]> sipb.mit.edu Git - ikiwiki.git/blobdiff - IkiWiki/Render.pm
Encode is used in nearly all code paths so stop loading it on demand
[ikiwiki.git] / IkiWiki / Render.pm
index f9730193b42c26ef3cacfd1eac51cddc73d794be..960d1da1e7fbd7c4c64fea6cb24082d1b41521e5 100644 (file)
@@ -5,58 +5,29 @@ package IkiWiki;
 use warnings;
 use strict;
 use File::Spec;
+use IkiWiki;
+use Encode;
 
-sub linkify ($$) { #{{{
-       my $content=shift;
+sub linkify ($$$) { #{{{
+       my $lpage=shift;
        my $page=shift;
+       my $content=shift;
 
        $content =~ s{(\\?)$config{wiki_link_regexp}}{
-               $2 ? ( $1 ? "[[$2|$3]]" : htmllink($page, titlepage($3), 0, 0, pagetitle($2)))
-                  : ( $1 ? "[[$3]]" :    htmllink($page, titlepage($3)))
+               $2 ? ( $1 ? "[[$2|$3]]" : htmllink($lpage, $page, titlepage($3), 0, 0, pagetitle($2)))
+                  : ( $1 ? "[[$3]]" :    htmllink($lpage, $page, titlepage($3)))
        }eg;
        
        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;
        
        if (! $INC{"/usr/bin/markdown"}) {
+               # Note: a proper perl module is available in Debian
+               # for markdown, but not upstream yet.
                no warnings 'once';
                $blosxom::version="is a proper perl module too much to ask?";
                use warnings 'all';
@@ -64,14 +35,21 @@ sub htmlize ($$) { #{{{
        }
        
        if ($type eq '.mdwn') {
+               # Workaround for perl bug (#376329)
+               $content=Encode::encode_utf8($content);
+               $content=Encode::encode_utf8($content);
                $content=Markdown::Markdown($content);
+               $content=Encode::decode_utf8($content);
+               $content=Encode::decode_utf8($content);
        }
        else {
                error("htmlization of $type not supported");
        }
 
-       if ($config{sanitize}) {
-               $content=scrubber()->scrub($content);
+       if (exists $hooks{sanitize}) {
+               foreach my $id (keys %{$hooks{sanitize}}) {
+                       $content=$hooks{sanitize}{$id}{call}->($content);
+               }
        }
        
        return $content;
@@ -122,9 +100,10 @@ sub parentlinks ($) { #{{{
        return @ret;
 } #}}}
 
-sub preprocess ($$) { #{{{
+sub preprocess ($$;$) { #{{{
        my $page=shift;
        my $content=shift;
+       my $onlystrip=shift || 0; # strip directives without processing
 
        my $handle=sub {
                my $escape=shift;
@@ -133,15 +112,20 @@ sub preprocess ($$) { #{{{
                if (length $escape) {
                        return "[[$command $params]]";
                }
-               elsif (exists $plugins{preprocess}{$command}) {
-                       my %params;
-                       while ($params =~ /(\w+)=\"([^"]+)"(\s+|$)/g) {
-                               $params{$1}=$2;
+               elsif ($onlystrip) {
+                       return "";
+               }
+               elsif (exists $hooks{preprocess}{$command}) {
+                       # Note: preserve order of params, some plugins may
+                       # consider it significant.
+                       my @params;
+                       while ($params =~ /(\w+)=\"?([^"]+)"?(\s+|$)/g) {
+                               push @params, $1, $2;
                        }
-                       return $plugins{preprocess}{$command}->(page => $page, %params);
+                       return $hooks{preprocess}{$command}{call}->(@params, page => $page);
                }
                else {
-                       return "[[bad directive $command]]";
+                       return "[[$command not processed]]";
                }
        };
        
@@ -157,35 +141,63 @@ sub add_depends ($$) { #{{{
                $depends{$page}=$globlist;
        }
        else {
-               $depends{$page}.=" ".$globlist;
+               $depends{$page}=globlist_merge($depends{$page}, $globlist);
        }
 } # }}}
 
+sub globlist_merge ($$) { #{{{
+       my $a=shift;
+       my $b=shift;
+
+       my $ret="";
+       # Only add negated globs if they are not matched by the other globlist.
+       foreach my $i ((map { [ $a, $_ ] } split(" ", $b)), 
+                      (map { [ $b, $_ ] } split(" ", $a))) {
+               if ($i->[1]=~/^!(.*)/) {
+                       if (! globlist_match($1, $i->[0])) {
+                               $ret.=" ".$i->[1];
+                       }
+               }
+               else {
+                       $ret.=" ".$i->[1];
+               }
+       }
+       
+       return $ret;
+} #}}}
+
 sub genpage ($$$) { #{{{
-       my $content=shift;
        my $page=shift;
+       my $content=shift;
        my $mtime=shift;
 
        my $title=pagetitle(basename($page));
        
-       my $template=HTML::Template->new(blind_cache => 1,
-               filename => "$config{templatedir}/page.tmpl");
-       
+       my $template=template("page.tmpl", blind_cache => 1);
+       my $actions=0;
+
        if (length $config{cgiurl}) {
                $template->param(editurl => cgiurl(do => "edit", page => $page));
                $template->param(prefsurl => cgiurl(do => "prefs"));
                if ($config{rcs}) {
                        $template->param(recentchangesurl => cgiurl(do => "recentchanges"));
                }
+               $actions++;
        }
 
        if (length $config{historyurl}) {
                my $u=$config{historyurl};
                $u=~s/\[\[file\]\]/$pagesources{$page}/g;
                $template->param(historyurl => $u);
+               $actions++;
+       }
+       if ($config{discussion}) {
+               $template->param(discussionlink => htmllink($page, $page, "Discussion", 1, 1));
+               $actions++;
        }
-       if ($config{hyperestraier}) {
-               $template->param(hyperestraierurl => cgiurl());
+
+       if ($actions) {
+               $template->param(have_actions => 1);
        }
 
        $template->param(
@@ -194,10 +206,15 @@ sub genpage ($$$) { #{{{
                parentlinks => [parentlinks($page)],
                content => $content,
                backlinks => [backlinks($page)],
-               discussionlink => htmllink($page, "Discussion", 1, 1),
-               mtime => scalar(gmtime($mtime)),
+               mtime => displaytime($mtime),
                styleurl => styleurl($page),
        );
+
+       if (exists $hooks{pagetemplate}) {
+               foreach my $id (keys %{$hooks{pagetemplate}}) {
+                       $hooks{pagetemplate}{$id}{call}->($page, $template);
+               }
+       }
        
        return $template->output;
 } #}}}
@@ -216,6 +233,16 @@ sub check_overwrite ($$) { #{{{
        }
 } #}}}
 
+sub displaytime ($) { #{{{
+       my $time=shift;
+
+       eval q{use POSIX};
+       # strftime doesn't know about encodings, so make sure
+       # its output is properly treated as utf8
+       return Encode::decode_utf8(POSIX::strftime(
+                       $config{timeformat}, localtime($time)));
+} #}}}
+
 sub mtime ($) { #{{{
        my $file=shift;
        
@@ -223,16 +250,21 @@ sub mtime ($) { #{{{
 } #}}}
 
 sub findlinks ($$) { #{{{
-       my $content=shift;
        my $page=shift;
+       my $content=shift;
 
        my @links;
        while ($content =~ /(?<!\\)$config{wiki_link_regexp}/g) {
                push @links, titlepage($2);
        }
-       # Discussion links are a special case since they're not in the text
-       # of the page, but on its template.
-       return @links, "$page/discussion";
+       if ($config{discussion}) {
+               # Discussion links are a special case since they're not in the
+               # text of the page, but on its template.
+               return @links, "$page/discussion";
+       }
+       else {
+               return @links;
+       }
 } #}}}
 
 sub render ($) { #{{{
@@ -243,17 +275,26 @@ sub render ($) { #{{{
        if ($type ne 'unknown') {
                my $content=readfile($srcfile);
                my $page=pagename($file);
-               
-               $links{$page}=[findlinks($content, $page)];
                delete $depends{$page};
                
-               $content=linkify($content, $page);
+               if (exists $hooks{filter}) {
+                       foreach my $id (keys %{$hooks{filter}}) {
+                               $content=$hooks{filter}{$id}{call}->(
+                                       page => $page,
+                                       content => $content
+                               );
+                       }
+               }
+               
+               $links{$page}=[findlinks($page, $content)];
+               
+               $content=linkify($page, $page, $content);
                $content=preprocess($page, $content);
                $content=htmlize($type, $content);
                
                check_overwrite("$config{destdir}/".htmlpage($page), $page);
                writefile(htmlpage($page), $config{destdir},
-                       genpage($content, $page, mtime($srcfile)));
+                       genpage($page, $content, mtime($srcfile)));
                $oldpagemtime{$page}=time;
                $renderedfiles{$page}=htmlpage($page);
        }
@@ -278,57 +319,6 @@ sub prune ($) { #{{{
        }
 } #}}}
 
-sub estcfg () { #{{{
-       my $estdir="$config{wikistatedir}/hyperestraier";
-       my $cgi=basename($config{cgiurl});
-       $cgi=~s/\..*$//;
-       open(TEMPLATE, ">$estdir/$cgi.tmpl") ||
-               error("write $estdir/$cgi.tmpl: $!");
-       print TEMPLATE misctemplate("search", 
-               "<!--ESTFORM-->\n\n<!--ESTRESULT-->\n\n<!--ESTINFO-->\n\n");
-       close TEMPLATE;
-       open(TEMPLATE, ">$estdir/$cgi.conf") ||
-               error("write $estdir/$cgi.conf: $!");
-       my $template=HTML::Template->new(
-               filename => "$config{templatedir}/estseek.conf"
-       );
-       eval q{use Cwd 'abs_path'};
-       $template->param(
-               index => $estdir,
-               tmplfile => "$estdir/$cgi.tmpl",
-               destdir => abs_path($config{destdir}),
-               url => $config{url},
-       );
-       print TEMPLATE $template->output;
-       close TEMPLATE;
-       $cgi="$estdir/".basename($config{cgiurl});
-       unlink($cgi);
-       symlink("/usr/lib/estraier/estseek.cgi", $cgi) ||
-               error("symlink $cgi: $!");
-} # }}}
-
-sub estcmd ($;@) { #{{{
-       my @params=split(' ', shift);
-       push @params, "-cl", "$config{wikistatedir}/hyperestraier";
-       if (@_) {
-               push @params, "-";
-       }
-       
-       my $pid=open(CHILD, "|-");
-       if ($pid) {
-               # parent
-               foreach (@_) {
-                       print CHILD "$_\n";
-               }
-               close(CHILD) || error("estcmd @params exited nonzero: $?");
-       }
-       else {
-               # child
-               open(STDOUT, "/dev/null"); # shut it up (closing won't work)
-               exec("estcmd", @params) || error("can't run estcmd");
-       }
-} #}}}
-
 sub refresh () { #{{{
        # find existing pages
        my %exists;
@@ -337,6 +327,7 @@ sub refresh () { #{{{
        find({
                no_chdir => 1,
                wanted => sub {
+                       $_ = Encode::decode_utf8($_);
                        if (/$config{wiki_file_prune_regexp}/) {
                                $File::Find::prune=1;
                        }
@@ -356,6 +347,7 @@ sub refresh () { #{{{
        find({
                no_chdir => 1,
                wanted => sub {
+                       $_ = Encode::decode_utf8($_);
                        if (/$config{wiki_file_prune_regexp}/) {
                                $File::Find::prune=1;
                        }
@@ -389,8 +381,12 @@ sub refresh () { #{{{
                        push @add, $file;
                        $links{$page}=[];
                        $pagesources{$page}=$file;
-                       $pagectime{$page}=mtime(srcfile($file))
-                               unless exists $pagectime{$page};
+                       if ($config{getctime} && -e "$config{srcdir}/$file") {
+                               $pagectime{$page}=rcs_getctime("$config{srcdir}/$file");
+                       }
+                       elsif (! exists $pagectime{$page}) {
+                               $pagectime{$page}=mtime(srcfile($file));
+                       }
                }
        }
        my @del;
@@ -411,7 +407,7 @@ sub refresh () { #{{{
                
                if (! exists $oldpagemtime{$page} ||
                    mtime(srcfile($file)) > $oldpagemtime{$page}) {
-                       debug("rendering changed file $file");
+                       debug("rendering $file");
                        render($file);
                        $rendered{$file}=1;
                }
@@ -440,7 +436,7 @@ FILE:               foreach my $file (@files) {
        }
 
        # Handle backlinks; if a page has added/removed links, update the
-       # pages it links to. Also handles rebuilding dependat pages.
+       # pages it links to. Also handles rebuilding dependant pages.
        # TODO: inefficient; pages may get rendered above and again here;
        # problem is the backlinks could be wrong in the first pass render
        # above
@@ -494,19 +490,15 @@ FILE:             foreach my $file (@files) {
                }
        }
 
-       if ($config{hyperestraier} && (%rendered || @del)) {
-               debug("updating hyperestraier search index");
-               if (%rendered) {
-                       estcmd("gather -cm -bc -cl -sd", 
-                               map { $config{destdir}."/".$renderedfiles{pagename($_)} }
-                               keys %rendered);
+       if (@del && exists $hooks{delete}) {
+               foreach my $id (keys %{$hooks{delete}}) {
+                       $hooks{delete}{$id}{call}->(@del);
                }
-               if (@del) {
-                       estcmd("purge -cl");
+       }
+       if (%rendered && exists $hooks{change}) {
+               foreach my $id (keys %{$hooks{change}}) {
+                       $hooks{change}{$id}{call}->(keys %rendered);
                }
-               
-               debug("generating hyperestraier cgi config");
-               estcfg();
        }
 } #}}}