[Nickle]Re: initializer bug?

[Bart Massey]

Nevermind.  Changing the initialization to

  cell[size, size] cells = {
      {(cell){val = cellval.BLANK,  wall = (bool[4]){false ...}} ...}
      ...
  };

seems to have done the trick.  Frustrating that it has
enough information to typecheck but not to parse, though.

	Bart

In message <E17kH7S-00019X-00@bart.cs.pdx.edu> you wrote:
> Attached is a code fragment initializing an array of
> structures containing an array: the initializer doesn't
> work.  Is this a bug in my code, or in Nickle?  I suspect
> the latter.  (Additionally, why should Nickle need the
> "(cell)" in the initializer?  It seems like it should be
> able to infer this from the lhs type, as does C, but it
> seems to get even more lost without it.)
> 
> Thanks,
> 
> 	Bart
> 
>   /*
>    * Ricochet Robots solver
>    * Bart Massey 2002/8
>    */
> 
>   int size = 16;
> 
>   typedef enum {
>       BLANK, ME, DRONE, TARGET
>   } cellval;
> 
>   typedef struct {
>       cellval val;
>       /* 0 == N, 1 == E, 2 == S, 3 == W */
>       bool[4] wall;
>   } cell;
> 
>   cell[size, size] cells = {
>       {(cell){val = cellval.BLANK,  wall = {false ...}} ...}
>       ...
>   };
> 
> 
> _______________________________________________
> Nickle mailing list
> Nickle@nickle.org
> http://nickle.org/mailman/listinfo/nickle