Add a note on the performance of APR::Request

[catagits/Catalyst-Runtime.git] / lib / Catalyst / Engine.pm

diff --git a/lib/Catalyst/Engine.pm b/lib/Catalyst/Engine.pm
index 564cb42..fd22ff0 100644 (file)
--- a/lib/Catalyst/Engine.pm
+++ b/lib/Catalyst/Engine.pm
@@ -20,6 +20,11 @@ use overload '""' => sub { return ref shift }, fallback => 1;
 # Amount of data to read from input on each pass
 our $CHUNKSIZE = 64 * 1024;
 
+# See if we can use libapreq2 for URI unescaping
+use constant HAS_APR => eval {
+    require APR::Request;
+};
+
 =head1 NAME
 
 Catalyst::Engine - The Catalyst Engine
@@ -636,14 +641,19 @@ sub write {
 
 =head2 $self->unescape_uri($uri)
 
-Unescapes a given URI using the most efficient method available.  Engines such
-as Apache may implement this using Apache's C-based modules, for example.
+Unescapes a given URI using the most efficient method available.  Engines
+can subclass to provide faster implementations.
 
 =cut
 
 sub unescape_uri {
     my $self = shift;
     
+    if ( HAS_APR ) {
+        # This function is ~12x faster than URI::Escape
+        return APR::Request::decode(@_);
+    }
+    
     my $e = URI::Escape::uri_unescape(@_);
     $e =~ s/\+/ /g;