[mercury-users] Array bounds

[Ralph Becket]

>From the interface, it looks like array.m may eventually handle arrays
that aren't indexed from 0.  This would be no bad thing, but because
arrays currently are indexed from 0 I suspect that assumption has been
hard coded into most of the code around that manipulates arrays (if
only because it's less verbose...)

That is, I suspect common practice is to write

    ..., foo(A, 0, X), ...

:- func foo(array(T1), int, T2) = T2.
:- mode foo(array_ui, in, in) = out is det.

foo(A, I, X) =
    ( if I < array__size(A) then 
        foo(A, I + 1, bar(X, array__lookup(A, I)))
      else
        X
    ).

rather than

    ..., foo(A, array__min(A), X), ...

:- func foo(array(T1), int, T2) = T2.
:- mode foo(array_ui, in, in) = out is det.

foo(A, I, X) =
    ( if I =< array__max(A) then 
        foo(A, I + 1, bar(X, array__lookup(A, I)))
      else
        X
    ).

If the array type interface is changed to allow non-0 based arrays
then code of the first type is going to break.  So what's the plan?

Ralph

--
Ralph Becket      |      MSR Cambridge      |      rbeck at microsoft.com 
--------------------------------------------------------------------------
mercury-users mailing list
post:  mercury-users at cs.mu.oz.au
administrative address: owner-mercury-users at cs.mu.oz.au
unsubscribe: Address: mercury-users-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-users-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------