[m-rev.] for review: dependency tracking

[Mark Brown]

On 23-Apr-2002, Zoltan Somogyi <zs at cs.mu.OZ.AU> wrote:
> diff -u browser/declarative_analyser.m browser/declarative_analyser.m
> --- browser/declarative_analyser.m
> +++ browser/declarative_analyser.m
> @@ -53,35 +53,29 @@
>  	--->	subterm_in
>  	;	subterm_out.
>  
> -:- type subterm_origin(R)
> +:- type subterm_origin(T)
>  
> -			% Subterm came from an output of a child or sibling
> -			% call. The first argument records the id of the exit
> -			% event of the call. The second and third arguments
> -			% state which part of which argument is the origin.
> +			% Subterm came from an output of a child or sibling.
>  			%
> -	--->	output(R, arg_pos, term_path)
> +	--->	output(T, arg_pos, term_path)
>  
> -			% Subterm came from an input of the parent. The
> -			% arguments identify which part of which argument of
> -			% the clause head is the origin.
> +			% Subterm came from an input of the parent.
>  			%
>  	;	input(arg_pos, term_path)

Both of these changes were okay.  The only problem is that you should
replace "the id of the exit event of the call" with "the child or sibling
EDT node".

>  	% The analyser state represents a set of suspects.  We
>  	% consider one incorrect node at a time, and store its suspect
>  	% children.
>  	%
> -:- type analyser_state(R)
> +:- type analyser_state(T)
>  	--->	analyser(
>  				% Current incorrect node (initially `no').
>  				% This is the most recent node that the
>  				% oracle has said is incorrect.
>  				%
> -			maybe_prime	:: maybe(prime_suspect(R)),
> +			maybe_prime	:: maybe(prime_suspect(T)),
>  
>  				% Current suspects.
>  				%
> -			suspects	:: list(suspect(R)),
> +			suspects	:: list(suspect(T)),
>  
>  				% Previous prime suspects.
>  				%
> -			previous	:: list(suspect(R))
> +			previous	:: list(suspect(T)),
> +
> +			debug_origin	:: maybe(subterm_origin(T))

This field needs a comment to document it.

After you have addressed these remarks, and if the change passes its
bootcheck, then you can go ahead and commit it.

Cheers,
Mark.

--------------------------------------------------------------------------
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
--------------------------------------------------------------------------