Right here’s a cool little problem introduced up this morning by a buddy. Think about the next code:
change foo {
case .a: return "a"
case .b(let str) the place str.hasPrefix("c"), .c: return "c"
case .b: return "b"
}
It gained’t compile.
Once you bind an emblem for one sample, you have to bind that image for each sample in a case. This prevents you, for instance, from binding str in a single sample after which making an attempt to make use of str within the shared case physique. For instance, contemplate this case. What would you count on to occur when foo is .c?
func switchTheFallthroughOrder(foo: Foo) -> String {
change foo {
case .a: return "a"
case .b(let str) the place str.hasPrefix("c"), .c:
// Utilizing `str` right here is unhealthy!
print(str)
return "c"
case .b: return "b"
}
}
Regardless of my first knee-jerk refactoring, transferring out the .c case to make use of fallthrough doesn’t work. Once more, it is because str will not be sure for .c and might be used within the successive case physique:
Nevertheless, as Greg Titus identified, when you change the order to make use of the binding case first with fallthrough, Swift is aware of at compile time that the binding gained’t keep it up past that scope. This resolves the error, since str is simply used within the the place clause to slender the sample matching:
Additional, when utilizing bindings in case assessments, a waterfall method the place the sure objects are used earlier than fallthrough can lengthen by means of a number of steps with the blessing of the compiler:
case .widest(let first, let second) the place first.satisfiesACondition():
// can use `first`, `second` right here
fallthrough
case .medium(let second) the place second.satisfiesAnotherCondition():
// can use `second` right here even when it was sure
// by way of `widest` above by way of fallthrough
fallthrough
case .narrowest: return someValue
My due to Greg Titus for figuring this all out!
Associated
