> On 29 Dec 2015, at 17:35, Rúbio Terra (via RT) <[email protected]>
> wrote:
>
> # New Ticket Created by Rúbio Terra
> # Please include the string: [perl #127075]
> # in the subject line of all future correspondence about this issue.
> # <URL: https://rt.perl.org/Ticket/Display.html?id=127075 >
>
>
> I can declare non-ascii tokens without problems on a grammar but I get an
> encoding error if I refer to the token using the <> syntax. This happens on
> MoarVM, it works fine on JVM.
>
> This runs OK:
>
> perl6 -c -e 'grammar { token TOP { <numero>+ }; token número {<< \d+ >>} }'
> Syntax OK
>
> Now, if I reference the new token (número) I get an error:
>
> perl6 -c -e 'grammar { token TOP { <número>+ }; token número {<< \d+ >>} }'
> ===SORRY!===
> Error encoding ASCII string: could not encode codepoint 250
>
> The same code on JVM runs OK:
>
> perl6-j -c -e 'grammar { token TOP { <número>+ }; token número {<< \d+ >>} }'
> Syntax OK
>
> I'm running Linux Fedora 22. Both Rakudo 2015.11 and 2015.12 exhibit the
> error.
This looks like it is happening deep in the bowels of NQP:
$ 6l 'grammar { token TOP { <número>+ } }'
Error encoding ASCII string: could not encode codepoint 250
at gen/moar/stage2/NQPP6QRegex.nqp:1382
(/Users/liz/Github/rakudo.moar/install/share/nqp/lib/NQPP6QRegex.moarvm:capnames:293)
from gen/moar/stage2/NQPP6QRegex.nqp:1397
(/Users/liz/Github/rakudo.moar/install/share/nqp/lib/NQPP6QRegex.moarvm:capnames:478)
from gen/moar/stage2/NQPP6QRegex.nqp:1361
(/Users/liz/Github/rakudo.moar/install/share/nqp/lib/NQPP6QRegex.moarvm:capnames:52)
from gen/moar/stage2/NQPP6QRegex.nqp:1313
(/Users/liz/Github/rakudo.moar/install/share/nqp/lib/NQPP6QRegex.moarvm:qbuildsub:112)
Hope somebody with NQP internals knowledge can pick this up.
Liz
