[m-users.] Inferring over dynamic data

[Zoltan Somogyi]

2022-01-18 20:07 GMT+11:00 "Zoltan Somogyi" <zoltan.somogyi at runbox.com>:
> Example Prolog predicate that operates on these two predicates:
> 
> add_new_component(Item, NewComponent) :-
>    m(Item, OldComponents),
>   retract(m(Item, OldComponents),
>   NewComponents = [NewComponent | OldComponents], 
>   assert(m(Item, NewComponents)),
>   assert(s(Item)).
> 
> In Mercury, the direct equivalent of this code is
> 
> add_new_component(Item, NewComponent, S0, S, M0, M) :-
>   map.lookup(Item, OldComponents),
>   NewComponents = [NewComponent | OldComponents],
>   map.det_update(Item, NewComponents, M0, M),
>   set.insert(Item, S0, S).

The first line of the Mercury definition has a typo, which I fear
may confuse people just learning Mercury: what now reads

  map.lookup(Item, OldComponents)

should read

  map.lookup(M0, Item, OldComponents)

The difference is really important, since M0 represents the current
state of the map in which the lookup is done; without it, the lookup
cannot be done.

Zoltan.