Configurable prelude

Add an attribute 'prelude' which defaults to 'use strictures 1;' so
use of strictures is configurable.

diff --git a/Changes b/Changes
index 27cefdf..b177418 100644 (file)
--- a/Changes
+++ b/Changes
@@ -1,3 +1,5 @@
+  - Make prelude configurable, so strictures can be optional
+
 1.1.0 2011-01-11 21:51:00
   - Add a #line directive so it's clearer where errors occurred (DGL)
   - Filter out lexicals called '&' since these are anon subs (MSTROUT)

diff --git a/lib/Eval/WithLexicals.pm b/lib/Eval/WithLexicals.pm
index 31703e6..fab265c 100644 (file)
--- a/lib/Eval/WithLexicals.pm
+++ b/lib/Eval/WithLexicals.pm
@@ -25,6 +25,10 @@ has in_package => (
   is => 'rw', default => quote_sub q{ 'Eval::WithLexicals::Scratchpad' }
 );
 
+has prelude => (
+  is => 'rw', default => quote_sub q{ 'use strictures 1;' }
+);
+
 sub eval {
   my ($self, $to_eval) = @_;
   local *Eval::WithLexicals::Cage::current_line;
@@ -32,7 +36,8 @@ sub eval {
   local *Eval::WithLexicals::Cage::grab_captures;
   my $setup = Sub::Quote::capture_unroll('$_[2]', $self->lexicals, 2);
   my $package = $self->in_package;
-  local our $current_code = qq!use strictures 1;
+  my $prelude = $self->prelude;
+  local our $current_code = qq!${prelude}
 ${setup}
 sub Eval::WithLexicals::Cage::current_line {
 package ${package};
@@ -155,6 +160,7 @@ Eval::WithLexicals - pure perl eval with persistent lexical variables
     lexicals => { '$x' => \1 },      # default {}
     in_package => 'PackageToEvalIn', # default Eval::WithLexicals::Scratchpad
     context => 'scalar',             # default 'list'
+    prelude => 'use warnings',       # default 'use strictures 1'
   );
 
 =head2 eval
@@ -179,6 +185,14 @@ Eval::WithLexicals - pure perl eval with persistent lexical variables
 
   $eval->context($new_context); # 'list', 'scalar' or 'void'
 
+=head2 prelude
+
+Code to run before evaling code. Loads L<strictures> by default.
+
+  my $current_prelude = $eval->prelude;
+
+  $eval->prelude(q{use warnings}); # only warnings, not strict.
+
 =head1 AUTHOR
 
 Matt S. Trout <mst@shadowcat.co.uk>