On Sat Jan 04 02:26:52 2014, masak wrote: > On Sun Dec 29 08:24:44 2013, FROGGS.de wrote: > > <FROGGS> p: say qq:to/EOF/ ~~ /\t/;\thelloEOF# the tab is > > preserved > > <camelia> rakudo-parrot c14a84: OUTPUT«「 」» > > <FROGGS> p: say qq:to/EOF/ ~~ /\t/; \thello EOF# now it is > > gone > > <camelia> rakudo-parrot c14a84: OUTPUT«Nil» > > > > In the second example it is meant to strip four space chars, but > > additionally it converts the tabs to space chars. Which is bad for > > e.g. Makefiles. > > I agree. S32/Str is a bit vague about what to do in this case. It > talks about different policies (spaces only, only some other \h char, > or mixed types) when *adding* spaces, but it doesn't mention what to > do when removing spaces. > > I'm fine with clarifying the spec, as long as we err on the side of > sanity/consistency. Feel free to run any patches through me for > review.
TimToady++ confirmed that \t falls under the same rule as \r, \n, and so forth - that is, we should be interpolating it after considering the dedent. I've implemented that now, and it's tested in S02-literals/quoting.t. /jnthn