On Sat Apr 19 20:42:11 2014, Mouq wrote: > On Wed Aug 21 05:40:22 2013, masak wrote: > > <lizmat> r: .say for < $< $> > # rakudobug ? > > <camelia> rakudo e84dda: OUTPUT«$<$>» > > <benabik> Yeah. I would have expected "$<\n$>\n" > > <benabik> r: < $< $> >.perl.say > > <camelia> rakudo e84dda: OUTPUT«("\$<\$>",).list» > > <jnthn> Yeah. Odd > > * masak submits rakudobug > > There's other fallout from the same thing that causes this: > > $ perl6 -e'say qx[ echo [ echo abcd ] ]' > /bin/sh: line 1: ]: command not found > [abcd > $ echo [ echo abcd ] > [ echo abcd ] > > This happens because postprocess_* is called on each <nibble>. > Something like "< < a b c > >" is parsed as: > "<" ← <starter> > " " ← . > "<" ← <starter> > " a b c " ← <nibbler> > ">" ← <stopper> > " " ← . > ">" ← <stopper> > > The inside <nibbler> post-processes " a b c ", turning it into > ("a","b","c"). The outer <nibbler> subsequently turns that into > "<"~("a","b","c")~">" (or "<a b c>"), and post-processes that into "<a > b c>". A similar process happens for qx[].
Rather it *post-processes that into ("<a","b","c>")