From @masak

<masak> awww.
<masak> who would have guessed, declaring a sub infix​:<;> causes problems...
<masak> rakudo​: our sub infix​:<;>($lhs, $rhs) {}; { say "A"; say "B" }
<p6eval> rakudo 5ac05e​: ( no output )
<masak> :(
<masak> rakudo​: our sub infix​:<;>($lhs, $rhs) { say $lhs, $rhs }; 1; 2; 3
<p6eval> rakudo 5ac05e​: OUTPUT«infix​:<;>1␤Bool​::True2␤Bool​::True3␤»
<masak> I believe this is backwards. the stopper ';' should take
precedence over any declared infixes, unless we're inside a
parenthetical thingy.
* masak submits rakudobug

From @masak

On Sat Apr 16 05​:37​:53 2011, masak wrote​:

<masak> awww.
<masak> who would have guessed, declaring a sub infix​:<;> causes
problems...
<masak> rakudo​: our sub infix​:<;>($lhs, $rhs) {}; { say "A"; say "B" }
<p6eval> rakudo 5ac05e​: ( no output )
<masak> :(
<masak> rakudo​: our sub infix​:<;>($lhs, $rhs) { say $lhs, $rhs }; 1;
2; 3
<p6eval> rakudo 5ac05e​: OUTPUT«infix​:<;>1␤Bool​::True2␤Bool​::True3␤»
<masak> I believe this is backwards. the stopper ';' should take
precedence over any declared infixes, unless we're inside a
parenthetical thingy.
* masak submits rakudobug

<jnthn> What does STD think?
<jnthn> std​: sub infix​:<;>($, $) {}; { say "A"; say "B" }
<p6eval> std 4608239​: OUTPUT«ok 00​:01 122m␤»
<jnthn> I'm thinking that example doesn't force it to reveal what it's
parsing that ; as though
<jnthn> And wondering how to construct one that does :)
<colomon> without executing the code, I can't see how to do it.
<jnthn> oh, maybe you can do it with statement modifiers
<jnthn> std​: sub infix​:<;>($, $) {}; { say "A" if 1; say "B" if 1; }
<p6eval> std 4608239​: OUTPUT«ok 00​:01 122m␤»
<jnthn> I expect that'd fail if the ; wasn't being parsed as a statement
sep there.
<masak> jnthn++
* masak submits rakudobug
<jnthn> You already did submit the Rakudo bug! :P
<masak> apparently I'm getting senile.
<masak> ok, I'll just amend that ticket, then.
<colomon> std​: (say "A" if 1) + 10
<p6eval> std 4608239​: OUTPUT«ok 00​:01 121m␤»
<colomon> jnthn​: are you sure?
<masak> colomon​: yes, it's the 'if' twice that makes it work.
<jnthn> std​: say "A" if 1 + say "B" if 1
<p6eval> std 4608239​: OUTPUT«�[31m===�[0mSORRY!�[31m===�[0m␤Confused [...]
FAILED 00​:01 120m␤»

@masak - Status changed from 'new' to 'open'

From @moritz

Now tested in S06-operator-overloading/semicolon.t

@moritz - Status changed from 'open' to 'resolved'