Get subset of associative array. Which of these is more readable? Is there another reason to use one over the other (or a different method)

Question

If I have an array that has more info than I need and I want to pull out a subset of its name value pairs I have come up with 2 methods.

Both rely on this:

  $detailkeys=array(title,summary,subtitle,url,imgsrc,description);

The method I came up with on my own:

  foreach ($detailkeys as $keyname) {
    $details[$keyname]=$data[$keyname];
  }

The method I found online:

  // get just the desired parts of $data and put them in $details.
  $details=array_intersect_key($data, array_flip($detailkeys)); 

The second method, with a decent comment, is what I chose, but wanted to see what anyone else had to say about it.

Answer 1

I think the second method is preferable. And this would seem perfectly readable, providing you are familiar with the array_intersect_key() function (although that's a quick lookup if you're not, and if you are it's almost self documenting). To be honest, I think your first method would also require a similar comment.

The second method also avoids an additional $keyname variable. And with it being a built-in function you would expect it to be optimised.

Have you tried benchmarking the results? (Although TBH with such a small set of "required" keys I can't imagine this being significant - I micro optimisation at best.)

You could perhaps avoid the need for your $detailkeys if you initialised your $details array with the required keys to begin with (if that is possible)...

$details = array (
    'title' => null,
    'summary' => null,
    // etc.
);

And then you also wouldn't need to call array_flip()...

$details = array_intersect_key($data, $details);

However, I would also query whether you needed to create a subset array to begin with. If you are only reading the data, passing it around, and not changing it, then it's not necessarily an optimisation and could even be the opposite.