Multi-byte unicode char leads to multi-line string alignment error

[tombentley]

Per https://github.com/ceylon/ceylon/issues/6284, and as shown by http://try.ceylon-lang.org/?gist=9b84d5688cef18c56bfa685efbedf3fd, a multi-byte character seems to upset the parser.

[bjansen]

As mentioned in https://github.com/ceylon/ceylon-ide-intellij/issues/153, this is because the indent is made of 1 tab + 3 spaces instead of 7 spaces. If the tab is replaced with 4 spaces, the problem disappears, so I don't think it's a bug.

[tombentley]

If that's the case surely tab in the web IDE should insert spaces and not tabs? And the error should be clear that there are tabs and spaces in the indentation.

[tombentley]

So it only worked in Eclipse because when I copied and pasted Eclipse kindly converted the tab into spaces.

Amusingly I can appease the typechecker by inserting extra spaces:

shared void m() {
    print("Hello ❄
              world!");
}

So it now looks unaligned in the source, but prints aligned when run.

[jvasileff]

the web IDE should insert spaces and not tabs

👍. I've also had tab alignment issues with the way code appears in gists vs. the web ide.