I just tried this with today's rakudo.moar build and got:

$ cat txt
foo bar
foo bar
$ cat foo.p6
my $layn = slurp "txt";

$layn ~~ s:g/(\w+) \s+ (\w+)/$1 $0/; say $layn;

$ ./perl6 foo.p6
bar foo
bar foo

On Fri, May 2, 2014 at 2:02 PM, Schwenn, Peter <[email protected]> wrote:
> Designation:  Non-Finmeccanica
>
> Moritz,
>
> Thanks for the response.  I get the same successful result as you do with:
>
> my $line = "bar foo"; $line ~~ s/(\w+) \s+ (\w+)/$1 $0/; say $line # output:
> foo bar
>
> All well and good.  And if the pattern is one line I get success with my 
> other s///; forms as well.
>
> However I'm working with a multi-line slurped source text, and in that case 
> s///; reports no error and does no substitution.
>
> So for example if file   foobar.txt   contains these two lines:
>
> foo bar
> foo far
>
> and is read with:
>
> my $layn = slurp "foobar.txt";
>
> then
>
> $layn ~~ s:g/(\w+) \s+ (\w+)/$1 $0/; say $layn;
>
> Makes no complaint and makes no substitution ($layn is still foo bar foo bar 
> ).  (Same result for no ':g').
>
> Perhaps Slurping is not what I should be doing -- I'm using it because great 
> speed gain.
>
> Thanks,
>
> Peter Schwenn
>
>
>
> -----Original Message-----
> From: Moritz Lenz via RT [mailto:[email protected]]
> Sent: Friday, May 02, 2014 3:51 AM
> To: Schwenn, Peter
> Subject: Re: [perl #121772] $x ~~ s///;
>
> On 04/30/2014 05:21 PM, [email protected] wrote:
>> # New Ticket Created by  Schwenn, Peter # Please include the string:
>> [perl #121772] # in the subject line of all future correspondence
>> about this issue.
>> # <URL: https://rt.perl.org/Ticket/Display.html?id=121772 >
>>
>>
>> Designation: Non-Finmeccanica
>> Dear Perl6'rs
>>
>> [In short: no variation of   $x ~~ s/...(...)../....$0..../;   seems to work 
>> (to substitute any capture).]
>
> With a current Rakudo, I cannot reproduce this:
>
> my $line = "bar foo"; $line ~~ s/(\w+) \s+ (\w+)/$1 $0/; say $line # output:
> foo bar
>
> So either the bug has been fixed in the mean time (and a Rakudo Star release 
> is upcoming today), or you need to produce a runnable piece of code 
> (including example data) with which we can replicate your problem.
>
> Cheers,
> Moritz
>
>
>
> 3.3.042312



-- 
Will "Coke" Coleda

Reply via email to