Unexpected interaction between infix forms and function calls

[shawwn]

It turns out that

((not foo) x y z)

is identical to

(not (foo x y z))

E.g. ((not string?) 42) evaluates to true.

At first I was ready to call this a bug, but it’s a convenient way to negate the result of a function call, since you don’t have to wrap the entire arglist in an extra set of parens.

Still, I’m inclined to file this as a bug since it breaks the macro system. If you define foo as both a macro and a function, then (not (foo x y z)) will expand foo properly, whereas ((not foo) x y z) won’t.