[m-dev.] for review: accumulator.m 2

[Peter David ROSS]

On 03-Jul-1998, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:
> On 03-Jul-1998, Peter David ROSS <petdr at cs.mu.OZ.AU> wrote:
> > On 03-Jul-1998, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:
> > > On 01-Jul-1998, Peter David ROSS <petdr at students.cs.mu.oz.au> wrote:
> > > > 	%
> > > > 	% This type is threaded through the processing of the compose
> > > > 	% goals.  Its purpose is to help identify whether or not the
> > > > 	% goals as a whole are assocative.
> > > ...
> > > > 	% The Prev Call Map records for the original dynamic var whether
> > > > 	% previous calls have been commutative or not.
> > > > 	%
> > > > :- type rename
> > > > 	--->	rename(
> > > > 			list(var),		% Ys
> > > > 			list(var),		% Y0s
> > > > 			module_info,
> > > > 			set(var),		% Static vars.
> > > > 			set(var),		% Dynamic vars
> > > > 			multi_map(var, var),	% Original Var map
> > > > 			map(var, commutative)	% Prev call map
> > > > 		).
> > > 
> > > The meaning of the Prev Call Map is still a bit mysterious.
> > > Which previous calls do you mean?
> > > Does it record whether *all* previous calls have been commutative
> > > or whether just *some* previous calls were commutative?
> >
> > All previous calls must have been commutative.
> 
> OK, you should document that.
> 
> That only answers my second question; I still don't know which
> calls you mean by the "previous calls".
> 
Sorry.

While processing the "Compose" goals their exists a set of already
processed goals.  For all goals in this set which are calls which modify
a variable descended from the original dynamic variable we record the
commutivity of those goals as a whole.

Pete.