Re^2: Pugs configuration - file locations

That's it. Setting PERL5LIB got past that problem. Now I get

pugs> my $z= "abc12345we678"
"abc12345we678"
pugs> $z ~~ ?/\d+/
*** Cannot parse regex: \d+
*** Error: Error: Can't use string ("1") as a subroutine ref while "st
+rict refs" in use at D:\pugs\perl5\Pugs-Compiler-Rule\lib/Pugs
/Compiler/Regex.pm line 85.

Bool::False
pugs>
[download]

Owell...

—John
in Allen, Texas (Dallas metroplex)

Comment on Re^2: Pugs configuration - file locations Download Code

Replies are listed 'Best First'.
Re^3: Pugs configuration - file locations by moritz (Cardinal) on Feb 27, 2008 at 08:59 UTC
The question mark shouldn't be there. Just `$z ~~ /\d+/`. If you want to coerce that to boolean context, the question mark has to be in front of the expression: `my $bool = ? $z ~~ m/\d+/` [download] (Don't know if ~~ or ? has the higher precedence, so you might need parenthesis. But I don't think so ;)	[reply] [d/l] [select]
Re^4: Pugs configuration - file locations by John M. Dlugosz (Monsignor) on Feb 27, 2008 at 09:08 UTC
Why? It seems more reasonable to tell the regex engine it has an easier job, close to where it is invoked rather than converting the result. IAC, both forms are taken by Pugs, and every variation (even with no ? ) give the same error.	[reply]
Re^5: Pugs configuration - file locations by moritz (Cardinal) on Feb 27, 2008 at 09:24 UTC
It seems more reasonable to tell the regex engine it has an easier job, close to where it is invoked rather than converting the result. Note that a regex match has side effects like setting `$0, $1` etc., and the full match object is still available in `$/` even if the match is performed in boolean context. So the optimatzion had to a lot smarter than just detecting boolean context (which propagates anyway, so it's not a question of the used syntax).	[reply] [d/l] [select]