From @moritz

08​:19 < moritz> nom​: my $x = '1}; say "pwnd"; #'; 'a' ~~ /<$x>/
08​:19 <+p6eval> nom 317f89​: OUTPUT«pwnd␤pwnd␤»
08​:19 < moritz> ouch
08​:20 < jnthn> ...ouch.
08​:23 * moritz submits rakudobug

From @FROGGS

better?

<FROGGS> r​: my $x = '1}; say "pwnd"; #'; 'a' ~~ /<$x>/
<p6eval> rakudo bf472b​: OUTPUT«pwnd␤»

The RT System itself - Status changed from 'new' to 'open'

From @moritz

On Wed Nov 21 08​:24​:32 2012, FROGGS.de wrote​:

better?

<FROGGS> r​: my $x = '1}; say "pwnd"; #'; 'a' ~~ /<$x>/
<p6eval> rakudo bf472b​: OUTPUT«pwnd␤»

Not at all. It should complain that
1}; say "pwnd"; #

is not a valid regex.

From @usev6

Nowadays this no longer says "pwnd", but complains about two terms in a row​:

$ perl6 -e 'my $x = Q[1}; say "pwnd"; #]; "a" ~~ /<$x>/'
===SORRY!=== Error while compiling EVAL_0
Two terms in a row
at EVAL_0​:1
------> anon regex { ^1};\c[32]⏏say\c[32]"pwnd";\c[32]#}
  expecting any of​:
  infix stopper
  infix or meta-infix
  statement end
  statement modifier
  statement modifier loop

skids++ added a test to S05-interpolation/regex-in-variable.t (commit Raku/roast@5c539db0a7) which checks whether the command dies.

Is that good enough to close the ticket?

From @usev6

Output has changed​:

$ perl6-m -e 'my $x = Q[1}; say "pwnd"; #]; "a" ~~ /<$x>/'
pwnd
Cannot find method 'CALL-ME'
  in block <unit> at -e​:1

$ perl6-m --version
This is perl6 version 2015.10-7-gca1cf74 built on MoarVM version 2015.10

From @TimToady

A new restricted dialect of regex is implemented in 28ab83f947b4899a4f8698eee5bc056742f356f1 and 19d84be0066978f616ace6fa9f506e742161a378

Tests in 1becd7c9b456b707a14bfba40d672ec28945f199

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