[m-rev.] for review: module qualification of equality preds

[Simon Taylor]

On 19-May-2003, Ralph Becket <rafe at cs.mu.OZ.AU> wrote:
> Simon Taylor, Monday, 19 May 2003:
> > 
> > Estimated hours taken: 4
> > Branches: main, release
> > 
> > The compiler was not module qualifying the `where equality is ...'
> > part of type declarations in interface files. This could cause the
> > compiler to use the wrong predicate in importing modules if the
> > user-defined equality predicate was not exported.
> > 
> > To avoid this problem we now require user-defined equality or
> > comparison predicates for a type to be defined in the same module
> > as the type. This is the only simple way to module qualify the
> > unification predicates in interface files without running
> > typechecking.
> 
> This restriction should be documented for users.


Estimated hours taken: 4
Branches: main, release

The compiler was not module qualifying the `where equality is ...'
part of type declarations in interface files. This could cause the
compiler to use the wrong predicate in importing modules if the
user-defined equality predicate was not exported.

To avoid this problem we now require user-defined equality or
comparison predicates for a type to be defined in the same module
as the type. This is the only simple way to module qualify the
unification predicates in interface files without running
typechecking.

compiler/prog_io.m:
compiler/prog_io_pragma.m:
	Require user-defined equality or comparison predicates for a
	type to be defined in the same module as the type.

NEWS:
doc/reference_manual.texi:
	Document the change.

compiler/typecheck.m:
compiler/post_typecheck.m:
compiler/hlds_pred.m:
compiler/make_hlds.m:
compiler/*.m:
	Calls in the compiler generated unification predicates are
	now guaranteed to be fully qualified, so don't attempt
	to search for more qualified predicates.

	Add a new pred_marker 'calls_are_fully_qualifed' to
	identify predicates whose bodies contain only fully
	qualified calls, rather than testing the import_status.

tests/invalid/Mmakefile:
tests/invalid/exported_unify.m:
tests/invalid/exported_unify2.m:
tests/invalid/exported_unify2.err_exp:
	Test case.

Index: NEWS
===================================================================
RCS file: /home/mercury1/repository/mercury/NEWS,v
retrieving revision 1.311
diff -u -u -r1.311 NEWS
--- NEWS	7 May 2003 06:42:31 -0000	1.311
+++ NEWS	19 May 2003 03:30:13 -0000
@@ -10,7 +10,9 @@
   declarations and in the head of a clause for a user-defined field access
   function.
 * We now support impure higher-order code.
-* We now allow user-defined comparison routines.
+* We now allow user-defined comparison predicates.
+* User-defined equality and comparison predicates for a type are now
+  required to be defined in the same module as the type.
 
 Changes to the Mercury compiler:
 * Better support for incremental program development:
@@ -42,11 +44,14 @@
 
 Changes to the Mercury language:
 
-* We now allow user-defined comparison routines, using the syntax
+* We now allow user-defined comparison predicates, using the syntax
   :- type t ---> t where equality is unify_t, comparison is compare_t.
 
   See the "User-defined equality and comparison" chapter of the
   Mercury Language Reference Manual for details.
+
+* User-defined equality and comparison predicates for a type are now
+  required to be defined in the same module as the type.
 
 * Infix `.' is now accepted as a module name separator.  Hence it is
   now possible to write io.write_string and list.member to mean the
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------