4 use Getopt::Long; # package for handling command-line parameters
10 @EXPORT = qw(pod2html htmlify);
15 use locale; # make \w work right in non-ASCII lands
23 Pod::Html - module to convert pod files to HTML
32 Converts files from pod format (see L<perlpod>) to HTML format. It
33 can automatically generate indexes and cross-references, and it keeps
34 a cache of things it knows how to cross-reference.
38 Pod::Html takes the following arguments:
46 Displays the usage message.
52 Sets the directory in which the resulting HTML file is placed. This
53 is used to generate relative links to other files. Not passing this
54 causes all links to be absolute, since this is the value that tells
55 Pod::Html the root of the documentation tree.
61 Sets the base URL for the HTML files. When cross-references are made,
62 the HTML root is prepended to the URL.
68 Specify the pod file to convert. Input is taken from STDIN if no
75 Specify the HTML file to create. Output goes to STDOUT if no outfile
82 Specify the base directory for finding library pods.
86 --podpath=name:...:name
88 Specify which subdirectories of the podroot contain pod files whose
89 HTML converted forms can be linked-to in cross-references.
93 --libpods=name:...:name
95 List of page names (eg, "perlfunc") which contain linkable C<=item>s.
101 Use Netscape HTML directives when applicable.
107 Do not use Netscape HTML directives (default).
113 Generate an index at the top of the HTML file (default behaviour).
119 Do not generate an index at the top of the HTML file.
126 Recurse into subdirectories specified in podpath (default behaviour).
132 Do not recurse into subdirectories specified in podpath.
138 Specify the title of the resulting HTML file.
144 Specify the URL of a cascading style sheet.
150 Display progress messages.
156 Don't display I<mostly harmless> warning messages.
163 "--podpath=lib:ext:pod:vms",
164 "--podroot=/usr/src/perl",
165 "--htmlroot=/perl/nmanual",
166 "--libpods=perlfunc:perlguts:perlvar:perlrun:perlop",
169 "--outfile=/perl/nmanual/foo.html");
173 Uses $Config{pod2html} to setup default options.
177 Tom Christiansen, E<lt>tchrist@perl.comE<gt>.
181 Has trouble with C<> etc in = commands.
189 This program is distributed under the Artistic License.
193 my $cache_ext = $^O eq 'VMS' ? ".tmp" : ".x~~";
194 my $dircache = "pod2htmd$cache_ext";
195 my $itemcache = "pod2htmi$cache_ext";
197 my @begin_stack = (); # begin/end stack
199 my @libpods = (); # files to search for links from C<> directives
200 my $htmlroot = "/"; # http-server base directory from which all
201 # relative paths in $podpath stem.
202 my $htmldir = ""; # The directory to which the html pages
203 # will (eventually) be written.
204 my $htmlfile = ""; # write to stdout by default
205 my $htmlfileurl = "" ; # The url that other files would use to
206 # refer to this file. This is only used
207 # to make relative urls that point to
209 my $podfile = ""; # read from stdin by default
210 my @podpath = (); # list of directories containing library pods.
211 my $podroot = "."; # filesystem base directory from which all
212 # relative paths in $podpath stem.
213 my $css = ''; # Cascading style sheet
214 my $recurse = 1; # recurse on subdirectories in $podpath.
215 my $quiet = 0; # not quiet by default
216 my $verbose = 0; # not verbose by default
217 my $doindex = 1; # non-zero if we should generate an index
218 my $listlevel = 0; # current list depth
219 my @listitem = (); # stack of HTML commands to use when a =item is
220 # encountered. the top of the stack is the
222 my @listdata = (); # similar to @listitem, but for the text after
224 my @listend = (); # similar to @listitem, but the text to use to
226 my $ignore = 1; # whether or not to format text. we don't
227 # format text until we hit our first pod
230 my %items_named = (); # for the multiples of the same item in perlfunc
232 my $netscape = 0; # whether or not to use netscape directives.
233 my $title; # title to give the pod(s)
234 my $header = 0; # produce block header/footer
235 my $top = 1; # true if we are at the top of the doc. used
236 # to prevent the first <HR> directive.
237 my $paragraph; # which paragraph we're processing (used
238 # for error messages)
239 my %pages = (); # associative array used to find the location
240 # of pages referenced by L<> links.
241 my %sections = (); # sections within this page
242 my %items = (); # associative array used to find the location
243 # of =item directives referenced by C<> links
244 my $Is83; # is dos with short filenames (8.3)
247 $dircache = "pod2html.d~~";
248 $itemcache = "pod2html.i~~";
250 @begin_stack = (); # begin/end stack
252 @libpods = (); # files to search for links from C<> directives
253 $htmlroot = "/"; # http-server base directory from which all
254 # relative paths in $podpath stem.
255 $htmlfile = ""; # write to stdout by default
256 $podfile = ""; # read from stdin by default
257 @podpath = (); # list of directories containing library pods.
258 $podroot = "."; # filesystem base directory from which all
259 # relative paths in $podpath stem.
260 $css = ''; # Cascading style sheet
261 $recurse = 1; # recurse on subdirectories in $podpath.
262 $quiet = 0; # not quiet by default
263 $verbose = 0; # not verbose by default
264 $doindex = 1; # non-zero if we should generate an index
265 $listlevel = 0; # current list depth
266 @listitem = (); # stack of HTML commands to use when a =item is
267 # encountered. the top of the stack is the
269 @listdata = (); # similar to @listitem, but for the text after
271 @listend = (); # similar to @listitem, but the text to use to
273 $ignore = 1; # whether or not to format text. we don't
274 # format text until we hit our first pod
279 $netscape = 0; # whether or not to use netscape directives.
280 $header = 0; # produce block header/footer
281 $title = ''; # title to give the pod(s)
282 $top = 1; # true if we are at the top of the doc. used
283 # to prevent the first <HR> directive.
284 $paragraph = ''; # which paragraph we're processing (used
285 # for error messages)
286 %sections = (); # sections within this page
288 # These are not reinitialised here but are kept as a cache.
289 # See get_cache and related cache management code.
290 #%pages = (); # associative array used to find the location
291 # of pages referenced by L<> links.
292 #%items = (); # associative array used to find the location
293 # of =item directives referenced by C<> links
304 $Is83 = 0 if (defined (&Dos::UseLFN) && Dos::UseLFN());
306 # cache of %pages and %items from last time we ran pod2html
308 #undef $opt_help if defined $opt_help;
310 # parse the command-line parameters
311 parse_command_line();
313 # set some variables to their default values if necessary
315 unless (@ARGV && $ARGV[0]) {
316 $podfile = "-" unless $podfile; # stdin
317 open(POD, "<$podfile")
318 || die "$0: cannot open $podfile file for input: $!\n";
320 $podfile = $ARGV[0]; # XXX: might be more filenames
323 $htmlfile = "-" unless $htmlfile; # stdout
324 $htmlroot = "" if $htmlroot eq "/"; # so we don't get a //
325 $htmldir =~ s#/$## ; # so we don't get a //
327 && defined( $htmldir )
329 && substr( $htmlfile, 0, length( $htmldir ) ) eq $htmldir
332 # Set the 'base' url for this file, so that we can use it
333 # as the location from which to calculate relative links
334 # to other files. If this is '', then absolute links will
335 # be used throughout.
336 $htmlfileurl= "$htmldir/" . substr( $htmlfile, length( $htmldir ) + 1);
339 # read the pod a paragraph at a time
340 warn "Scanning for sections in input file(s)\n" if $verbose;
345 # scan the pod for =head[1-6] directives and build an index
346 my $index = scan_headings(\%sections, @poddata);
349 warn "No pod in $podfile\n" if $verbose;
353 # open the output file
354 open(HTML, ">$htmlfile")
355 || die "$0: cannot open $htmlfile file for output: $!\n";
357 # put a title in the HTML file if one wasn't specified
360 for (my $i = 0; $i < @poddata; $i++) {
361 if ($poddata[$i] =~ /^=head1\s*NAME\b/m) {
362 for my $para ( @poddata[$i, $i+1] ) {
364 if ($title) = $para =~ /(\S+\s+-+.*\S)/s;
371 if (!$title and $podfile =~ /\.pod$/) {
372 # probably a split pod so take first =head[12] as title
373 for (my $i = 0; $i < @poddata; $i++) {
374 last if ($title) = $poddata[$i] =~ /^=head[12]\s*(.*)/;
376 warn "adopted '$title' as title for $podfile\n"
377 if $verbose and $title;
380 $title =~ s/\s*\(.*\)//;
382 warn "$0: no title for $podfile" unless $quiet;
383 $podfile =~ /^(.*)(\.[^.\/]+)?$/;
384 $title = ($podfile eq "-" ? 'No Title' : $1);
385 warn "using $title" if $verbose;
387 my $csslink = $css ? qq(\n<LINK REL="stylesheet" HREF="$css" TYPE="text/css">) : '';
388 $csslink =~ s,\\,/,g;
389 $csslink =~ s,(/.):,$1|,;
391 my $block = $header ? <<END_OF_BLOCK : '';
392 <TABLE BORDER=0 CELLPADDING=0 CELLSPACING=0 WIDTH=100%>
393 <TR><TD CLASS=block VALIGN=MIDDLE WIDTH=100% BGCOLOR="#cccccc">
394 <FONT SIZE=+1><STRONG><P CLASS=block> $title</P></STRONG></FONT>
399 print HTML <<END_OF_HEAD;
402 <TITLE>$title</TITLE>$csslink
403 <LINK REV="made" HREF="mailto:$Config{perladmin}">
410 # load/reload/validate/cache %pages and %items
411 get_cache($dircache, $itemcache, \@podpath, $podroot, $recurse);
413 # scan the pod for =item directives
414 scan_items("", \%items, @poddata);
416 # put an index at the top of the file. note, if $doindex is 0 we
417 # still generate an index, but surround it with an html comment.
418 # that way some other program can extract it if desired.
420 print HTML "<!-- INDEX BEGIN -->\n";
421 print HTML "<!--\n" unless $doindex;
423 print HTML "-->\n" unless $doindex;
424 print HTML "<!-- INDEX END -->\n\n";
425 print HTML "<HR>\n" if $doindex;
427 # now convert this file
428 warn "Converting input file\n" if $verbose;
429 foreach my $i (0..$#poddata) {
432 if (/^(=.*)/s) { # is it a pod directive?
435 if (/^=begin\s+(\S+)\s*(.*)/si) {# =begin
436 process_begin($1, $2);
437 } elsif (/^=end\s+(\S+)\s*(.*)/si) {# =end
439 } elsif (/^=cut/) { # =cut
441 } elsif (/^=pod/) { # =pod
444 next if @begin_stack && $begin_stack[-1] ne 'html';
446 if (/^=(head[1-6])\s+(.*\S)/s) { # =head[1-6] heading
447 process_head($1, $2);
448 } elsif (/^=item\s*(.*\S)/sm) { # =item text
450 } elsif (/^=over\s*(.*)/) { # =over N
452 } elsif (/^=back/) { # =back
454 } elsif (/^=for\s+(\S+)\s+(.*)/si) {# =for
458 warn "$0: $podfile: unknown pod directive '$1' in "
459 . "paragraph $paragraph. ignoring.\n";
466 next if @begin_stack && $begin_stack[-1] ne 'html';
468 process_text(\$text, 1);
469 print HTML "<P>\n$text</P>\n";
473 # finish off any pending directives
475 print HTML <<END_OF_TAIL;
482 # close the html file
485 warn "Finished\n" if $verbose;
488 ##############################################################################
490 my $usage; # see below
493 warn "$0: $podfile: @_\n" if @_;
497 $usage =<<END_OF_USAGE;
498 Usage: $0 --help --htmlroot=<name> --infile=<name> --outfile=<name>
499 --podpath=<name>:...:<name> --podroot=<name>
500 --libpods=<name>:...:<name> --recurse --verbose --index
501 --netscape --norecurse --noindex
503 --flush - flushes the item and directory caches.
504 --help - prints this message.
505 --htmlroot - http-server base directory from which all relative paths
506 in podpath stem (default is /).
507 --index - generate an index at the top of the resulting html
509 --infile - filename for the pod to convert (input taken from stdin
511 --libpods - colon-separated list of pages to search for =item pod
512 directives in as targets of C<> and implicit links (empty
513 by default). note, these are not filenames, but rather
514 page names like those that appear in L<> links.
515 --netscape - will use netscape html directives when applicable.
516 --nonetscape - will not use netscape directives (default).
517 --outfile - filename for the resulting html file (output sent to
519 --podpath - colon-separated list of directories containing library
520 pods. empty by default.
521 --podroot - filesystem base directory from which all relative paths
522 in podpath stem (default is .).
523 --noindex - don't generate an index at the top of the resulting html.
524 --norecurse - don't recurse on those subdirectories listed in podpath.
525 --recurse - recurse on those subdirectories listed in podpath
527 --title - title that will appear in resulting html file.
528 --header - produce block header/footer
529 --css - stylesheet URL
530 --verbose - self-explanatory
531 --quiet - supress some benign warning messages
535 sub parse_command_line {
536 my ($opt_flush,$opt_help,$opt_htmldir,$opt_htmlroot,$opt_index,$opt_infile,$opt_libpods,$opt_netscape,$opt_outfile,$opt_podpath,$opt_podroot,$opt_norecurse,$opt_recurse,$opt_title,$opt_verbose,$opt_css,$opt_header,$opt_quiet);
537 unshift @ARGV, split ' ', $Config{pod2html} if $Config{pod2html};
538 my $result = GetOptions(
539 'flush' => \$opt_flush,
540 'help' => \$opt_help,
541 'htmldir=s' => \$opt_htmldir,
542 'htmlroot=s' => \$opt_htmlroot,
543 'index!' => \$opt_index,
544 'infile=s' => \$opt_infile,
545 'libpods=s' => \$opt_libpods,
546 'netscape!' => \$opt_netscape,
547 'outfile=s' => \$opt_outfile,
548 'podpath=s' => \$opt_podpath,
549 'podroot=s' => \$opt_podroot,
550 'norecurse' => \$opt_norecurse,
551 'recurse!' => \$opt_recurse,
552 'title=s' => \$opt_title,
553 'header' => \$opt_header,
554 'css=s' => \$opt_css,
555 'verbose' => \$opt_verbose,
556 'quiet' => \$opt_quiet,
558 usage("-", "invalid parameters") if not $result;
560 usage("-") if defined $opt_help; # see if the user asked for help
561 $opt_help = ""; # just to make -w shut-up.
563 $podfile = $opt_infile if defined $opt_infile;
564 $htmlfile = $opt_outfile if defined $opt_outfile;
565 $htmldir = $opt_htmldir if defined $opt_outfile;
567 @podpath = split(":", $opt_podpath) if defined $opt_podpath;
568 @libpods = split(":", $opt_libpods) if defined $opt_libpods;
570 warn "Flushing item and directory caches\n"
571 if $opt_verbose && defined $opt_flush;
572 unlink($dircache, $itemcache) if defined $opt_flush;
574 $htmlroot = $opt_htmlroot if defined $opt_htmlroot;
575 $podroot = $opt_podroot if defined $opt_podroot;
577 $doindex = $opt_index if defined $opt_index;
578 $recurse = $opt_recurse if defined $opt_recurse;
579 $title = $opt_title if defined $opt_title;
580 $header = defined $opt_header ? 1 : 0;
581 $css = $opt_css if defined $opt_css;
582 $verbose = defined $opt_verbose ? 1 : 0;
583 $quiet = defined $opt_quiet ? 1 : 0;
584 $netscape = $opt_netscape if defined $opt_netscape;
591 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
592 my @cache_key_args = @_;
594 # A first-level cache:
595 # Don't bother reading the cache files if they still apply
596 # and haven't changed since we last read them.
598 my $this_cache_key = cache_key(@cache_key_args);
600 return if $saved_cache_key and $this_cache_key eq $saved_cache_key;
602 # load the cache of %pages and %items if possible. $tests will be
603 # non-zero if successful.
605 if (-f $dircache && -f $itemcache) {
606 warn "scanning for item cache\n" if $verbose;
607 $tests = load_cache($dircache, $itemcache, $podpath, $podroot);
610 # if we didn't succeed in loading the cache then we must (re)build
613 warn "scanning directories in pod-path\n" if $verbose;
614 scan_podpath($podroot, $recurse, 0);
616 $saved_cache_key = cache_key(@cache_key_args);
620 my($dircache, $itemcache, $podpath, $podroot, $recurse) = @_;
621 return join('!', $dircache, $itemcache, $recurse,
622 @$podpath, $podroot, stat($dircache), stat($itemcache));
626 # load_cache - tries to find if the caches stored in $dircache and $itemcache
627 # are valid caches of %pages and %items. if they are valid then it loads
628 # them and returns a non-zero value.
632 my($dircache, $itemcache, $podpath, $podroot) = @_;
638 open(CACHE, "<$itemcache") ||
639 die "$0: error opening $itemcache for reading: $!\n";
642 # is it the same podpath?
645 $tests++ if (join(":", @$podpath) eq $_);
647 # is it the same podroot?
650 $tests++ if ($podroot eq $_);
652 # load the cache if its good
658 warn "loading item cache\n" if $verbose;
665 warn "scanning for directory cache\n" if $verbose;
666 open(CACHE, "<$dircache") ||
667 die "$0: error opening $dircache for reading: $!\n";
671 # is it the same podpath?
674 $tests++ if (join(":", @$podpath) eq $_);
676 # is it the same podroot?
679 $tests++ if ($podroot eq $_);
681 # load the cache if its good
687 warn "loading directory cache\n" if $verbose;
699 # scan_podpath - scans the directories specified in @podpath for directories,
700 # .pod files, and .pm files. it also scans the pod files specified in
701 # @libpods for =item directives.
704 my($podroot, $recurse, $append) = @_;
706 my($libpod, $dirname, $pod, @files, @poddata);
713 # scan each directory listed in @podpath
716 || die "$0: error changing to directory $podroot: $!\n";
717 foreach $dir (@podpath) {
718 scan_dir($dir, $recurse);
721 # scan the pods listed in @libpods for =item directives
722 foreach $libpod (@libpods) {
723 # if the page isn't defined then we won't know where to find it
725 next unless defined $pages{$libpod} && $pages{$libpod};
727 # if there is a directory then use the .pod and .pm files within it.
728 # NOTE: Only finds the first so-named directory in the tree.
729 # if ($pages{$libpod} =~ /([^:]*[^(\.pod|\.pm)]):/) {
730 if ($pages{$libpod} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
731 # find all the .pod and .pm files within the directory
733 opendir(DIR, $dirname) ||
734 die "$0: error opening directory $dirname: $!\n";
735 @files = grep(/(\.pod|\.pm)$/ && ! -d $_, readdir(DIR));
738 # scan each .pod and .pm file for =item directives
739 foreach $pod (@files) {
740 open(POD, "<$dirname/$pod") ||
741 die "$0: error opening $dirname/$pod for input: $!\n";
745 scan_items("$dirname/$pod", @poddata);
748 # use the names of files as =item directives too.
749 foreach $pod (@files) {
750 $pod =~ /^(.*)(\.pod|\.pm)$/;
751 $items{$1} = "$dirname/$1.html" if $1;
753 } elsif ($pages{$libpod} =~ /([^:]*\.pod):/ ||
754 $pages{$libpod} =~ /([^:]*\.pm):/) {
755 # scan the .pod or .pm file for =item directives
757 open(POD, "<$pod") ||
758 die "$0: error opening $pod for input: $!\n";
762 scan_items("$pod", @poddata);
764 warn "$0: shouldn't be here (line ".__LINE__."\n";
767 @poddata = (); # clean-up a bit
770 || die "$0: error changing to directory $pwd: $!\n";
772 # cache the item list for later use
773 warn "caching items for later use\n" if $verbose;
774 open(CACHE, ">$itemcache") ||
775 die "$0: error open $itemcache for writing: $!\n";
777 print CACHE join(":", @podpath) . "\n$podroot\n";
778 foreach my $key (keys %items) {
779 print CACHE "$key $items{$key}\n";
784 # cache the directory list for later use
785 warn "caching directories for later use\n" if $verbose;
786 open(CACHE, ">$dircache") ||
787 die "$0: error open $dircache for writing: $!\n";
789 print CACHE join(":", @podpath) . "\n$podroot\n";
790 foreach my $key (keys %pages) {
791 print CACHE "$key $pages{$key}\n";
798 # scan_dir - scans the directory specified in $dir for subdirectories, .pod
799 # files, and .pm files. notes those that it finds. this information will
800 # be used later in order to figure out where the pages specified in L<>
801 # links are on the filesystem.
804 my($dir, $recurse) = @_;
805 my($t, @subdirs, @pods, $pod, $dirname, @dirs);
811 opendir(DIR, $dir) ||
812 die "$0: error opening directory $dir: $!\n";
813 while (defined($_ = readdir(DIR))) {
814 if (-d "$dir/$_" && $_ ne "." && $_ ne "..") { # directory
815 $pages{$_} = "" unless defined $pages{$_};
816 $pages{$_} .= "$dir/$_:";
818 } elsif (/\.pod$/) { # .pod
820 $pages{$_} = "" unless defined $pages{$_};
821 $pages{$_} .= "$dir/$_.pod:";
822 push(@pods, "$dir/$_.pod");
823 } elsif (/\.pm$/) { # .pm
825 $pages{$_} = "" unless defined $pages{$_};
826 $pages{$_} .= "$dir/$_.pm:";
827 push(@pods, "$dir/$_.pm");
832 # recurse on the subdirectories if necessary
834 foreach my $subdir (@subdirs) {
835 scan_dir("$dir/$subdir", $recurse);
841 # scan_headings - scan a pod file for head[1-6] tags, note the tags, and
845 my($sections, @data) = @_;
846 my($tag, $which_head, $title, $listdepth, $index);
848 # here we need local $ignore = 0;
849 # unfortunately, we can't have it, because $ignore is lexical
855 # scan for =head directives, note their name, and build an index
856 # pointing to each of them.
857 foreach my $line (@data) {
858 if ($line =~ /^=(head)([1-6])\s+(.*)/) {
859 ($tag,$which_head, $title) = ($1,$2,$3);
861 $$sections{htmlify(0,$title)} = 1;
863 while ($which_head != $listdepth) {
864 if ($which_head > $listdepth) {
865 $index .= "\n" . ("\t" x $listdepth) . "<UL>\n";
867 } elsif ($which_head < $listdepth) {
869 $index .= "\n" . ("\t" x $listdepth) . "</UL>\n";
873 $index .= "\n" . ("\t" x $listdepth) . "<LI>" .
874 "<A HREF=\"#" . htmlify(0,$title) . "\">" .
875 html_escape(process_text(\$title, 0)) . "</A></LI>";
879 # finish off the lists
880 while ($listdepth--) {
881 $index .= "\n" . ("\t" x $listdepth) . "</UL>\n";
884 # get rid of bogus lists
885 $index =~ s,\t*<UL>\s*</UL>\n,,g;
887 $ignore = 1; # restore old value;
893 # scan_items - scans the pod specified by $pod for =item directives. we
894 # will use this information later on in resolving C<> links.
897 my($pod, @poddata) = @_;
902 $pod .= ".html" if $pod;
904 foreach $i (0..$#poddata) {
907 # remove any formatting instructions
908 s,[A-Z]<([^<>]*)>,$1,g;
910 # figure out what kind of item it is and get the first word of
912 if (/^=item\s+(\w*)\s*.*$/s) {
913 if ($1 eq "*") { # bullet list
914 /\A=item\s+\*\s*(.*?)\s*\Z/s;
916 } elsif ($1 =~ /^\d+/) { # numbered list
917 /\A=item\s+\d+\.?(.*?)\s*\Z/s;
920 # /\A=item\s+(.*?)\s*\Z/s;
925 $items{$item} = "$pod" if $item;
931 # process_head - convert a pod head[1-6] tag and convert it to HTML format.
934 my($tag, $heading) = @_;
937 # figure out the level of the =head
938 $tag =~ /head([1-6])/;
941 # can't have a heading full of spaces and speechmarks and so on
942 $firstword = $heading; $firstword =~ s/\s*(\w+)\s.*/$1/;
944 print HTML "<P>\n" unless $listlevel;
945 print HTML "<HR>\n" unless $listlevel || $top;
946 print HTML "<H$level>"; # unless $listlevel;
947 #print HTML "<H$level>" unless $listlevel;
948 my $convert = $heading; process_text(\$convert, 0);
949 $convert = html_escape($convert);
950 print HTML '<A NAME="' . htmlify(0,$heading) . "\">$convert</A>";
951 print HTML "</H$level>"; # unless $listlevel;
956 # process_item - convert a pod item tag and convert it to HTML format.
960 my($i, $quote, $name);
962 my $need_preamble = 0;
966 # lots of documents start a list without doing an =over. this is
967 # bad! but, the proper thing to do seems to be to just assume
968 # they did do an =over. so warn them once and then continue.
969 warn "$0: $podfile: unexpected =item directive in paragraph $paragraph. ignoring.\n"
971 process_over() unless $listlevel;
973 return unless $listlevel;
975 # remove formatting instructions from the text
976 1 while $text =~ s/[A-Z]<([^<>]*)>/$1/g;
979 $need_preamble = $items_seen[$listlevel]++ == 0;
981 # check if this is the first =item after an =over
983 my $need_new = $listlevel >= @listitem;
985 if ($text =~ /\A\*/) { # bullet
987 if ($need_preamble) {
988 push(@listend, "</UL>");
993 if ($text =~ /\A\*\s*(.+)\Z/s) {
994 print HTML '<STRONG>';
995 if ($items_named{$1}++) {
996 print HTML html_escape($1);
998 my $name = 'item_' . htmlify(1,$1);
999 print HTML qq(<A NAME="$name">), html_escape($1), '</A>';
1001 print HTML '</STRONG>';
1004 } elsif ($text =~ /\A[\d#]+/) { # numbered list
1006 if ($need_preamble) {
1007 push(@listend, "</OL>");
1008 print HTML "<OL>\n";
1012 if ($text =~ /\A\d+\.?\s*(.+)\Z/s) {
1013 print HTML '<STRONG>';
1014 if ($items_named{$1}++) {
1015 print HTML html_escape($1);
1017 my $name = 'item_' . htmlify(0,$1);
1018 print HTML qq(<A NAME="$name">), html_escape($1), '</A>';
1020 print HTML '</STRONG>';
1023 } else { # all others
1025 if ($need_preamble) {
1026 push(@listend, '</DL>');
1027 print HTML "<DL>\n";
1031 if ($text =~ /(\S+)/) {
1032 print HTML '<STRONG>';
1033 if ($items_named{$1}++) {
1034 print HTML html_escape($text);
1036 my $name = 'item_' . htmlify(1,$text);
1037 print HTML qq(<A NAME="$name">), html_escape($text), '</A>';
1039 print HTML '</STRONG>';
1048 # process_over - process a pod over tag and start a corresponding HTML
1057 # process_back - process a pod back tag and convert it to HTML format.
1060 warn "$0: $podfile: unexpected =back directive in paragraph $paragraph. ignoring.\n"
1062 return unless $listlevel;
1064 # close off the list. note, I check to see if $listend[$listlevel] is
1065 # defined because an =item directive may have never appeared and thus
1066 # $listend[$listlevel] may have never been initialized.
1068 print HTML $listend[$listlevel] if defined $listend[$listlevel];
1071 # don't need the corresponding perl code anymore
1080 # process_cut - process a pod cut tag, thus stop ignoring pod directives.
1087 # process_pod - process a pod pod tag, thus ignore pod directives until we see a
1088 # corresponding cut.
1091 # no need to set $ignore to 0 cause the main loop did it
1095 # process_for - process a =for pod tag. if it's for html, split
1096 # it out verbatim, if illustration, center it, otherwise ignore it.
1099 my($whom, $text) = @_;
1100 if ( $whom =~ /^(pod2)?html$/i) {
1102 } elsif ($whom =~ /^illustration$/i) {
1103 1 while chomp $text;
1104 for my $ext (qw[.png .gif .jpeg .jpg .tga .pcl .bmp]) {
1105 $text .= $ext, last if -r "$text$ext";
1107 print HTML qq{<p align = "center"><img src = "$text" alt = "$text illustration"></p>};
1112 # process_begin - process a =begin pod tag. this pushes
1113 # whom we're beginning on the begin stack. if there's a
1114 # begin stack, we only print if it us.
1117 my($whom, $text) = @_;
1119 push (@begin_stack, $whom);
1120 if ( $whom =~ /^(pod2)?html$/) {
1121 print HTML $text if $text;
1126 # process_end - process a =end pod tag. pop the
1127 # begin stack. die if we're mismatched.
1130 my($whom, $text) = @_;
1132 if ($begin_stack[-1] ne $whom ) {
1133 die "Unmatched begin/end at chunk $paragraph\n"
1139 # process_text - handles plaintext that appears in the input pod file.
1140 # there may be pod commands embedded within the text so those must be
1141 # converted to html commands.
1144 my($text, $escapeQuotes) = @_;
1145 my($result, $rest, $s1, $s2, $s3, $s4, $match, $bf);
1146 my($podcommand, $params, $tag, $quote);
1150 $quote = 0; # status of double-quote conversion
1154 if ($rest =~ /^\s+/) { # preformatted text, no pod directives
1158 1 while $line =~ s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e;
1162 $rest =~ s/&/&/g;
1163 $rest =~ s/</</g;
1164 $rest =~ s/>/>/g;
1165 $rest =~ s/"/"/g;
1167 # try and create links for all occurrences of perl.* within
1168 # the preformatted text.
1172 if (defined $pages{$2}) { # is a link
1173 qq($1<A HREF="$htmlroot/$pages{$2}">$2</A>);
1174 } elsif (defined $pages{dosify($2)}) { # is a link
1175 qq($1<A HREF="$htmlroot/$pages{dosify($2)}">$2</A>);
1180 # $rest =~ s/(<A HREF=)([^>:]*:)?([^>:]*)\.pod:([^>:]*:)?/$1$3.html/g;
1182 (<A\ HREF="?) ([^>:]*:)? ([^>:]*) \.pod: ([^>:]*:)?
1185 if ( $htmlfileurl ne '' ) {
1186 # Here, we take advantage of the knowledge
1187 # that $htmlfileurl ne '' implies $htmlroot eq ''.
1188 # Since $htmlroot eq '', we need to prepend $htmldir
1189 # on the fron of the link to get the absolute path
1190 # of the link's target. We check for a leading '/'
1191 # to avoid corrupting links that are #, file:, etc.
1193 $old_url = "$htmldir$old_url"
1194 if ( $old_url =~ m{^\/} ) ;
1195 $url = relativize_url( "$old_url.html", $htmlfileurl );
1196 # print( " a: [$old_url.html,$htmlfileurl,$url]\n" ) ;
1204 # Look for embedded URLs and make them in to links. We don't
1205 # relativize them since they are best left as the author intended.
1206 my $urls = '(' . join ('|', qw{
1219 my $gunk = '/#~:.?+=&%@!\-';
1221 my $any = "${ltrs}${gunk}${punc}";
1224 \b # start at word boundary
1226 $urls : # need resource and a colon
1227 (?!:) # Ignore File::, among others.
1228 [$any] +? # followed by on or more
1229 # of any valid character, but
1230 # be conservative and take only
1231 # what you need to....
1233 (?= # look-ahead non-consumptive assertion
1234 [$punc]* # either 0 or more puntuation
1235 [^$any] # followed by a non-url char
1237 $ # then end of the string
1239 }{<A HREF="$1">$1</A>}igox;
1241 $result = "<PRE>" # text should be as it is (verbatim)
1244 } else { # formatted text
1245 # parse through the string, stopping each time we find a
1246 # pod-escape. once the string has been throughly processed
1248 while (length $rest) {
1249 # check to see if there are any possible pod directives in
1250 # the remaining part of the text.
1251 if ($rest =~ m/[BCEIFLSZ]</) {
1252 warn "\$rest\t= $rest\n" unless
1259 $s1 = $1; # pure text
1260 $s2 = $2; # the type of pod-escape that follows
1262 $s4 = $3; # the rest of the string
1270 if ($s3 eq '<' && $s2) { # a pod-escape
1271 $result .= ($escapeQuotes ? process_puretext($s1, \$quote) : $s1);
1272 $podcommand = "$s2<";
1275 # find the matching '>'
1278 while ($match && !$bf) {
1280 if ($rest =~ /\A([^<>]*[BCEIFLSZ]<)(.*)\Z/s) {
1285 } elsif ($rest =~ /\A([^>]*>)(.*)\Z/s) {
1295 $0: $podfile: cannot find matching > for $s2 in paragraph $paragraph.
1297 $result .= substr $podcommand, 0, 2;
1298 $rest = substr($podcommand, 2) . $rest;
1302 # pull out the parameters to the pod-escape
1303 $podcommand =~ /^([BCFEILSZ]?)<(.*)>$/s;
1307 # process the text within the pod-escape so that any escapes
1308 # which must occur do.
1309 process_text(\$params, 0) unless $tag eq 'L';
1312 if (!$tag || $tag eq " ") { # <> : no tag
1313 $s1 = "<$params>";
1314 } elsif ($tag eq "L") { # L<> : link
1315 $s1 = process_L($params);
1316 } elsif ($tag eq "I" || # I<> : italicize text
1317 $tag eq "B" || # B<> : bold text
1318 $tag eq "F") { # F<> : file specification
1319 $s1 = process_BFI($tag, $params);
1320 } elsif ($tag eq "C") { # C<> : literal code
1321 $s1 = process_C($params, 1);
1322 } elsif ($tag eq "E") { # E<> : escape
1323 $s1 = process_E($params);
1324 } elsif ($tag eq "Z") { # Z<> : zero-width character
1325 $s1 = process_Z($params);
1326 } elsif ($tag eq "S") { # S<> : non-breaking space
1327 $s1 = process_S($params);
1328 } elsif ($tag eq "X") { # S<> : non-breaking space
1329 $s1 = process_X($params);
1331 warn "$0: $podfile: unhandled tag '$tag' in paragraph $paragraph\n";
1336 # for pure text we must deal with implicit links and
1337 # double-quotes among other things.
1338 $result .= ($escapeQuotes ? process_puretext("$s1$s2$s3", \$quote) : "$s1$s2$s3");
1348 $rest =~ s/&(?!\w+;|#)/&/g; # XXX not bulletproof
1349 $rest =~ s/</</g;
1350 $rest =~ s/>/>/g;
1351 $rest =~ s/"/"/g;
1356 # process_puretext - process pure text (without pod-escapes) converting
1357 # double-quotes and handling implicit C<> links.
1359 sub process_puretext {
1360 my($text, $quote) = @_;
1361 my(@words, $result, $rest, $lead, $trail);
1363 # convert double-quotes to single-quotes
1364 $text =~ s/\A([^"]*)"/$1''/s if $$quote;
1365 while ($text =~ s/\A([^"]*)["]([^"]*)["]/$1``$2''/sg) {}
1367 $$quote = ($text =~ m/"/ ? 1 : 0);
1368 $text =~ s/\A([^"]*)"/$1``/s if $$quote;
1370 # keep track of leading and trailing white-space
1371 $lead = ($text =~ /\A(\s*)/s ? $1 : "");
1372 $trail = ($text =~ /(\s*)\Z/s ? $1 : "");
1374 # collapse all white space into a single space
1376 @words = split(" ", $text);
1378 # process each word individually
1379 foreach my $word (@words) {
1380 # see if we can infer a link
1381 if ($word =~ /^\w+\(/) {
1382 # has parenthesis so should have been a C<> ref
1383 $word = process_C($word);
1384 # $word =~ /^[^()]*]\(/;
1385 # if (defined $items{$1} && $items{$1}) {
1386 # $word = "\n<CODE><A HREF=\"$htmlroot/$items{$1}#item_"
1387 # . htmlify(0,$word)
1388 # . "\">$word</A></CODE>";
1389 # } elsif (defined $items{$word} && $items{$word}) {
1390 # $word = "\n<CODE><A HREF=\"$htmlroot/$items{$word}#item_"
1391 # . htmlify(0,$word)
1392 # . "\">$word</A></CODE>";
1394 # $word = "\n<CODE><A HREF=\"#item_"
1395 # . htmlify(0,$word)
1396 # . "\">$word</A></CODE>";
1398 } elsif ($word =~ /^[\$\@%&*]+\w+$/) {
1399 # perl variables, should be a C<> ref
1400 $word = process_C($word, 1);
1401 } elsif ($word =~ m,^\w+://\w,) {
1403 # Don't relativize it: leave it as the author intended
1404 $word = qq(<A HREF="$word">$word</A>);
1405 } elsif ($word =~ /[\w.-]+\@[\w-]+\.\w/) {
1406 # looks like an e-mail address
1407 my ($w1, $w2, $w3) = ("", $word, "");
1408 ($w1, $w2, $w3) = ("(", $1, ")$2") if $word =~ /^\((.*?)\)(,?)/;
1409 ($w1, $w2, $w3) = ("<", $1, ">$2") if $word =~ /^<(.*?)>(,?)/;
1410 $word = qq($w1<A HREF="mailto:$w2">$w2</A>$w3);
1411 } elsif ($word !~ /[a-z]/ && $word =~ /[A-Z]/) { # all uppercase?
1412 $word = html_escape($word) if $word =~ /["&<>]/;
1413 $word = "\n<FONT SIZE=-1>$word</FONT>" if $netscape;
1415 $word = html_escape($word) if $word =~ /["&<>]/;
1419 # build a new string based upon our conversion
1421 $rest = join(" ", @words);
1422 while (length($rest) > 75) {
1423 if ( $rest =~ m/^(.{0,75})\s(.*?)$/o ||
1424 $rest =~ m/^(\S*)\s(.*?)$/o) {
1429 $result .= "$rest\n";
1433 $result .= $rest if $rest;
1435 # restore the leading and trailing white-space
1436 $result = "$lead$result$trail";
1442 # pre_escape - convert & in text to $amp;
1446 $$str =~ s/&(?!\w+;|#)/&/g; # XXX not bulletproof
1450 # dosify - convert filenames to 8.3
1454 return lc($str) if $^O eq 'VMS'; # VMS just needs casing
1457 $str =~ s/(\.\w+)/substr ($1,0,4)/ge;
1458 $str =~ s/(\w+)/substr ($1,0,8)/ge;
1464 # process_L - convert a pod L<> directive to a corresponding HTML link.
1465 # most of the links made are inferred rather than known about directly
1466 # (i.e it's not known whether the =head\d section exists in the target file,
1467 # or whether a .pod file exists in the case of split files). however, the
1468 # guessing usually works.
1470 # Unlike the other directives, this should be called with an unprocessed
1471 # string, else tags in the link won't be matched.
1475 my($s1, $s2, $linktext, $page, $page83, $section, $link); # work strings
1477 $str =~ s/\n/ /g; # undo word-wrapped tags
1480 # LREF: a la HREF L<show this text|man/section>
1481 $linktext = $1 if s:^([^|]+)\|::;
1483 # make sure sections start with a /
1485 s,^,/,g if (!m,/, && / /);
1487 # check if there's a section specified
1488 if (m,^(.*?)/"?(.*?)"?$,) { # yes
1489 ($page, $section) = ($1, $2);
1491 ($page, $section) = ($str, "");
1494 # check if we know that this is a section in this page
1495 if (!defined $pages{$page} && defined $sections{$page}) {
1500 # remove trailing punctuation, like ()
1501 $section =~ s/\W*$// ;
1504 $page83=dosify($page);
1505 $page=$page83 if (defined $pages{$page83});
1507 $link = "#" . htmlify(0,$section);
1508 $linktext = $section unless defined($linktext);
1509 } elsif ( $page =~ /::/ ) {
1510 $linktext = ($section ? "$section" : "$page");
1512 # Search page cache for an entry keyed under the html page name,
1513 # then look to see what directory that page might be in. NOTE:
1514 # this will only find one page. A better solution might be to produce
1515 # an intermediate page that is an index to all such pages.
1516 my $page_name = $page ;
1517 $page_name =~ s,^.*/,, ;
1518 if ( defined( $pages{ $page_name } ) &&
1519 $pages{ $page_name } =~ /([^:]*$page)\.(?:pod|pm):/
1524 # NOTE: This branch assumes that all A::B pages are located in
1525 # $htmlroot/A/B.html . This is often incorrect, since they are
1526 # often in $htmlroot/lib/A/B.html or such like. Perhaps we could
1527 # analyze the contents of %pages and figure out where any
1528 # cousins of A::B are, then assume that. So, if A::B isn't found,
1529 # but A::C is found in lib/A/C.pm, then A::B is assumed to be in
1530 # lib/A/B.pm. This is also limited, but it's an improvement.
1531 # Maybe a hints file so that the links point to the correct places
1533 # Also, maybe put a warn "$0: cannot resolve..." here.
1535 $link = "$htmlroot/$page.html";
1536 $link .= "#" . htmlify(0,$section) if ($section);
1537 } elsif (!defined $pages{$page}) {
1538 warn "$0: $podfile: cannot resolve L<$str> in paragraph $paragraph: no such page '$page'\n" unless $quiet;
1540 $linktext = $page unless defined($linktext);
1542 $linktext = ($section ? "$section" : "the $page manpage") unless defined($linktext);
1543 $section = htmlify(0,$section) if $section ne "";
1545 # if there is a directory by the name of the page, then assume that an
1546 # appropriate section will exist in the subdirectory
1547 # if ($section ne "" && $pages{$page} =~ /([^:]*[^(\.pod|\.pm)]):/) {
1548 if ($section ne "" && $pages{$page} =~ /([^:]*(?<!\.pod)(?<!\.pm)):/) {
1549 $link = "$htmlroot/$1/$section.html";
1551 # since there is no directory by the name of the page, the section will
1552 # have to exist within a .html of the same name. thus, make sure there
1553 # is a .pod or .pm that might become that .html
1555 $section = "#$section";
1556 # check if there is a .pod with the page name
1557 if ($pages{$page} =~ /([^:]*)\.pod:/) {
1558 $link = "$htmlroot/$1.html$section";
1559 } elsif ($pages{$page} =~ /([^:]*)\.pm:/) {
1560 $link = "$htmlroot/$1.html$section";
1562 warn "$0: $podfile: cannot resolve L$str in paragraph $paragraph: ".
1563 "no .pod or .pm found\n";
1565 $linktext = $section unless defined($linktext);
1570 process_text(\$linktext, 0);
1572 # Here, we take advantage of the knowledge that $htmlfileurl ne ''
1573 # implies $htmlroot eq ''. This means that the link in question
1574 # needs a prefix of $htmldir if it begins with '/'. The test for
1575 # the initial '/' is done to avoid '#'-only links, and to allow
1576 # for other kinds of links, like file:, ftp:, etc.
1578 if ( $htmlfileurl ne '' ) {
1579 $link = "$htmldir$link"
1580 if ( $link =~ m{^/} ) ;
1582 $url = relativize_url( $link, $htmlfileurl ) ;
1583 # print( " b: [$link,$htmlfileurl,$url]\n" ) ;
1589 $s1 = "<A HREF=\"$url\">$linktext</A>";
1591 $s1 = "<EM>$linktext</EM>";
1597 # relativize_url - convert an absolute URL to one relative to a base URL.
1598 # Assumes both end in a filename.
1600 sub relativize_url {
1601 my ($dest,$source) = @_ ;
1603 my ($dest_volume,$dest_directory,$dest_file) =
1604 File::Spec::Unix->splitpath( $dest ) ;
1605 $dest = File::Spec::Unix->catpath( $dest_volume, $dest_directory, '' ) ;
1607 my ($source_volume,$source_directory,$source_file) =
1608 File::Spec::Unix->splitpath( $source ) ;
1609 $source = File::Spec::Unix->catpath( $source_volume, $source_directory, '' ) ;
1612 if ( $dest ne '' ) {
1613 $rel_path = File::Spec::Unix->abs2rel( $dest, $source ) ;
1616 if ( $rel_path ne '' &&
1617 substr( $rel_path, -1 ) ne '/' &&
1618 substr( $dest_file, 0, 1 ) ne '#'
1620 $rel_path .= "/$dest_file" ;
1623 $rel_path .= "$dest_file" ;
1630 # process_BFI - process any of the B<>, F<>, or I<> pod-escapes and
1631 # convert them to corresponding HTML directives.
1634 my($tag, $str) = @_;
1635 my($s1); # work string
1636 my(%repltext) = ( 'B' => 'STRONG',
1640 # extract the modified text and convert to HTML
1641 $s1 = "<$repltext{$tag}>$str</$repltext{$tag}>";
1646 # process_C - process the C<> pod-escape.
1649 my($str, $doref) = @_;
1653 $s1 =~ s/\([^()]*\)//g; # delete parentheses
1655 $s1 =~ s/\W//g; # delete bogus characters
1656 $str = html_escape($str);
1658 # if there was a pod file that we found earlier with an appropriate
1659 # =item directive, then create a link to that page.
1660 if ($doref && defined $items{$s1}) {
1661 if ( $items{$s1} ) {
1662 my $link = "$htmlroot/$items{$s1}#item_" . htmlify(0,$s2) ;
1663 # Here, we take advantage of the knowledge that $htmlfileurl ne ''
1664 # implies $htmlroot eq ''.
1666 if ( $htmlfileurl ne '' ) {
1667 $link = "$htmldir$link" ;
1668 $url = relativize_url( $link, $htmlfileurl ) ;
1673 $s1 = "<A HREF=\"$url\">$str</A>" ;
1676 $s1 = "<A HREF=\"#item_" . htmlify(0,$s2) . "\">$str</A>" ;
1678 $s1 =~ s,(perl\w+/(\S+)\.html)#item_\2\b,$1,;
1679 confess "s1 has space: $s1" if $s1 =~ /HREF="[^"]*\s[^"]*"/;
1681 $s1 = "<CODE>$str</CODE>";
1682 # warn "$0: $podfile: cannot resolve C<$str> in paragraph $paragraph\n" if $verbose
1690 # process_E - process the E<> pod directive which seems to escape a character.
1696 s,([^/].*),\&$1\;,g;
1703 # process_Z - process the Z<> pod directive which really just amounts to
1704 # ignoring it. this allows someone to start a paragraph with an =
1709 # there is no equivalent in HTML for this so just ignore it.
1715 # process_S - process the S<> pod directive which means to convert all
1716 # spaces in the string to non-breaking spaces (in HTML-eze).
1721 # convert all spaces in the text to non-breaking spaces in HTML.
1722 $str =~ s/ / /g;
1727 # process_X - this is supposed to make an index entry. we'll just
1736 # Adapted from Nick Ing-Simmons' PodToHtml package.
1738 my $source_file = shift ;
1739 my $destination_file = shift;
1741 my $source = URI::file->new_abs($source_file);
1742 my $uo = URI::file->new($destination_file,$source)->abs;
1743 return $uo->rel->as_string;
1748 # finish_list - finish off any pending HTML lists. this should be called
1749 # after the entire pod file has been read and converted.
1752 while ($listlevel > 0) {
1753 print HTML "</DL>\n";
1759 # htmlify - converts a pod section specification to a suitable section
1760 # specification for HTML. if first arg is 1, only takes 1st word.
1763 my($compact, $heading) = @_;
1766 $heading =~ /^(\w+)/;
1770 # $heading = lc($heading);
1771 $heading =~ s/[^\w\s]/_/g;
1772 $heading =~ s/(\s+)/ /g;
1773 $heading =~ s/^\s*(.*?)\s*$/$1/s;
1774 $heading =~ s/ /_/g;
1775 $heading =~ s/\A(.{32}).*\Z/$1/s;
1776 $heading =~ s/\s+\Z//;
1777 $heading =~ s/_{2,}/_/g;