Utf.iter8 throws for 0x7F on Neko

[Aurel300]

class Main {
  public static function main():Void {
    var str = "\x7Fa";
    haxe.Utf8.iter(str, cp -> trace(cp));
  }
}

Expected output: 127 and 97 (works on Python, JS, Java, eval, C++, …)

Neko v2.2.0:

Called from ? line 1
Called from Main.hx line 4
Called from /.../haxe/std/neko/_std/haxe/Utf8.hx line 72
Uncaught exception - std@utf8_iter

[Aurel300]

This line is the culprit:

https://github.com/HaxeFoundation/neko/blob/master/libs/std/utf8.c#L362

(should be <= 0x7F)

There is a similar problem in HashLink, though it does not throw:

https://github.com/HaxeFoundation/hashlink/blob/master/src/std/string.c#L89

Additionally neither C0, C1 (overlong sequence start), nor F5 – FF (codepoints higher than allowed in Unicode) are rejected.

[ncannasse]

I have fixed the 0x7F in HashLink. I'm not sure what to do with C0-C1 and F5-FF, you can submit a PR for hashlink string.c if you want to improve things here I will not fix Neko but you can submit a PR there as well.

…

On Wed, Apr 10, 2019 at 12:11 PM Aurel ***@***.***> wrote: This line is the culprit: https://github.com/HaxeFoundation/neko/blob/master/libs/std/utf8.c#L362 (should be <= 0x7F) There is a similar problem in HashLink, though it does not throw: https://github.com/HaxeFoundation/hashlink/blob/master/src/std/string.c#L89 Additionally neither C0, C1 (overlong sequence start), nor F5 – FF (codepoints higher than allowed in Unicode) are rejected. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#8139 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA-bwAb17Zs8Gb2-nLPHfHHBP9JSm6tWks5vfbjogaJpZM4cmjqD> .

[ncannasse]

@Aurel300 can you submit a PR for Neko?

[ncannasse]

Nevermind I fixed it directly online.