web commit by joey

[git.ikiwiki.info.git] / IkiWiki / Render.pm

diff --git a/IkiWiki/Render.pm b/IkiWiki/Render.pm
index 7d1e8ee53fb98884b7a2e8a45157f9b0b83bb0af..a4efa0b74511a1b88faefd4e0519242e04736fe4 100644 (file)
--- a/IkiWiki/Render.pm
+++ b/IkiWiki/Render.pm
@@ -1,8 +1,11 @@
+#!/usr/bin/perl
+

 package IkiWiki;
 
 use warnings;
 use strict;
 use File::Spec;
 package IkiWiki;
 
 use warnings;
 use strict;
 use File::Spec;

+use IkiWiki;

 
 sub linkify ($$) { #{{{
        my $content=shift;
 
 sub linkify ($$) { #{{{
        my $content=shift;


@@ -16,6 +19,40 @@ sub linkify ($$) { #{{{
        return $content;
 } #}}}
 
        return $content;
 } #}}}
 

+my $_scrubber;
+sub scrubber { #{{{
+       return $_scrubber if defined $_scrubber;
+       
+       eval q{use HTML::Scrubber};
+       # Lists based on http://feedparser.org/docs/html-sanitization.html
+       $_scrubber = HTML::Scrubber->new(
+               allow => [qw{
+                       a abbr acronym address area b big blockquote br
+                       button caption center cite code col colgroup dd del
+                       dfn dir div dl dt em fieldset font form h1 h2 h3 h4
+                       h5 h6 hr i img input ins kbd label legend li map
+                       menu ol optgroup option p pre q s samp select small
+                       span strike strong sub sup table tbody td textarea
+                       tfoot th thead tr tt u ul var
+               }],
+               default => [undef, { map { $_ => 1 } qw{
+                       abbr accept accept-charset accesskey action
+                       align alt axis border cellpadding cellspacing
+                       char charoff charset checked cite class
+                       clear cols colspan color compact coords
+                       datetime dir disabled enctype for frame
+                       headers height href hreflang hspace id ismap
+                       label lang longdesc maxlength media method
+                       multiple name nohref noshade nowrap prompt
+                       readonly rel rev rows rowspan rules scope
+                       selected shape size span src start summary
+                       tabindex target title type usemap valign
+                       value vspace width
+               }}],
+       );
+       return $_scrubber;
+} # }}}
+

 sub htmlize ($$) { #{{{
        my $type=shift;
        my $content=shift;
 sub htmlize ($$) { #{{{
        my $type=shift;
        my $content=shift;


@@ -28,11 +65,17 @@ sub htmlize ($$) { #{{{
        }
        
        if ($type eq '.mdwn') {
        }
        
        if ($type eq '.mdwn') {

-               return Markdown::Markdown($content);
+               $content=Markdown::Markdown($content);

        }
        else {
                error("htmlization of $type not supported");
        }
        }
        else {
                error("htmlization of $type not supported");
        }

+
+       if ($config{sanitize}) {
+               $content=scrubber()->scrub($content);
+       }
+       
+       return $content;

 } #}}}
 
 sub backlinks ($) { #{{{
 } #}}}
 
 sub backlinks ($) { #{{{


@@ -80,35 +123,26 @@ sub parentlinks ($) { #{{{
        return @ret;
 } #}}}
 
        return @ret;
 } #}}}
 

-sub rsspage ($) { #{{{
-       my $page=shift;
-
-       return $page.".rss";
-} #}}}
-
-sub postprocess { #{{{
-       # Takes content to postprocess followed by a list of postprocessor
-       # commands and subroutine references to run for the commands.
+sub preprocess ($$) { #{{{

        my $page=shift;
        my $content=shift;
        my $page=shift;
        my $content=shift;

-       my %commands=@_;
-       
+

        my $handle=sub {
                my $escape=shift;
                my $command=shift;
                my $params=shift;
                if (length $escape) {
        my $handle=sub {
                my $escape=shift;
                my $command=shift;
                my $params=shift;
                if (length $escape) {

-                       "[[$command $params]]";
+                       return "[[$command $params]]";

                }
                }

-               elsif (exists $commands{$command}) {
+               elsif (exists $hooks{preprocess}{$command}) {

                        my %params;
                        while ($params =~ /(\w+)=\"([^"]+)"(\s+|$)/g) {
                                $params{$1}=$2;
                        }
                        my %params;
                        while ($params =~ /(\w+)=\"([^"]+)"(\s+|$)/g) {
                                $params{$1}=$2;
                        }

-                       $commands{$command}->($page, %params);
+                       return $hooks{preprocess}{$command}{call}->(page => $page, %params);

                }
                else {
                }
                else {

-                       "[[bad directive $command]]";
+                       return "[[$command not processed]]";

                }
        };
        
                }
        };
        


@@ -116,77 +150,37 @@ sub postprocess { #{{{
        return $content;
 } #}}}
 
        return $content;
 } #}}}
 

-sub blog_list ($$) { #{{{
-       my $globlist=shift;
-       my $maxitems=shift;
-       
-       my @list;
-       foreach my $page (keys %pagesources) {
-               if (globlist_match($page, $globlist)) {
-                       push @list, $page;
-               }
-       }
-
-       @list=sort { $pagectime{$b} <=> $pagectime{$a} } @list;
-       return @list if ! $maxitems || @list <= $maxitems;
-       return @list[0..$maxitems - 1];
-} #}}}
-
-sub get_inline_content ($$) { #{{{
-       my $parentpage=shift;
+sub add_depends ($$) { #{{{

        my $page=shift;
        my $page=shift;

+       my $globlist=shift;

        
        

-       my $file=$pagesources{$page};
-       my $type=pagetype($file);
-       if ($type ne 'unknown') {
-               return htmlize($type, linkify(readfile("$config{srcdir}/$file"), $parentpage));
+       if (! exists $depends{$page}) {
+               $depends{$page}=$globlist;

        }
        else {
        }
        else {

-               return "";
+               $depends{$page}=globlist_merge($depends{$page}, $globlist);

        }
        }

-} #}}}
+} # }}}
+
+sub globlist_merge ($$) { #{{{
+       my $a=shift;
+       my $b=shift;

 
 

-sub postprocess_html_inline { #{{{
-       my $parentpage=shift;
-       my %params=@_;
-       
-       if (! exists $params{pages}) {
-               return "";
-       }
-       if (! exists $params{archive}) {
-               $params{archive}="no";
-       }
-       if (! exists $params{show} && $params{archive} eq "no") {
-               $params{show}=10;
-       }
-       $inlinepages{$parentpage}=$params{pages};
-       

        my $ret="";
        my $ret="";

-       
-       if (exists $params{rootpage}) {
-               my $formtemplate=HTML::Template->new(blind_cache => 1,
-                       filename => "$config{templatedir}/blogpost.tmpl");
-               $formtemplate->param(cgiurl => $config{cgiurl});
-               $formtemplate->param(rootpage => $params{rootpage});
-               my $form=$formtemplate->output;
-               $ret.=$form;
-       }
-       
-       my $template=HTML::Template->new(blind_cache => 1,
-               filename => (($params{archive} eq "no") 
-                               ? "$config{templatedir}/inlinepage.tmpl"
-                               : "$config{templatedir}/inlinepagetitle.tmpl"));
-       
-       foreach my $page (blog_list($params{pages}, $params{show})) {
-               next if $page eq $parentpage;
-               $template->param(pagelink => htmllink($parentpage, $page));
-               $template->param(content => get_inline_content($parentpage, $page))
-                       if $params{archive} eq "no";
-               $template->param(ctime => scalar(gmtime($pagectime{$page})));
-               $ret.=$template->output;
+       # Only add negated globs if they are not matched by the other globlist.
+       foreach my $i ((map { [ $a, $_ ] } split(" ", $b)), 
+                      (map { [ $b, $_ ] } split(" ", $a))) {
+               if ($i->[1]=~/^!(.*)/) {
+                       if (! globlist_match($1, $i->[0])) {
+                               $ret.=" ".$i->[1];
+                       }
+               }
+               else {
+                       $ret.=" ".$i->[1];
+               }

        }
        
        }
        

-       return "</p>$ret<p>";
+       return $ret;

 } #}}}
 
 sub genpage ($$$) { #{{{
 } #}}}
 
 sub genpage ($$$) { #{{{


@@ -194,8 +188,6 @@ sub genpage ($$$) { #{{{
        my $page=shift;
        my $mtime=shift;
 
        my $page=shift;
        my $mtime=shift;
 

-       $content = postprocess($page, $content, inline => \&postprocess_html_inline);
-       

        my $title=pagetitle(basename($page));
        
        my $template=HTML::Template->new(blind_cache => 1,
        my $title=pagetitle(basename($page));
        
        my $template=HTML::Template->new(blind_cache => 1,


@@ -214,11 +206,8 @@ sub genpage ($$$) { #{{{
                $u=~s/\[\[file\]\]/$pagesources{$page}/g;
                $template->param(historyurl => $u);
        }
                $u=~s/\[\[file\]\]/$pagesources{$page}/g;
                $template->param(historyurl => $u);
        }

+       $template->param(headercontent => $config{headercontent});

 
 

-       if ($config{rss} && $inlinepages{$page}) {
-               $template->param(rssurl => rsspage(basename($page)));
-       }
-       

        $template->param(
                title => $title,
                wikiname => $config{wikiname},
        $template->param(
                title => $title,
                wikiname => $config{wikiname},


@@ -233,73 +222,6 @@ sub genpage ($$$) { #{{{
        return $template->output;
 } #}}}
 
        return $template->output;
 } #}}}
 

-sub date_822 ($) { #{{{
-       my $time=shift;
-
-       eval q{use POSIX};
-       return POSIX::strftime("%a, %d %b %Y %H:%M:%S %z", localtime($time));
-} #}}}
-
-sub absolute_urls ($$) { #{{{
-       # sucky sub because rss sucks
-       my $content=shift;
-       my $url=shift;
-
-       $url=~s/[^\/]+$//;
-       
-       $content=~s/<a\s+href="(?!http:\/\/)([^"]+)"/<a href="$url$1"/ig;
-       $content=~s/<img\s+src="(?!http:\/\/)([^"]+)"/<img src="$url$1"/ig;
-       return $content;
-} #}}}
-
-sub genrss ($$$) { #{{{
-       my $content=shift;
-       my $page=shift;
-       my $mtime=shift;
-       
-       my $url="$config{url}/".htmlpage($page);
-       
-       my $template=HTML::Template->new(blind_cache => 1,
-               filename => "$config{templatedir}/rsspage.tmpl");
-       
-       my @items;
-       my $isblog=0;
-       my $gen_blog=sub {
-               my $parentpage=shift;
-               my %params=@_;
-               
-               if (! exists $params{show}) {
-                       $params{show}=10;
-               }
-               if (! exists $params{pages}) {
-                       return "";
-               }
-               
-               $isblog=1;
-               foreach my $page (blog_list($params{pages}, $params{show})) {
-                       next if $page eq $parentpage;
-                       push @items, {
-                               itemtitle => pagetitle(basename($page)),
-                               itemurl => "$config{url}/$renderedfiles{$page}",
-                               itempubdate => date_822($pagectime{$page}),
-                               itemcontent => absolute_urls(get_inline_content($parentpage, $page), $url),
-                       } if exists $renderedfiles{$page};
-               }
-               
-               return "";
-       };
-       
-       $content = postprocess($page, $content, inline => $gen_blog);
-
-       $template->param(
-               title => $config{wikiname},
-               pageurl => $url,
-               items => \@items,
-       );
-       
-       return $template->output;
-} #}}}
-

 sub check_overwrite ($$) { #{{{
        # Important security check. Make sure to call this before saving
        # any files to the source directory.
 sub check_overwrite ($$) { #{{{
        # Important security check. Make sure to call this before saving
        # any files to the source directory.


@@ -337,34 +259,30 @@ sub render ($) { #{{{
        my $file=shift;
        
        my $type=pagetype($file);
        my $file=shift;
        
        my $type=pagetype($file);

-       my $content=readfile("$config{srcdir}/$file");
+       my $srcfile=srcfile($file);

        if ($type ne 'unknown') {
        if ($type ne 'unknown') {

+               my $content=readfile($srcfile);

                my $page=pagename($file);
                
                $links{$page}=[findlinks($content, $page)];
                my $page=pagename($file);
                
                $links{$page}=[findlinks($content, $page)];

-               delete $inlinepages{$page};
+               delete $depends{$page};

                
                $content=linkify($content, $page);
                
                $content=linkify($content, $page);

+               $content=preprocess($page, $content);

                $content=htmlize($type, $content);
                
                check_overwrite("$config{destdir}/".htmlpage($page), $page);
                $content=htmlize($type, $content);
                
                check_overwrite("$config{destdir}/".htmlpage($page), $page);

-               writefile("$config{destdir}/".htmlpage($page),
-                       genpage($content, $page, mtime("$config{srcdir}/$file")));
+               writefile(htmlpage($page), $config{destdir},
+                       genpage($content, $page, mtime($srcfile)));

                $oldpagemtime{$page}=time;
                $renderedfiles{$page}=htmlpage($page);
                $oldpagemtime{$page}=time;
                $renderedfiles{$page}=htmlpage($page);

-
-               # TODO: should really add this to renderedfiles and call
-               # check_overwrite, as above, but currently renderedfiles
-               # only supports listing one file per page.
-               if ($config{rss} && exists $inlinepages{$page}) {
-                       writefile("$config{destdir}/".rsspage($page),
-                               genrss($content, $page, mtime("$config{srcdir}/$file")));
-               }

        }
        else {
        }
        else {

+               my $content=readfile($srcfile, 1);

                $links{$file}=[];
                $links{$file}=[];

+               delete $depends{$file};

                check_overwrite("$config{destdir}/$file", $file);
                check_overwrite("$config{destdir}/$file", $file);

-               writefile("$config{destdir}/$file", $content);
+               writefile($file, $config{destdir}, $content, 1);

                $oldpagemtime{$file}=time;
                $renderedfiles{$file}=$file;
        }
                $oldpagemtime{$file}=time;
                $renderedfiles{$file}=$file;
        }


@@ -389,9 +307,7 @@ sub refresh () { #{{{
                no_chdir => 1,
                wanted => sub {
                        if (/$config{wiki_file_prune_regexp}/) {
                no_chdir => 1,
                wanted => sub {
                        if (/$config{wiki_file_prune_regexp}/) {

-                               no warnings 'once';

                                $File::Find::prune=1;
                                $File::Find::prune=1;

-                               use warnings "all";

                        }
                        elsif (! -d $_ && ! -l $_) {
                                my ($f)=/$config{wiki_file_regexp}/; # untaint
                        }
                        elsif (! -d $_ && ! -l $_) {
                                my ($f)=/$config{wiki_file_regexp}/; # untaint


@@ -406,6 +322,30 @@ sub refresh () { #{{{
                        }
                },
        }, $config{srcdir});
                        }
                },
        }, $config{srcdir});

+       find({
+               no_chdir => 1,
+               wanted => sub {
+                       if (/$config{wiki_file_prune_regexp}/) {
+                               $File::Find::prune=1;
+                       }
+                       elsif (! -d $_ && ! -l $_) {
+                               my ($f)=/$config{wiki_file_regexp}/; # untaint
+                               if (! defined $f) {
+                                       warn("skipping bad filename $_\n");
+                               }
+                               else {
+                                       # Don't add files that are in the
+                                       # srcdir.
+                                       $f=~s/^\Q$config{underlaydir}\E\/?//;
+                                       if (! -e "$config{srcdir}/$f" && 
+                                           ! -l "$config{srcdir}/$f") {
+                                               push @files, $f;
+                                               $exists{pagename($f)}=1;
+                                       }
+                               }
+                       }
+               },
+       }, $config{underlaydir});

 
        my %rendered;
 
 
        my %rendered;
 


@@ -418,7 +358,7 @@ sub refresh () { #{{{
                        push @add, $file;
                        $links{$page}=[];
                        $pagesources{$page}=$file;
                        push @add, $file;
                        $links{$page}=[];
                        $pagesources{$page}=$file;

-                       $pagectime{$page}=mtime("$config{srcdir}/$file") 
+                       $pagectime{$page}=mtime(srcfile($file))

                                unless exists $pagectime{$page};
                }
        }
                                unless exists $pagectime{$page};
                }
        }


@@ -439,7 +379,7 @@ sub refresh () { #{{{
                my $page=pagename($file);
                
                if (! exists $oldpagemtime{$page} ||
                my $page=pagename($file);
                
                if (! exists $oldpagemtime{$page} ||

-                   mtime("$config{srcdir}/$file") > $oldpagemtime{$page}) {
+                   mtime(srcfile($file)) > $oldpagemtime{$page}) {

                        debug("rendering changed file $file");
                        render($file);
                        $rendered{$file}=1;
                        debug("rendering changed file $file");
                        render($file);
                        $rendered{$file}=1;


@@ -469,19 +409,21 @@ FILE:             foreach my $file (@files) {
        }
 
        # Handle backlinks; if a page has added/removed links, update the
        }
 
        # Handle backlinks; if a page has added/removed links, update the

-       # pages it links to. Also handle inlining here.
+       # pages it links to. Also handles rebuilding dependat pages.

        # TODO: inefficient; pages may get rendered above and again here;
        # problem is the backlinks could be wrong in the first pass render
        # above
        if (%rendered || @del) {
                foreach my $f (@files) {
                        my $p=pagename($f);
        # TODO: inefficient; pages may get rendered above and again here;
        # problem is the backlinks could be wrong in the first pass render
        # above
        if (%rendered || @del) {
                foreach my $f (@files) {
                        my $p=pagename($f);

-                       if (exists $inlinepages{$p}) {
+                       if (exists $depends{$p}) {

                                foreach my $file (keys %rendered, @del) {
                                foreach my $file (keys %rendered, @del) {

+                                       next if $f eq $file;

                                        my $page=pagename($file);
                                        my $page=pagename($file);

-                                       if (globlist_match($page, $inlinepages{$p})) {
-                                               debug("rendering $f, which inlines $page");
+                                       if (globlist_match($page, $depends{$p})) {
+                                               debug("rendering $f, which depends on $page");

                                                render($f);
                                                render($f);

+                                               $rendered{$f}=1;

                                                last;
                                        }
                                }
                                                last;
                                        }
                                }


@@ -495,8 +437,8 @@ FILE:               foreach my $file (@files) {
                        if (exists $links{$page}) {
                                foreach my $link (map { bestlink($page, $_) } @{$links{$page}}) {
                                        if (length $link &&
                        if (exists $links{$page}) {
                                foreach my $link (map { bestlink($page, $_) } @{$links{$page}}) {
                                        if (length $link &&

-                                           ! exists $oldlinks{$page} ||
-                                           ! grep { $_ eq $link } @{$oldlinks{$page}}) {
+                                           (! exists $oldlinks{$page} ||
+                                            ! grep { bestlink($page, $_) eq $link } @{$oldlinks{$page}})) {

                                                $linkchanged{$link}=1;
                                        }
                                }
                                                $linkchanged{$link}=1;
                                        }
                                }


@@ -504,8 +446,8 @@ FILE:               foreach my $file (@files) {
                        if (exists $oldlinks{$page}) {
                                foreach my $link (map { bestlink($page, $_) } @{$oldlinks{$page}}) {
                                        if (length $link &&
                        if (exists $oldlinks{$page}) {
                                foreach my $link (map { bestlink($page, $_) } @{$oldlinks{$page}}) {
                                        if (length $link &&

-                                           ! exists $links{$page} ||
-                                           ! grep { $_ eq $link } @{$links{$page}}) {
+                                           (! exists $links{$page} || 
+                                            ! grep { bestlink($page, $_) eq $link } @{$links{$page}})) {

                                                $linkchanged{$link}=1;
                                        }
                                }
                                                $linkchanged{$link}=1;
                                        }
                                }


@@ -516,9 +458,21 @@ FILE:              foreach my $file (@files) {
                        if (defined $linkfile) {
                                debug("rendering $linkfile, to update its backlinks");
                                render($linkfile);
                        if (defined $linkfile) {
                                debug("rendering $linkfile, to update its backlinks");
                                render($linkfile);

+                               $rendered{$linkfile}=1;

                        }
                }
        }
                        }
                }
        }

+
+       if (@del && exists $hooks{delete}) {
+               foreach my $id (keys %{$hooks{delete}}) {
+                       $hooks{delete}{$id}{call}->(@del);
+               }
+       }
+       if (%rendered && exists $hooks{render}) {
+               foreach my $id (keys %{$hooks{render}}) {
+                       $hooks{render}{$id}{call}->(keys %rendered);
+               }
+       }

 } #}}}
 
 1
 } #}}}
 
 1