| Subject: | docs of refaddr() |
| Date: | Wed, 23 Jun 2010 10:00:05 +1000 |
| To: | bug-Scalar-List-Utils [...] rt.cpan.org |
| From: | Kevin Ryde <user42 [...] zip.com.au> |
At the risk of showing my ignorance, for a long time I didn't quite
realize refaddr() was the same as numizing but the point of it is to
bypass overloading. It might be worth a couple of words in the docs to
make that clear, and also perhaps a bit about what can be assumed about
an address:
=item refaddr EXPR
If EXPR evaluates to a reference then return the internal memory
address of the thing referred to by EXPR. Otherwise return C<undef>.
$addr = refaddr "string"; # undef
$addr = refaddr \$var; # eg 12345678
$addr = refaddr []; # eg 23456784
$obj = bless {}, "Foo";
$addr = refaddr $obj; # eg 88123488
The address is the same as using a reference numerically C<$ref+0> per
L<perlref/Using References>, but C<refaddr> by-passes any overloaded
numizing a blessed reference might have from L<overload>.
The address of a target thing is constant and no two things have the
same address, though an address may be re-used after the thing is
destroyed.