Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:62678
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.217.170 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <CAPaKmRkJ0QRkADd857jLikthqDqrwhfH-EfNdKF_62T0-sJ2YA@mail.gmail.com>
References: <CAPaKmRkJ0QRkADd857jLikthqDqrwhfH-EfNdKF_62T0-sJ2YA@mail.gmail.com>
Date: Sun, 2 Sep 2012 14:04:54 -0400
Message-ID: <CAL+t6cHZXC8oZnMpvWaGW0Uj_W1o2Jjfre+ynX8bENww_qPMHw@mail.gmail.com>
To: Amaury Bouchard <amaury@amaury.net>
Cc: PHP Internals <internals@lists.php.net>
Content-Type: text/plain; charset=ISO-8859-1
Subject: Re: [PHP-DEV] Question about hashtables implementation
From: theanomaly.is@gmail.com (Sherif Ramadan)

> Hi all,
>
> I have a question about the internal implementation of PHP's hashtables. I
> did some researches, but I didn't find the answer.
>
> Here is an example of what I would like to understand.
> Start by creating an array:
>  $a = array();
>
> Fill it, using implicit and explicit keys:
>  $a[] = 'cc';
>  $a[1] = 'dd';
>
> If we look at what's inside, we get:
>  Array
>  (
>     [0] => cc
>     [1] => dd
>  )
>
> OK, everything is obvious. Now, I add a value at the beginning of the array:
>  array_unshift($a, 'ee');
>
> Do a print_r() again:
>  Array
>  (
>     [0] => ee
>     [1] => cc
>     [2] => dd
>  )
>
> As you can see, the keys for 'cc' and 'dd' have been recalculated. It works
> as espected.
> My question is how does it work? Are all numeric keys computed when the
> array_shift() is done? Or is the iterator calculating them on-the-fly?
>
> Thanks.
>
> Amaury

To clarify, this particular functionality you're using as an example
"array_unshift" really isn't specific to the internal implementation
of hashtables in PHP. That is to say that this side-effect you're
describing is specific to that function and not necessarily
hashtables' internal implementation.

Essentially, array_unshift() just rebuilds the array. From
http://php.net/array-unshift "All numerical array keys will be
modified to start counting from zero while literal keys won't be
touched"

If we were to write a PHP user-land implementation of array_unshift()
it would probably look something like this...

function array_unshift(&$array) {
  $new_array = array();
  foreach (array_slice(func_get_args(),1) as $e) {
    $new_array[] = $e;
  }
  foreach ($array as $k => $v) {
    if (is_numeric($k)) {
      $new_array[] = $v;
    } else {
      $new_array[$k] = $v;
    }
  }
  $array = $new_array;
  return count($array);
}

If you're interested in seeing the actual code for array_unshift's
internal implementation you can take a look here:

http://lxr.php.net/xref/PHP_5_4/ext/standard/array.c#2049

As for how the internal implementation of hashtables in PHP, that can
be quite an extensive topic. PHP uses hashtables for pretty much
everything internally. The hashtable struct can be seen here
http://lxr.php.net/xref/PHP_5_4/Zend/zend_hash.h#66

If you're interested in a more detailed explanation of the
implementation of PHP's Array type you might find this article by
Nikic useful: http://nikic.github.com/2012/03/28/Understanding-PHPs-internal-array-implementation.html