So, this is a mystery... I'd look at /usr/lib/perl5/site_perl/5.8.8/Data/Serializer.pm line 597 and start tracing back from there. The later messages may or may not be the consequence of the first unitialised value.
I had a quick look at the current version (Data::Serializer v0.48) on CPAN. Around line 597 I see:
so, if that's the version you're using, it appears that $line is undefined. Seems that _get_token is used only in deserialize(), where it operates on its first argument (after $self), so:592: sub _get_token { 593: my $self = (shift); 594: my $line = (shift); 595: #Should be anchored to beginning 596: #my ($token) = $line =~ /\^([^\^]+?)\^/; 597: my ($token) = $line =~ /^\^([^\^]{1,120}?)\^/; 598: return $token; 599: }
I see that deserialise() is used in a number of places...my $self = (shift); if ($self->raw) { return $self->raw_deserialize(@_); } my $value = (shift); my $token = $self->_get_token($value);
I would force the initial warning to confess (eg: local $SIG{__WARN__} = sub { confess(@_) };) and see what the stack traceback tells me.
In reply to Re: Magic number checking on storable string failed
by gone2015
in thread Magic number checking on storable string failed
by shanu_040
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |