $tests will be
718 # non-zero if successful.
720 if (-f $dircache && -f $itemcache) {
721 warn "scanning for item cache\n" if $verbose;
722 $tests = load_cache($dircache, $itemcache, $podpath, $podroot);
725 # if we didn't succeed in loading the cache then we must (re)build
728 warn "scanning directories in pod-path\n" if $verbose;
729 scan_podpath($podroot, $recurse, 0);
731 $saved_cache_key = cache_key(@cache_key_args);
735 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
736 return join('!', $dircache, $itemcache, $recurse,
737 @$podpath, $podroot, stat($dircache), stat($itemcache));
741 # load_cache - tries to find if the caches stored in $dircache and $itemcache
742 # are valid caches of %pages and %items. if they are valid then it loads
743 # them and returns a non-zero value.
746 my($dircache, $itemcache, $podpath, $podroot) = @_;
752 open(CACHE, "<$itemcache") ||
753 die "$0: error opening $itemcache for reading: $!\n";
756 # is it the same podpath?
759 $tests++ if (join(":", @$podpath) eq $_);
761 # is it the same podroot?
764 $tests++ if ($podroot eq $_);
766 # load the cache if its good
772 warn "loading item cache\n" if $verbose;
779 warn "scanning for directory cache\n" if $verbose;
780 open(CACHE, "<$dircache") ||
781 die "$0: error opening $dircache for reading: $!\n";
785 # is it the same podpath?
788 $tests++ if (join(":", @$podpath) eq $_);
790 # is it the same podroot?
793 $tests++ if ($podroot eq $_);
795 # load the cache if its good
801 warn "loading directory cache\n" if $verbose;
813 # scan_podpath - scans the directories specified in @podpath for directories,
814 # .pod files, and .pm files. it also scans the pod files specified in
815 # @libpods for =item directives.
818 my($podroot, $recurse, $append) = @_;
820 my($libpod, $dirname, $pod, @files, @poddata);
827 # scan each directory listed in @podpath
830 || die "$0: error changing to directory $podroot: $!\n";
831 foreach $dir (@podpath) {
832 scan_dir($dir, $recurse);
835 # scan the pods listed in @libpods for =item directives
836 foreach $libpod (@libpods) {
837 # if the page isn't defined then we won't know where to find it
839 next unless defined $pages{$libpod} && $pages{$libpod};
841 # if there is a directory then use the .pod and .pm files within it.
842 # NOTE: Only finds the first so-named directory in the tree.
843 # if ($pages{$libpod} =~ /([^:]*[^(\.pod|\.pm)]):/) {
844 if ($pages{$libpod} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
845 # find all the .pod and .pm files within the directory
847 opendir(DIR, $dirname) ||
848 die "$0: error opening directory $dirname: $!\n";
849 @files = grep(/(\.pod|\.pm)\z/ && ! -d $_, readdir(DIR));
852 # scan each .pod and .pm file for =item directives
853 foreach $pod (@files) {
854 open(POD, "<$dirname/$pod") ||
855 die "$0: error opening $dirname/$pod for input: $!\n";
858 clean_data( \@poddata );
860 scan_items( \%items, "$dirname/$pod", @poddata);
863 # use the names of files as =item directives too.
864 ### Don't think this should be done this way - confuses issues.(WL)
865 ### foreach $pod (@files) {
866 ### $pod =~ /^(.*)(\.pod|\.pm)$/;
867 ### $items{$1} = "$dirname/$1.html" if $1;
869 } elsif ($pages{$libpod} =~ /([^:]*\.pod):/ ||
870 $pages{$libpod} =~ /([^:]*\.pm):/) {
871 # scan the .pod or .pm file for =item directives
873 open(POD, "<$pod") ||
874 die "$0: error opening $pod for input: $!\n";
877 clean_data( \@poddata );
879 scan_items( \%items, "$pod", @poddata);
881 warn "$0: shouldn't be here (line ".__LINE__."\n";
884 @poddata = (); # clean-up a bit
887 || die "$0: error changing to directory $pwd: $!\n";
889 # cache the item list for later use
890 warn "caching items for later use\n" if $verbose;
891 open(CACHE, ">$itemcache") ||
892 die "$0: error open $itemcache for writing: $!\n";
894 print CACHE join(":", @podpath) . "\n$podroot\n";
895 foreach my $key (keys %items) {
896 print CACHE "$key $items{$key}\n";
901 # cache the directory list for later use
902 warn "caching directories for later use\n" if $verbose;
903 open(CACHE, ">$dircache") ||
904 die "$0: error open $dircache for writing: $!\n";
906 print CACHE join(":", @podpath) . "\n$podroot\n";
907 foreach my $key (keys %pages) {
908 print CACHE "$key $pages{$key}\n";
915 # scan_dir - scans the directory specified in $dir for subdirectories, .pod
916 # files, and .pm files. notes those that it finds. this information will
917 # be used later in order to figure out where the pages specified in L<>
918 # links are on the filesystem.
921 my($dir, $recurse) = @_;
922 my($t, @subdirs, @pods, $pod, $dirname, @dirs);
928 opendir(DIR, $dir) ||
929 die "$0: error opening directory $dir: $!\n";
930 while (defined($_ = readdir(DIR))) {
931 if (-d "$dir/$_" && $_ ne "." && $_ ne "..") { # directory
932 $pages{$_} = "" unless defined $pages{$_};
933 $pages{$_} .= "$dir/$_:";
935 } elsif (/\.pod\z/) { # .pod
937 $pages{$_} = "" unless defined $pages{$_};
938 $pages{$_} .= "$dir/$_.pod:";
939 push(@pods, "$dir/$_.pod");
940 } elsif (/\.html\z/) { # .html
942 $pages{$_} = "" unless defined $pages{$_};
943 $pages{$_} .= "$dir/$_.pod:";
944 } elsif (/\.pm\z/) { # .pm
946 $pages{$_} = "" unless defined $pages{$_};
947 $pages{$_} .= "$dir/$_.pm:";
948 push(@pods, "$dir/$_.pm");
953 # recurse on the subdirectories if necessary
955 foreach my $subdir (@subdirs) {
956 scan_dir("$dir/$subdir", $recurse);
962 # scan_headings - scan a pod file for head[1-6] tags, note the tags, and
966 my($sections, @data) = @_;
967 my($tag, $which_head, $otitle, $listdepth, $index);
969 # here we need local $ignore = 0;
970 # unfortunately, we can't have it, because $ignore is lexical
976 # scan for =head directives, note their name, and build an index
977 # pointing to each of them.
978 foreach my $line (@data) {
979 if ($line =~ /^=(head)([1-6])\s+(.*)/) {
980 ($tag, $which_head, $otitle) = ($1,$2,$3);
982 my $title = depod( $otitle );
983 my $name = anchorify( $title );
984 $$sections{$name} = 1;
985 $title = process_text( \$otitle );
987 while ($which_head != $listdepth) {
988 if ($which_head > $listdepth) {
989 $index .= "\n" . ("\t" x $listdepth) . "<ul>\n";
991 } elsif ($which_head < $listdepth) {
993 $index .= "\n" . ("\t" x $listdepth) . "</ul>\n";
997 $index .= "\n" . ("\t" x $listdepth) . "<li>" .
998 "<a href=\"#" . $name . "\">" .
999 $title . "</a></li>";
1003 # finish off the lists
1004 while ($listdepth--) {
1005 $index .= "\n" . ("\t" x $listdepth) . "</ul>\n";
1008 # get rid of bogus lists
1009 $index =~ s,\t*<ul>\s*</ul>\n,,g;
1011 $ignore = 1; # restore old value;
1017 # scan_items - scans the pod specified by $pod for =item directives. we
1018 # will use this information later on in resolving C<> links.
1021 my( $itemref, $pod, @poddata ) = @_;
1025 $pod =~ s/\.pod\z//;
1026 $pod .= ".html" if $pod;
1028 foreach $i (0..$#poddata) {
1029 my $txt = depod( $poddata[$i] );
1031 # figure out what kind of item it is.
1032 # Build string for referencing this item.
1033 if ( $txt =~ /\A=item\s+\*\s*(.*)\Z/s ) { # bullet
1036 } elsif( $txt =~ /\A=item\s+(?>\d+\.?)\s*(.*)\Z/s ) { # numbered list
1038 } elsif( $txt =~ /\A=item\s+(.*)\Z/s ) { # plain item
1043 my $fid = fragment_id( $item );
1044 $$itemref{$fid} = "$pod" if $fid;
1049 # process_head - convert a pod head[1-6] tag and convert it to HTML format.
1052 my($tag, $heading, $hasindex) = @_;
1054 # figure out the level of the =head
1055 $tag =~ /head([1-6])/;
1059 warn "$0: $podfile: unterminated list at =head in paragraph $paragraph. ignoring.\n";
1060 while( $listlevel ){
1066 if( $level == 1 && ! $top ){
1067 print HTML "<a href=\"#__index__\"><small>$backlink</small></a>\n"
1068 if $hasindex and $backlink;
1069 print HTML "</p>\n<hr />\n"
1071 print HTML "</p>\n";
1074 my $name = anchorify( depod( $heading ) );
1075 my $convert = process_text( \$heading );
1076 print HTML "<h$level><a name=\"$name\">$convert</a></h$level>\n";
1081 # emit_item_tag - print an =item's text
1082 # Note: The global $EmittedItem is used for inhibiting self-references.
1086 sub emit_item_tag($$$){
1087 my( $otext, $text, $compact ) = @_;
1088 my $item = fragment_id( $text );
1090 $EmittedItem = $item;
1091 ### print STDERR "emit_item_tag=$item ($text)\n";
1093 print HTML '<strong>';
1094 if ($items_named{$item}++) {
1095 print HTML process_text( \$otext );
1097 my $name = 'item_' . $item;
1098 $name = anchorify($name);
1099 print HTML qq{<a name="$name">}, process_text( \$otext ), '</a>';
1101 print HTML "</strong><br />\n";
1102 undef( $EmittedItem );
1107 if( $items_seen[$listlevel]++ == 0 ){
1108 push( @listend, "</$tag>" );
1109 print HTML "<$tag>\n";
1111 my $emitted = $tag eq 'dl' ? 'dt' : 'li';
1112 print HTML "<$emitted>";
1117 # process_item - convert a pod item tag and convert it to HTML format.
1121 my $need_dd = 0; # set to 1 if we need a <dd></dd> after an item
1123 # lots of documents start a list without doing an =over. this is
1124 # bad! but, the proper thing to do seems to be to just assume
1125 # they did do an =over. so warn them once and then continue.
1126 if( $listlevel == 0 ){
1127 warn "$0: $podfile: unexpected =item directive in paragraph $paragraph. ignoring.\n";
1131 # formatting: insert a paragraph if preceding item has >1 paragraph
1133 print HTML "<p></p>\n";
1137 # remove formatting instructions from the text
1138 my $text = depod( $otext );
1140 my $emitted; # the tag actually emitted, used for closing
1142 # all the list variants:
1143 if( $text =~ /\A\*/ ){ # bullet
1144 $emitted = emit_li( 'ul' );
1145 if ($text =~ /\A\*\s+(.+)\Z/s ) { # with additional text
1147 $otext =~ s/\A\*\s+//;
1148 emit_item_tag( $otext, $tag, 1 );
1151 } elsif( $text =~ /\A\d+/ ){ # numbered list
1152 $emitted = emit_li( 'ol' );
1153 if ($text =~ /\A(?>\d+\.?)\s*(.+)\Z/s ) { # with additional text
1155 $otext =~ s/\A\d+\.?\s*//;
1156 emit_item_tag( $otext, $tag, 1 );
1159 } else { # definition list
1160 $emitted = emit_li( 'dl' );
1161 if ($text =~ /\A(.+)\Z/s ){ # should have text
1162 emit_item_tag( $otext, $text, 1 );
1166 print HTML "</$emitted>" if $emitted;
1172 # process_over - process a pod over tag and start a corresponding HTML list.
1177 push( @items_seen, 0 );
1182 # process_back - process a pod back tag and convert it to HTML format.
1185 if( $listlevel == 0 ){
1186 warn "$0: $podfile: unexpected =back directive in paragraph $paragraph. ignoring.\n";
1190 # close off the list. note, I check to see if $listend[$listlevel] is
1191 # defined because an =item directive may have never appeared and thus
1192 # $listend[$listlevel] may have never been initialized.
1194 if( defined $listend[$listlevel] ){
1195 print HTML '<p></p>' if $after_lpar;
1196 print HTML $listend[$listlevel];
1202 # clean up item count
1207 # process_cut - process a pod cut tag, thus start ignoring pod directives.
1214 # process_pod - process a pod tag, thus stop ignoring pod directives
1215 # until we see a corresponding cut.
1218 # no need to set $ignore to 0 cause the main loop did it
1222 # process_for - process a =for pod tag. if it's for html, spit
1223 # it out verbatim, if illustration, center it, otherwise ignore it.
1226 my($whom, $text) = @_;
1227 if ( $whom =~ /^(pod2)?html$/i) {
1229 } elsif ($whom =~ /^illustration$/i) {
1230 1 while chomp $text;
1231 for my $ext (qw[.png .gif .jpeg .jpg .tga .pcl .bmp]) {
1232 $text .= $ext, last if -r "$text$ext";
1234 print HTML qq{<p align="center"><img src="$text" alt="$text illustration" /></p>};
1239 # process_begin - process a =begin pod tag. this pushes
1240 # whom we're beginning on the begin stack. if there's a
1241 # begin stack, we only print if it us.
1244 my($whom, $text) = @_;
1246 push (@begin_stack, $whom);
1247 if ( $whom =~ /^(pod2)?html$/) {
1248 print HTML $text if $text;
1253 # process_end - process a =end pod tag. pop the
1254 # begin stack. die if we're mismatched.
1257 my($whom, $text) = @_;
1259 if ($begin_stack[-1] ne $whom ) {
1260 die "Unmatched begin/end at chunk $paragraph\n"
1262 pop( @begin_stack );
1266 # process_pre - indented paragraph, made into <pre></pre>
1275 # insert spaces in place of tabs
1278 1 while $line =~ s/(\t+)/' ' x ((length($1) * 8) - $-[0] % 8)/e;
1282 # convert some special chars to HTML escapes
1283 $rest = html_escape($rest);
1285 # try and create links for all occurrences of perl.* within
1286 # the preformatted text.
1290 if ( defined $pages{$2} ){ # is a link
1291 qq($1<a href="$htmlroot/$pages{$2}">$2</a>);
1292 } elsif (defined $pages{dosify($2)}) { # is a link
1293 qq($1<a href="$htmlroot/$pages{dosify($2)}">$2</a>);
1299 (<a\ href="?) ([^>:]*:)? ([^>:]*) \.pod: ([^>:]*:)?
1302 if ( $htmlfileurl ne '' ){
1303 # Here, we take advantage of the knowledge
1304 # that $htmlfileurl ne '' implies $htmlroot eq ''.
1305 # Since $htmlroot eq '', we need to prepend $htmldir
1306 # on the fron of the link to get the absolute path
1307 # of the link's target. We check for a leading '/'
1308 # to avoid corrupting links that are #, file:, etc.
1310 $old_url = "$htmldir$old_url" if $old_url =~ m{^\/};
1311 $url = relativize_url( "$old_url.html", $htmlfileurl );
1318 # Look for embedded URLs and make them into links. We don't
1319 # relativize them since they are best left as the author intended.
1321 my $urls = '(' . join ('|', qw{
1334 my $gunk = '/#~:.?+=&%@!\-';
1335 my $punc = '.:!?\-;';
1336 my $any = "${ltrs}${gunk}${punc}";
1339 \b # start at word boundary
1341 $urls : # need resource and a colon
1342 (?!:) # Ignore File::, among others.
1343 [$any] +? # followed by one or more of any valid
1344 # character, but be conservative and
1345 # take only what you need to....
1347 (?= # look-ahead non-consumptive assertion
1348 [$punc]* # either 0 or more punctuation
1350 [^$any] # by a non-url char
1352 $ # end of the string
1355 $ # then end of the string
1357 }{<a href="$1">$1</a>}igox;
1359 # text should be as it is (verbatim)
1365 # pure text processing
1367 # pure_text/inIS_text: differ with respect to automatic C<> recognition.
1368 # we don't want this to happen within IS
1372 process_puretext( $text, \$ptQuote, 1 );
1377 process_puretext( $text, \$ptQuote, 0 );
1381 # process_puretext - process pure text (without pod-escapes) converting
1382 # double-quotes and handling implicit C<> links.
1384 sub process_puretext {
1385 my($text, $quote, $notinIS) = @_;
1387 ## Guessing at func() or [$@%&]*var references in plain text is destined
1388 ## to produce some strange looking ref's. uncomment to disable:
1391 my(@words, $lead, $trail);
1393 # convert double-quotes to single-quotes
1394 if( $$quote && $text =~ s/"/''/s ){
1397 while ($text =~ s/"([^"]*)"/``$1''/sg) {};
1398 $$quote = 1 if $text =~ s/"/``/s;
1400 # keep track of leading and trailing white-space
1401 $lead = ($text =~ s/\A(\s+)//s ? $1 : "");
1402 $trail = ($text =~ s/(\s+)\Z//s ? $1 : "");
1404 # split at space/non-space boundaries
1405 @words = split( /(?<=\s)(?=\S)|(?<=\S)(?=\s)/, $text );
1407 # process each word individually
1408 foreach my $word (@words) {
1410 next if $word =~ /^\s*$/;
1411 # see if we can infer a link
1412 if( $notinIS && $word =~ /^(\w+)\((.*)\)$/ ) {
1413 # has parenthesis so should have been a C<> ref
1414 ## try for a pagename (perlXXX(1))?
1415 my( $func, $args ) = ( $1, $2 );
1416 if( $args =~ /^\d+$/ ){
1417 my $url = page_sect( $word, '' );
1419 $word = "<a href=\"$url\">the $word manpage</a>";
1423 ## try function name for a link, append tt'ed argument list
1424 $word = emit_C( $func, '', "($args)");
1426 #### disabled. either all (including $\W, $\w+{.*} etc.) or nothing.
1427 ## } elsif( $notinIS && $word =~ /^[\$\@%&*]+\w+$/) {
1428 ## # perl variables, should be a C<> ref
1429 ## $word = emit_C( $word );
1431 } elsif ($word =~ m,^\w+://\w,) {
1433 # Don't relativize it: leave it as the author intended
1434 $word = qq(<a href="$word">$word</a>);
1435 } elsif ($word =~ /[\w.-]+\@[\w-]+\.\w/) {
1436 # looks like an e-mail address
1437 my ($w1, $w2, $w3) = ("", $word, "");
1438 ($w1, $w2, $w3) = ("(", $1, ")$2") if $word =~ /^\((.*?)\)(,?)/;
1439 ($w1, $w2, $w3) = ("<", $1, ">$2") if $word =~ /^<(.*?)>(,?)/;
1440 $word = qq($w1<a href="mailto:$w2">$w2</a>$w3);
1442 $word = html_escape($word) if $word =~ /["&<>]/;
1446 # put everything back together
1447 return $lead . join( '', @words ) . $trail;
1452 # process_text - handles plaintext that appears in the input pod file.
1453 # there may be pod commands embedded within the text so those must be
1454 # converted to html commands.
1457 sub process_text1($$;$$);
1458 sub pattern ($) { $_[0] ? '[^\S\n]+'.('>' x ($_[0] + 1)) : '>' }
1459 sub closing ($) { local($_) = shift; (defined && s/\s+$//) ? length : 0 }
1464 my $res = process_text1( 0, $tref );
1468 sub process_text1($$;$$){
1469 my( $lev, $rstr, $func, $closing ) = @_;
1472 unless (defined $func) {
1478 # B<text> - boldface
1479 $res = '<strong>' . process_text1( $lev, $rstr ) . '</strong>';
1481 } elsif( $func eq 'C' ){
1482 # C<code> - can be a ref or <code></code>
1483 # need to extract text
1484 my $par = go_ahead( $rstr, 'C', $closing );
1486 ## clean-up of the link target
1487 my $text = depod( $par );
1489 ### my $x = $par =~ /[BI]</ ? 'yes' : 'no' ;
1490 ### print STDERR "-->call emit_C($par) lev=$lev, par with BI=$x\n";
1492 $res = emit_C( $text, $lev > 1 || ($par =~ /[BI]</) );
1494 } elsif( $func eq 'E' ){
1495 # E<x> - convert to character
1496 $$rstr =~ s/^([^>]*)>//;
1498 $escape =~ s/^(\d+|X[\dA-F]+)$/#$1/i;
1501 } elsif( $func eq 'F' ){
1502 # F<filename> - italizice
1503 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1505 } elsif( $func eq 'I' ){
1506 # I<text> - italizice
1507 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1509 } elsif( $func eq 'L' ){
1511 ## L<text|cross-ref> => produce text, use cross-ref for linking
1512 ## L<cross-ref> => make text from cross-ref
1513 ## need to extract text
1514 my $par = go_ahead( $rstr, 'L', $closing );
1516 # some L<>'s that shouldn't be:
1517 # a) full-blown URL's are emitted as-is
1518 if( $par =~ m{^\w+://}s ){
1519 return make_URL_href( $par );
1521 # b) C<...> is stripped and treated as C<>
1522 if( $par =~ /^C<(.*)>$/ ){
1523 my $text = depod( $1 );
1524 return emit_C( $text, $lev > 1 || ($par =~ /[BI]</) );
1527 # analyze the contents
1528 $par =~ s/\n/ /g; # undo word-wrapped tags
1531 if( $par =~ s{^([^|]+)\|}{} ){
1535 # make sure sections start with a /
1538 my( $page, $section, $ident );
1540 # check for link patterns
1541 if( $par =~ m{^([^/]+?)/(?!")(.*?)$} ){ # name/ident
1542 # we've got a name/ident (no quotes)
1543 ( $page, $ident ) = ( $1, $2 );
1544 ### print STDERR "--> L<$par> to page $page, ident $ident\n";
1546 } elsif( $par =~ m{^(.*?)/"?(.*?)"?$} ){ # [name]/"section"
1547 # even though this should be a "section", we go for ident first
1548 ( $page, $ident ) = ( $1, $2 );
1549 ### print STDERR "--> L<$par> to page $page, section $section\n";
1551 } elsif( $par =~ /\s/ ){ # this must be a section with missing quotes
1552 ( $page, $section ) = ( '', $par );
1553 ### print STDERR "--> L<$par> to void page, section $section\n";
1556 ( $page, $section ) = ( $par, '' );
1557 ### print STDERR "--> L<$par> to page $par, void section\n";
1560 # now, either $section or $ident is defined. the convoluted logic
1561 # below tries to resolve L<> according to what the user specified.
1562 # failing this, we try to find the next best thing...
1563 my( $url, $ltext, $fid );
1566 if( defined $ident ){
1567 ## try to resolve $ident as an item
1568 ( $url, $fid ) = coderef( $page, $ident );
1570 if( ! defined( $linktext ) ){
1572 $linktext .= " in " if $ident && $page;
1573 $linktext .= "the $page manpage" if $page;
1575 ### print STDERR "got coderef url=$url\n";
1578 ## no luck: go for a section (auto-quoting!)
1581 ## now go for a section
1582 my $htmlsection = htmlify( $section );
1583 $url = page_sect( $page, $htmlsection );
1585 if( ! defined( $linktext ) ){
1586 $linktext = $section;
1587 $linktext .= " in " if $section && $page;
1588 $linktext .= "the $page manpage" if $page;
1590 ### print STDERR "got page/section url=$url\n";
1593 ## no luck: go for an ident
1600 ( $url, $fid ) = coderef( $page, $ident );
1602 if( ! defined( $linktext ) ){
1604 $linktext .= " in " if $ident && $page;
1605 $linktext .= "the $page manpage" if $page;
1607 ### print STDERR "got section=>coderef url=$url\n";
1611 # warning; show some text.
1612 $linktext = $opar unless defined $linktext;
1613 warn "$0: $podfile: cannot resolve L<$opar> in paragraph $paragraph.\n";
1616 # now we have a URL or just plain code
1617 $$rstr = $linktext . '>' . $$rstr;
1618 if( defined( $url ) ){
1619 $res = "<a href=\"$url\">" . process_text1( $lev, $rstr ) . '</a>';
1621 $res = '<em>' . process_text1( $lev, $rstr ) . '</em>';
1624 } elsif( $func eq 'S' ){
1625 # S<text> - non-breaking spaces
1626 $res = process_text1( $lev, $rstr );
1627 $res =~ s/ / /g;
1629 } elsif( $func eq 'X' ){
1631 $$rstr =~ s/^[^>]*>//;
1633 } elsif( $func eq 'Z' ){
1635 warn "$0: $podfile: invalid X<> in paragraph $paragraph.\n"
1636 unless $$rstr =~ s/^>//;
1639 my $term = pattern $closing;
1640 while( $$rstr =~ s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|$term)//s ){
1641 # all others: either recurse into new function or
1642 # terminate at closing angle bracket(s)
1644 $pt .= $2 if !$3 && $lev == 1;
1645 $res .= $lev == 1 ? pure_text( $pt ) : inIS_text( $pt );
1646 return $res if !$3 && $lev > 1;
1648 $res .= process_text1( $lev, $rstr, $3, closing $4 );
1652 $res .= pure_text( $$rstr );
1654 warn "$0: $podfile: undelimited $func<> in paragraph $paragraph.\n";
1661 # go_ahead: extract text of an IS (can be nested)
1664 my( $rstr, $func, $closing ) = @_;
1666 my @closing = ($closing);
1668 s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|@{[pattern $closing[0]]})//s ){
1672 return $res unless @closing;
1674 unshift @closing, closing $4;
1678 warn "$0: $podfile: undelimited $func<> in paragraph $paragraph.\n";
1683 # emit_C - output result of C<text>
1684 # $text is the depod-ed text
1687 my( $text, $nocode, $args ) = @_;
1688 $args = '' unless defined $args;
1690 my( $url, $fid ) = coderef( undef(), $text );
1692 # need HTML-safe text
1693 my $linktext = html_escape( "$text$args" );
1695 if( defined( $url ) &&
1696 (!defined( $EmittedItem ) || $EmittedItem ne $fid ) ){
1697 $res = "<a href=\"$url\"><code>$linktext</code></a>";
1698 } elsif( 0 && $nocode ){
1701 $res = "<code>$linktext</code>";
1707 # html_escape: make text safe for HTML
1711 $rest =~ s/&/&/g;
1712 $rest =~ s/</</g;
1713 $rest =~ s/>/>/g;
1714 $rest =~ s/"/"/g;
1715 # ' is only in XHTML, not HTML4. Be conservative
1716 #$rest =~ s/'/'/g;
1722 # dosify - convert filenames to 8.3
1726 return lc($str) if $^O eq 'VMS'; # VMS just needs casing
1729 $str =~ s/(\.\w+)/substr ($1,0,4)/ge;
1730 $str =~ s/(\w+)/substr ($1,0,8)/ge;
1736 # page_sect - make a URL from the text of a L<>
1739 my( $page, $section ) = @_;
1740 my( $linktext, $page83, $link); # work strings
1742 # check if we know that this is a section in this page
1743 if (!defined $pages{$page} && defined $sections{$page}) {
1746 ### print STDERR "reset page='', section=$section\n";
1749 $page83=dosify($page);
1750 $page=$page83 if (defined $pages{$page83});
1752 $link = "#" . anchorify( $section );
1753 } elsif ( $page =~ /::/ ) {
1755 # Search page cache for an entry keyed under the html page name,
1756 # then look to see what directory that page might be in. NOTE:
1757 # this will only find one page. A better solution might be to produce
1758 # an intermediate page that is an index to all such pages.
1759 my $page_name = $page ;
1760 $page_name =~ s,^.*/,,s ;
1761 if ( defined( $pages{ $page_name } ) &&
1762 $pages{ $page_name } =~ /([^:]*$page)\.(?:pod|pm):/
1767 # NOTE: This branch assumes that all A::B pages are located in
1768 # $htmlroot/A/B.html . This is often incorrect, since they are
1769 # often in $htmlroot/lib/A/B.html or such like. Perhaps we could
1770 # analyze the contents of %pages and figure out where any
1771 # cousins of A::B are, then assume that. So, if A::B isn't found,
1772 # but A::C is found in lib/A/C.pm, then A::B is assumed to be in
1773 # lib/A/B.pm. This is also limited, but it's an improvement.
1774 # Maybe a hints file so that the links point to the correct places
1778 $link = "$htmlroot/$page.html";
1779 $link .= "#" . anchorify( $section ) if ($section);
1780 } elsif (!defined $pages{$page}) {
1783 $section = anchorify( $section ) if $section ne "";
1784 ### print STDERR "...section=$section\n";
1786 # if there is a directory by the name of the page, then assume that an
1787 # appropriate section will exist in the subdirectory
1788 # if ($section ne "" && $pages{$page} =~ /([^:]*[^(\.pod|\.pm)]):/) {
1789 if ($section ne "" && $pages{$page} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
1790 $link = "$htmlroot/$1/$section.html";
1791 ### print STDERR "...link=$link\n";
1793 # since there is no directory by the name of the page, the section will
1794 # have to exist within a .html of the same name. thus, make sure there
1795 # is a .pod or .pm that might become that .html
1797 $section = "#$section" if $section;
1798 ### print STDERR "...section=$section\n";
1800 # check if there is a .pod with the page name
1801 if ($pages{$page} =~ /([^:]*)\.pod:/) {
1802 $link = "$htmlroot/$1.html$section";
1803 } elsif ($pages{$page} =~ /([^:]*)\.pm:/) {
1804 $link = "$htmlroot/$1.html$section";
1812 # Here, we take advantage of the knowledge that $htmlfileurl ne ''
1813 # implies $htmlroot eq ''. This means that the link in question
1814 # needs a prefix of $htmldir if it begins with '/'. The test for
1815 # the initial '/' is done to avoid '#'-only links, and to allow
1816 # for other kinds of links, like file:, ftp:, etc.
1818 if ( $htmlfileurl ne '' ) {
1819 $link = "$htmldir$link" if $link =~ m{^/}s;
1820 $url = relativize_url( $link, $htmlfileurl );
1821 # print( " b: [$link,$htmlfileurl,$url]\n" );
1834 # relativize_url - convert an absolute URL to one relative to a base URL.
1835 # Assumes both end in a filename.
1837 sub relativize_url {
1838 my ($dest,$source) = @_ ;
1840 my ($dest_volume,$dest_directory,$dest_file) =
1841 File::Spec::Unix->splitpath( $dest ) ;
1842 $dest = File::Spec::Unix->catpath( $dest_volume, $dest_directory, '' ) ;
1844 my ($source_volume,$source_directory,$source_file) =
1845 File::Spec::Unix->splitpath( $source ) ;
1846 $source = File::Spec::Unix->catpath( $source_volume, $source_directory, '' ) ;
1849 if ( $dest ne '' ) {
1850 $rel_path = File::Spec::Unix->abs2rel( $dest, $source ) ;
1853 if ( $rel_path ne '' &&
1854 substr( $rel_path, -1 ) ne '/' &&
1855 substr( $dest_file, 0, 1 ) ne '#'
1857 $rel_path .= "/$dest_file" ;
1860 $rel_path .= "$dest_file" ;
1868 # coderef - make URL from the text of a C<>
1871 my( $page, $item ) = @_;
1874 my $fid = fragment_id( $item );
1875 if( defined( $page ) ){
1876 # we have been given a $page...
1879 # Do we take it? Item could be a section!
1880 my $base = $items{$fid} || "";
1881 $base =~ s{[^/]*/}{};
1882 if( $base ne "$page.html" ){
1883 ### print STDERR "coderef( $page, $item ): items{$fid} = $items{$fid} = $base => discard page!\n";
1888 # no page - local items precede cached items
1889 if( defined( $fid ) ){
1890 if( exists $local_items{$fid} ){
1891 $page = $local_items{$fid};
1893 $page = $items{$fid};
1898 # if there was a pod file that we found earlier with an appropriate
1899 # =item directive, then create a link to that page.
1900 if( defined $page ){
1902 if( exists $pages{$page} and $pages{$page} =~ /([^:.]*)\.[^:]*:/){
1903 $page = $1 . '.html';
1905 my $link = "$htmlroot/$page#item_" . anchorify($fid);
1907 # Here, we take advantage of the knowledge that $htmlfileurl
1908 # ne '' implies $htmlroot eq ''.
1909 if ( $htmlfileurl ne '' ) {
1910 $link = "$htmldir$link" ;
1911 $url = relativize_url( $link, $htmlfileurl ) ;
1916 $url = "#item_" . anchorify($fid);
1919 confess "url has space: $url" if $url =~ /"[^"]*\s[^"]*"/;
1921 return( $url, $fid );
1927 # Adapted from Nick Ing-Simmons' PodToHtml package.
1929 my $source_file = shift ;
1930 my $destination_file = shift;
1932 my $source = URI::file->new_abs($source_file);
1933 my $uo = URI::file->new($destination_file,$source)->abs;
1934 return $uo->rel->as_string;
1939 # finish_list - finish off any pending HTML lists. this should be called
1940 # after the entire pod file has been read and converted.
1943 while ($listlevel > 0) {
1944 print HTML "</dl>\n";
1950 # htmlify - converts a pod section specification to a suitable section
1951 # specification for HTML. Note that we keep spaces and special characters
1952 # except ", ? (Netscape problem) and the hyphen (writer's problem...).
1956 $heading =~ s/(\s+)/ /g;
1957 $heading =~ s/\s+\Z//;
1958 $heading =~ s/\A\s+//;
1959 # The hyphen is a disgrace to the English language.
1960 $heading =~ s/[-"?]//g;
1961 $heading = lc( $heading );
1966 # similar to htmlify, but turns spaces into underscores
1970 $anchor = htmlify($anchor);
1971 $anchor =~ s/\s/_/g; # fixup spaces left by htmlify
1976 # depod - convert text by eliminating all interior sequences
1977 # Note: can be called with copy or modify semantics
1984 $E2c{amp} = '&'; # in Tk's pods
1992 ${$_[0]} = depod1( \$string );
2000 my( $rstr, $func, $closing ) = @_;
2002 return $res unless defined $$rstr;
2003 if( ! defined( $func ) ){
2004 # skip to next begin of an interior sequence
2005 while( $$rstr =~ s/\A(.*?)([BCEFILSXZ])<(<+[^\S\n]+)?// ){
2006 # recurse into its text
2007 $res .= $1 . depod1( $rstr, $2, closing $3);
2010 } elsif( $func eq 'E' ){
2011 # E<x> - convert to character
2012 $$rstr =~ s/^([^>]*)>//;
2013 $res .= $E2c{$1} || "";
2014 } elsif( $func eq 'X' ){
2016 $$rstr =~ s/^[^>]*>//;
2017 } elsif( $func eq 'Z' ){
2021 # all others: either recurse into new function or
2022 # terminate at closing angle bracket
2023 my $term = pattern $closing;
2024 while( $$rstr =~ s/\A(.*?)(([BCEFILSXZ])<(<+[^\S\n]+)?|$term)// ){
2027 $res .= depod1( $rstr, $3, closing $4 );
2029 ## If we're here and $2 ne '>': undelimited interior sequence.
2030 ## Ignored, as this is called without proper indication of where we are.
2031 ## Rely on process_text to produce diagnostics.
2037 # fragment_id - construct a fragment identifier from:
2039 # b) contents of C<...>
2044 $text =~ s/\s+\Z//s;
2046 # a method or function?
2047 return $1 if $text =~ /(\w+)\s*\(/;
2048 return $1 if $text =~ /->\s*(\w+)\s*\(?/;
2051 return $1 if $text =~ /^([$@%*]\S+)/;
2053 # some pattern matching operator?
2054 return $1 if $text =~ m|^(\w+/).*/\w*$|;
2056 # fancy stuff... like "do { }"
2057 return $1 if $text =~ m|^(\w+)\s*{.*}$|;
2059 # honour the perlfunc manpage: func [PAR[,[ ]PAR]...]
2060 # and some funnies with ... Module ...
2061 return $1 if $text =~ m{^([a-z\d]+)(\s+[A-Z\d,/& ]+)?$};
2062 return $1 if $text =~ m{^([a-z\d]+)\s+Module(\s+[A-Z\d,/& ]+)?$};
2065 $text =~ s/\s+/_/sg;
2067 defined( $hc[ord($1)] ) ? $hc[ord($1)]
2068 : ( $hc[ord($1)] = sprintf( "%%%02X", ord($1) ) ) }gxe;
2069 $text = substr( $text, 0, 50 );
2076 # make_URL_href - generate HTML href from URL
2077 # Special treatment for CGI queries.
2079 sub make_URL_href($){
2082 s{^(http:[-\w/#~:.+=&%@!]+)(\?.*)$}{<a href="$1$2">$1</a>}i ){
2083 $url = "<a href=\"$url\">$url</a>";