Hi Damian --
I'm heading for the door, but wanted to attempt to get a response into
your hands before the day was through:
> // this fails <-----------------!!!!!!
> // proc testOfTuple1(b : (int(16), int(16), int(?w))) : real(w)
I believe this ought to work and am _slightly_ surprised it does not. That
said, as I've mentioned on another thread recently, our wildcard formal
argument pattern matching support is a bit piecemeal at present and needs
to be beefed up, so I'd guess that's what you're hitting. A wildcard-free
way to handle this case _may_ be (sorry I don't have time to test):
proc testOfTuple(b ...): real(numBits(b(3).type))
Ugly, though... For this reason, I might tend to use the more
type-inferred version and create 'chpldoc' comments to describe what's
expected (but there are good reasons to want the more explicitly typed
form as well, so I'm not suggesting that's a solution for all cases).
> var w : int;
...
> w = numBits(b(3).type);
...
> // this fails <-----------------!!!!!!
> // return (e * m):real(w);
Here, the issue is much simpler: the argument to real() must be a param
(compile-time known value). Happily, numBits() returns a param, so the
fix should simply be:
param w: int = numBits(b(3).type);
return (e * m): real(w);
or simply:
return (e * m): real(numBits(b(3).type));
Hope this is helpful...
-Brad
------------------------------------------------------------------------------
_______________________________________________
Chapel-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-users
