diff: bug fix for empty typeclass decls

[David Glen JEFFERY]

Hi,

Could you please review this, Fergus?



Estimated hours taken: 1

Bug fix. When there are no methods for a class, the maybe(list(pred_proc_id))
field of the hlds_instance_defn was left as no, rather than yes([]), so
base_typeclass_info.m bombed out when trying to generate the pointers to the
typeclass methods.

compiler/base_typeclass_info.m:
	Explicitly check for typeclasses with no methods, and set the
	mybe(list(pred_proc_id)) field to yes([]).


Index: compiler//check_typeclass.m
===================================================================
RCS file: /home/staff/zs/imp/mercury/compiler/check_typeclass.m,v
retrieving revision 1.2
diff -u -r1.2 check_typeclass.m
--- check_typeclass.m	1998/01/13 09:57:16	1.2
+++ check_typeclass.m	1998/01/22 04:06:34
@@ -92,8 +92,21 @@
 		Errors0, Errors):-
 		
 		% check conformance of the instance interface
-	list__foldl2(check_instance_pred(Vars, ClassInterface, ModuleInfo), 
-		PredIds, InstanceDefn0, InstanceDefn1, Errors0, Errors1),
+	(
+		PredIds \= []
+	->
+		list__foldl2(check_instance_pred(Vars, ClassInterface,
+			ModuleInfo), PredIds, InstanceDefn0, InstanceDefn1,
+			Errors0, Errors1)
+	;
+		% there are no methods for this class
+		InstanceDefn0 = hlds_instance_defn(A, B, C, D, 
+				_MaybeInstancePredProcs, F, G),
+		InstanceDefn1 = hlds_instance_defn(A, B, C, D, 
+				yes([]), F, G),
+		Errors1 = Errors0
+	),
+
 
 		% check that the superclass constraints are satisfied for the
 		% types in this instance declaration


love and cuddles,
dgj
-- 
David Jeffery (dgj at cs.mu.oz.au) |  Marge: Did you just call everyone "chicken"?
MEngSc student,                 |  Homer: Noooo.  I swear on this Bible!
Department of Computer Science  |  Marge: That's not a Bible; that's a book of
University of Melbourne         |         carpet samples!
Australia                       |  Homer: Ooooh... Fuzzy.