Parse `&x`as a call expression

[GunnarFarneback]

In #6080 (comment) @JeffBezanson wrote:

Deprecation is in for 0.7. For 1.0, we should change the parsing of &x to a call expression.

This doesn't seem to have happened, cf https://discourse.julialang.org/t/ampersand-operator-not-consistent-with-help/34266. Is this something that can be changed for 1.x or would it have to wait for 2.0?

[JeffBezanson]

Given possible uses of &x syntax, the question is whether it's more useful for &(1, 2) to be the same as 1 & 2, or for it to apply & to the single tuple (1, 2).

[JeffreySarnoff]

If &(true, false) is going to remain a syntax error until ampersand(..) is designed and implemented, then please reconsider introducing and, or [ like xor, if \xor were ascii ] and supporting or(p, qs....), and(p, qs...) as Boolean operators.