I think excluding the upper bound (..<) does not address the problem correctly.
The problem occurs on including the upper bound.
Consider the following simple program
#!/usr/bin/env swift
if let n: UInt8 = UInt8(Process.arguments[1])
{
for i in 0...n
{
print (i)
}
}
this program compiles and may even pass some testing. It will run correctly for
all values except 255, for which it crashes. The problem is that no one knows
how many of these cases we do have in current production code and tracking
these issues down may be almost impossible.
I would say, the take home message for know is, don’t use 0…n (i.e. include the
upper bound) unless you are sure that the upper bound cannot be equal to the
max value of the type.
I have not checked out the Swift3 branch which should fix this. Currently, with
Swift2.2 I think this is a serious issue.
Daniel
> On 08.04.2016, at 19:47, Cody Weaver <codyleewea...@gmail.com> wrote:
>
> I think an easy answer if I understand the question right is.
>
> let a: UInt8 = UInt8.min
> let b = UInt8 = UInt8.max
>
> for c in a..<b {
> print(c)
> }
>
> On Fri, Apr 8, 2016 at 10:27 AM Daniel Strobusch via swift-users
> <swift-users@swift.org> wrote:
> Hi,
> just as a side note, I filed a bug on this some time ago:
>
> https://bugs.swift.org/browse/SR-1091
>
> so let’s hope, this gets fixed in Swift3. Its really annoying and dangerous
> because the error is not recovered at compile time.
>
> Daniel
> _______________________________________________
> swift-users mailing list
> swift-users@swift.org
> https://lists.swift.org/mailman/listinfo/swift-users
_______________________________________________
swift-users mailing list
swift-users@swift.org
https://lists.swift.org/mailman/listinfo/swift-users
