fixed most of the xapian todos
authorJoey Hess <joey@kodama.kitenet.net>
Wed, 4 Jun 2008 03:52:56 +0000 (23:52 -0400)
committerJoey Hess <joey@kodama.kitenet.net>
Wed, 4 Jun 2008 03:52:56 +0000 (23:52 -0400)
IkiWiki/Plugin/search.pm

index fc77a7d2fad4cb3e56dbbdd589d3fc3c722a4317..0d4a8f552abb9bf7f38d7e1112e1278308df1353 100644 (file)
@@ -61,6 +61,7 @@ sub needsbuild ($) { #{{{
        %toindex = map { pagename($_) => 1 } @{shift()};
 } #}}}
 
+my $scrubber;
 sub filter (@) { #{{{
        my %params=@_;
        
@@ -77,10 +78,31 @@ sub filter (@) { #{{{
                        $title=IkiWiki::pagetitle($params{page});
                }
 
+               # Remove any html from text to be indexed.
+               # TODO: This removes html that is in eg, a markdown pre,
+               # which should not be removed.
+               if (! defined $scrubber) {
+                       eval q{use HTML::Scrubber};
+                       error($@) if $@;
+                       $scrubber=HTML::Scrubber->new(allow => []);
+               }
+               my $toindex=$scrubber->scrub($params{content});
+
+               # Take 512 characters for a sample, then extend it out
+               # if it stopped in the middle of a word.
+               my $size=512;
+               my ($sample)=substr($toindex, 0, $size);
+               my $next=substr($toindex, $size++, 1);
+               while ($next !~ /\s/) {
+                       $sample.=$next;
+                       $next=substr($toindex, $size++, 1);
+               }
+               $sample=~s/\n/ /g;
+
                # data used by omega
                $doc->set_data(
                        "url=".urlto($params{page}, "")."\n".
-                       "sample=\n". # TODO
+                       "sample=$sample\n".
                        "caption=$title\n".
                        "modtime=$IkiWiki::pagemtime{$params{page}}\n".
                        "size=".length($params{content})."\n"
@@ -91,7 +113,7 @@ sub filter (@) { #{{{
                $tg->set_document($doc);
                $tg->index_text($params{page}, 2);
                $tg->index_text($title, 2);
-               $tg->index_text($params{content}); # TODO html strip; preprocessor too
+               $tg->index_text($toindex);
 
                my $pageterm=pageterm($params{page});
                $doc->add_term($pageterm);