The 42 xx 999999999999999 case is fixed with https://github.com/rakudo/rakudo/commit/f190f241e0 .
Not having an Empty indicate the end can only be done by breaking at least two spectests. Also: (|(),) xx 42 would then hang. So this feels like something that cannot be fixed generally, unless we accept the hanging behaviour of this example. I also have a patch for the thunked version of xx. But alas, this still breaks a couple of spectests that do: my $a = 0; ($a++,) xx 42; # note the xx Seq is in sink context So for some reason, it looks like the Seq.sink-all is not doing any pull-one’s. Still investigating! > On 8 Mar 2017, at 16:56, Zoffix Znet (via RT) <perl6-bugs-follo...@perl.org> > wrote: > > # New Ticket Created by Zoffix Znet > # Please include the string: [perl #130956] > # in the subject line of all future correspondence about this issue. > # <URL: https://rt.perl.org/Ticket/Display.html?id=130956 > > > > > > 15:53 m: dd (($++ ?? 42 !! |()) xx 9999999999999999999)[^4] > 15:53 camelia rakudo-moar 0633f0: OUTPUT: «(Nil, Nil, Nil, > Nil)» > 15:53 IOninja m: dd (($++ ?? 42 !! |()) xx 999)[^4] > 15:53 camelia rakudo-moar 0633f0: OUTPUT: «(42, 42, 42, 42)» > > > Also, giving huge values that don't yet trigger the above bug causes a hang, > perhaps it shouldn't be refying everything? > > m: dd (($++ ?? 42 !! |()) xx 99999999999999)[^4] > camelia │ rakudo-moar 0633f0: OUTPUT: «(timeout)»
