[Grace-core] Some comments from Tijs on parsing Grace

[James Noble]

On 27/07/2016, at 23:09PM, Tijs van der Storm <storm at cwi.nl> wrote:

> but I guess this is about having both varags and lineups

not quite - the issue is that if you allow multiple-args "destructuring" 
(don't have to be variadic) in definitions: 

method single(a) { ... } 
method double(a,b) { ... } 

then when you're dynamically typed, there are problem cases 

single(1)  // works fine - "a" is bound to 1 
double(1) // probably error from destructuring on receiver side
double (1,2) // ,  builds tuple, descrtructres fine
single (1,2) // "a" is bound to tuple (1,2)
single (1,) // "a is bound to oneple (1,) 

from here, it's an obvious fix to allow only single arguments,
and have things explicitly destructure --- which is close to lineups
(except we also allow multiple flat arguments).

Either way, you have to make *another* distinction on the *caller* side
between calling

single(1)  // works fine - "a" is bound to 1 
single[1]  //works fine - "a" is bound to [1] 

we already have that of course with 

single{1}  but we seem to get away with that... 

> (Btw, comma would then just be cons, not flat-cons, if you assume it's right associative, but I guess this is off limits for Grace ;). 

yep.  J