Perlbal::Fields provide a simple way to override field::new that uses…

… Hash::Util::lock_ref_keys

set env PERLBAL_REMOVE_FIELDS=1 to unactivate fields stuff
fields are fast for reading access, but slow when need to create many objects

lib/Perlbal.pm

@@ -126,6 +126,13 @@ if ($ENV{PERLBAL_XS_HEADERS} && $XSModules{headers}) {
     Perlbal::XS::HTTPHeaders::enable();
 }
 
+# unactivate field::new 
+if ($ENV{PERLBAL_REMOVE_FIELDS}) {
+	use Perlbal::Fields;
+	Perlbal::Fields->remove();
+}
+
+
 # setup a USR1 signal handler that tells us to dump some basic statistics
 # of how we're doing to the syslog
 $SIG{'USR1'} = sub {

lib/Perlbal/Fields.pm

@@ -0,0 +1,47 @@
+package Perlbal::Fields;
+use strict;
+use warnings;
+use fields;
+
+# allow package to be called in command line
+__PACKAGE__->run(@ARGV) unless caller();
+
+# should be the main method called, extra sub could be triggered from this point
+sub run {
+    my ( $package, @options ) = @_;
+
+    # unactivate fields if launch in command line
+    $package->remove();
+
+    1;
+}
+
+# hash with keys and undef val for each class
+my $cache_for_class = {};
+
+# replace fields::new method which uses Hash::Util::lock_ref_keys
+# - it's a good idea to keep using the original fields::new during development stage
+# - but during production we can avoid locking hash and wasting time doing this ( ~ 30 % )
+sub remove {
+    my ($class) = @_;
+
+    no warnings "redefine";
+    no strict 'refs';
+    *fields::new = sub {
+        my $class = shift;
+        $class = ref $class if ref $class;
+
+        if ( !defined( $cache_for_class->{$class} ) ) {
+            my $h    = {};
+            my @keys = keys %{ $class . "::FIELDS" };
+            map { $h->{$_} = undef; } @keys;
+            $cache_for_class->{$class} = $h;
+        }
+        my %h = %{ $cache_for_class->{$class} };
+
+        return bless \%h, $class;
+    };
+}
+
+1;
+