Front page | perl.perl5.porters |
Postings from July 2009
Citeren Dave Mitchell <davem@iabyn.com>:
> On Tue, Jul 21, 2009 at 12:45:04PM +0200, Rafael Garcia-Suarez wrote:
>> Ah, you're right. The parser has this line :
>> switch : label GIVEN '(' remember mydefsv mexpr ')' mblock
>> where mydefsv is an empty rule that lexicalizes $_.
>> But then the code I pasted above could probably be simplified.
>
>
> So if I've understood this correctly, given() adds a lexical $_ to the
> scope, and there's a bug in List::Util::first in that it doesn't work with
> a lexical $_? So not a bug in given/smartmatch?
I'm not sure if 'bug' is the correct word.
What happens: (example with for instead of given)
#!/usr/bin/perl -l
for my $_ (1) {
my $c = sub { print $_ };
$c->();
}
sub foo {
my $code = shift;
$_ = "abc";
$code->();
}
__END__
$c is a closure and closes over $_.
It then passes the code reference to the function foo which modifies
the global $_ and calls the closure.
But: the closure closes over $_ so it will always uses the value 1...
There are, as fas as I can see, two solutions:
1) modify given() so it doesn't use a lexical $_.
2) update the documentation for given() and make it more explicit that
it uses a lexical $_ and that every block/code reference inside it
needs our $_; before using $_.
Best regards,
Bram
Thread Previous
|
Thread Next