Elizabeth Mattijsen
Elizabeth Mattijsen
To make it less likely that you accidentally overwrite it / make it possible to create your own "default" ?
Agree a test case could be good. Although I am not sure the reason the RakuAST version doesn't bomb for the right reason :-)
Test added with https://github.com/rakudo/rakudo/commit/314eb11437
It should really produce the same error as: ``` % raku -e 'sub a() { (^10).hyper(:1batch).map: { return if $_ == 5 } }; a; say "after"' A worker in...
This issue is really that it gets an `IntStr` back. It shouldn't.
At one point we need to test *all* subclassing of core classes, not just `Str`. For instance, I'm not sure subclassing `Date` and `DateTime` is currently tested. But yeah, we...
Test added with https://github.com/rakudo/rakudo/commit/d2f12f3520
```raku # lib/A.rakumod sub MAIN(|c) is export { dd c; say "BUG!" } ``` then: ``` % raku -c -Ilib -MA -e '' foo bar \("foo", "bar") BUG! Syntax OK...
Indeed, changing the signature to `sub foo(@a, Bool :$sorted-keys) {` makes the problem go away. Anyways, it seems to affect things that are pre-compiled: This code produces the correct error...
No, I don't think so. In the end in this case, the error message is LTA. My code example *was* doing something wrong, and the error message sent me on...