--- trunk/spider/swishspider 2003/03/24 09:04:57 29 +++ trunk/spider/swishspider 2003/03/24 09:57:44 30 @@ -59,10 +59,20 @@ my $contents = $response->content(); open( CONTENTS, ">$localpath.contents" ) || die( "Could not open contents file $localpath.contents\n" ); + # if you don't want content to be indexed, include it in + # foobar tags or surround it with comments + # foobar $contents =~ s,.+?,,isg; $contents =~ s,.+?,,isg; + # this will remove all script from indexing content $contents =~ s,,,isg; + # remap Windows charset to ISO-8859-2 $contents =~ tr/šðžèæŠÐŽÈÆ/¹ð¾èæ©Ð®ÈÆ/; # 1250 -> iso8859-2 + # this will fix badly formatted html in form: + # some text + # which will confuse indexer (or libxml2?) + $contents =~ s/[\n\r]^(>)/$1\n/msg; print CONTENTS $contents; close( CONTENTS );