Hello fionbarr,

Suppose you were silly unfortunate enough to have code like this:

 2:56 >perl -wE "my %h = ( A => 'B' ); my $h = { A => 'C' }; say $h{A}
+; say $$h{A}; say $h->{A};"
B
C
C

 3:15 >
[download]

That is, two (separate, unrelated) hashes both called h: a hash named %h, and an unnamed hash accessible via the scalar reference $h. $h{A} and $$h{A} would then be quite different things, but they would look so similar that it would be only too easy to confuse them. In a case like this, the value of the dereference syntax $h->{A} becomes apparent: it makes the dereference operation visually distinctive, and so reduces the danger of confusion.

Anyway, those of use who came to Perl from C++ find the arrow dereference syntax comfortingly familiar; it just “looks right.” ;-)

But, as always, TMTOWTDI, and YMMV.

Hope that helps,

informative reply: thanks

Whenever I see code like: print $$hash_ref{$key}; I have to think hard as to whether that is equivalent to:

print ${ $hash_ref }{ $key };
[download]

Or:

print ${ $hash_ref{ $key } };
[download]

With the second form, I don't have that indecision.

In a similar vein, given this sub:

sub x{ print $$_[0][1] };;
[download]

What do you think these two calls print?:

x( [['a','b'],['c','d']] );;

x( ['a','b'],['c','d'] );;
[download]

Produces:

Is it clearer if the sub is defined this way?:

sub x{ print $_[0]->[1] };;

xx( [['a','b'],['c','d']] );;

xx( ['a','b'],['c','d'] );;
[download]

Produces:

Look at references quick reference for a minute, then look at your question dereference hash, which syntax do you remember and are 100% sure what it does?

Great responses already...

I use what is easier to spot when I'm glancing at code. This:

$hash->{thing};
[download]

is easier to see at a glance than:

$$hash{thing};
[download]

Likewise, I find:

%{$hash};
[download]

easier to understand in a quick pass than:

%$hash
[download]

-stevieb