Mostly, this is a refactoring commit. There is one significant new feature,

though: overrides now only work within a class heirarchy, which is to say that
a subclass can override an element in a base class, but a child scope cannot
otherwise override an element in a base scope.

I've also done a good bit of refactoring, though; notably, AST#evaluate now
takes named arguments, and I changed the 'name' parameter to 'type' in all of
the Component classes (this was all internal, but was confusing as it was).

I also removed the need for the autonaming stuff -- it's now acceptable for
components not to have names, and everything behaves correctly.  I haven't yet
removed the autoname code, though; I'll do that on the next commit.


git-svn-id: https://reductivelabs.com/svn/puppet/trunk@952 980ebf18-57e1-0310-9a29-db15c13687c0

1 files changed, 4 insertions, 3 deletions

diff --git a/lib/puppet/parser/ast/objectref.rb b/lib/puppet/parser/ast/objectref.rb
index 6418c14fb..12e36360f 100644
--- a/lib/puppet/parser/ast/objectref.rb
+++ b/lib/puppet/parser/ast/objectref.rb
@@ -12,9 +12,10 @@ class Puppet::Parser::AST
 
         # Evaluate our object, but just return a simple array of the type
         # and name.
-        def evaluate(scope)
-            objtype = @type.safeevaluate(scope)
-            objnames = @name.safeevaluate(scope)
+        def evaluate(hash)
+            scope = hash[:scope]
+            objtype = @type.safeevaluate(:scope => scope)
+            objnames = @name.safeevaluate(:scope => scope)
 
             # it's easier to always use an array, even for only one name
             unless objnames.is_a?(Array)