Update to Swift 5

[lewisbar]

I've made a few updates to switch to Swift 5, the biggest being replacing "Hashable.hashValue" with "hash(into:)". I hope I did everything correctly there.
I have not yet seen into updating Completer, I'd have to do some reading about Strings first.

[lewisbar]

Oh, the tests cause a crash because Completer is not fixed yet.

[lewisbar]

That was just a fix for the warnings in Completer. The tests still crash.

[lewisbar]

Only these two tests crash:

• test_ColonAfterCase_WithEmojis_AdoptsSwitchIndentation()
• test_ColonAfterDefault_WithEmojis_AdoptsSwitchIndentation()

It's an index-out-of-bounds error, by the way.
Feels like I'm getting closer to to root of the problem.

[lewisbar]

Another thought, for anyone reading this, maybe someone knows: The cause must be some change between Swift 3 and Swift 4.2, I think, because the tests used to pass back then. I hope I find the time to look further into it.

[lewisbar]

Oh, of course the problem could have been introduced by dff687b ... That's actually quite likely, I almost forgot about that.

[lewisbar]

dff687b seems correct, though. Next step: I must look at the two crashing tests and inspect their results. Maybe at some point in the next days.

[lewisbar]

I wonder if this is a Swift bug.
When I do

var newText = text.replacingCharacters(in: selection, with: insertion)

the test crashes with "String index is out of bounds" when trying to use the selection, for example with newText[selection]. (selection is in this case a Range<String.Index> with a length of 0, so you could think of it as the cursor or insertion point. The insertion is a colon ":".)

But when I do

var newText = text
newText.insert(Character(insertion), at: selection.lowerBound)

the test passes, and newText[selection] works as expected.

But the utf16.count is the same for both strings, as well as the count, and they are even ==.

Still when I do

let newRange = newText.index(stringRange.lowerBound, offsetBy: 1)..<newText.index(stringRange.upperBound, offsetBy: 1)
newText[..<newRange.lowerBound]

it works for the string that has been changed by insert(at:), but crashes for the string that has been created by replacingCharacters(in:with:). How is that possible when both are ==?

The problem is, I need to use replacingCharacters(in:with:), because the user could have selected multiple characters when entering a new one. Also, insert(at:) can only insert single characters, so it wouldn't work if the user pastes several characters. Maybe I could build a workaround, but the question is: Is this a bug in Swift and should be reported?

EDIT:
In the playground, the error is not "String index is out of bounds" but "error: Execution was interrupted, reason: EXC_BAD_INSTRUCTION (code=EXC_I386_INVOP, subcode=0x0).
The process has been left at the point where it was interrupted, use "thread return -x" to return to the state before expression evaluation." (But still only for the string that's created by replacingCharacters(in:with:).)

[lewisbar]

Here is a playground in which I tried to show the bug as clearly as possible. https://github.com/lewisbar/ReplacingCharactersBug