Restructure XML in PHP

Question

I'm provided a XML file with this structure:

<items>
  <item>
    <images>
      <image>A</image>
      <image>B</image>
      <image>C</image>
    </images>
    .
    .
    .
  </item>
</items>

However the import function of my Shop requires the following format:

<items>
  <item>
    <images>
      <image>A</image>
      <image1>B</image>
      <image2>C</image>
    </images>
    .
    .
    .
  </item>
</items>

At first I was thinking I would do this simply in Java since it would be pretty easy to read line by line and restructure the document but I would love to have it so I can just visit a url and this is done for me.

Here is the approach I took:

<?php
$xml = simplexml_load_file('data.xml');


// Loop over items in original xml
for($i = 0; $i < count($xml->item); $i++)
{
    $images;
    if( ($c = count($xml->item[$i]->images->image)) > 1)
    {
        $images = $xml->item[$i]->images;

        // Remove entry
        unset($xml->item[$i]->images);

        $xml->item[$i]->addChild('images');

        for($y = 0; $y < count($images->image); $y++)
        {
            if($y == 0)
            {
                $xx = $xml->item[$i]->images->addChild('image', $images->image[$y]);

            }else {
                $xml->item[$i]->images->addChild('image' . $y, $images->image[$y]);
            }
        }

        var_dump($images);
    }
}

$xml->asXML('POTO.xml');

The dilemma I have tho is that none of the childs get added to images. I have been told I need to restructure the whole document but this is kind of silly if I var_dump just after removing the images node the node and it's children are all removed however when I go to add images node back to item node it and var_dump the node it shows the node was added to item node as a child however when I try to add image to images nothing gets added.

Do I really need to restructure the whole document because it seems simpler to do it in Java then. Or did I miss something.

What is the value of count($xml->item) and count($images->image)? Are your loop bodies executed? — George Cummins
– George Cummins, Commented Feb 12, 2014 at 23:09
@GeorgeCummins The count syntax looked funny to me, too, but seems to be valid. Looks like PHP 5.3 brought support for a SimpleXMLElement::count() object method, but using the basic count() was actually standard before that. Who knew? Ref: us3.php.net/manual/en/simplexmlelement.count.php — eimajenthat
– eimajenthat, Commented Feb 13, 2014 at 0:13

eimajenthat · Accepted Answer · 2014-02-13 00:08:51Z

Okay, I think the key problem here is that assigning the images to $images doesn't work like you'd expect. While such an assignment of a primitive makes a copy, assigning an object makes a reference to the same object (similar to a pointer you've worked with lower level languages). They do this because an object can be larger than a primitive, and unnecessary duplication of that is just wasteful.

However, in this case, the result is that when you unset the images from the main SimpleXML object, you're also destroying the instance in $images. Fortunately, there is a solution. You can clone the object, rather than assigning it. See http://www.php.net/manual/en/language.oop5.cloning.php

I modified your sample code here and it seems to work: http://3v4l.org/ZmRZV

Edit: cleaned up my code a little and included it:

<?php
//$xml = simplexml_load_file('data.xml');
$xml = simplexml_load_string('<items>
  <item>
    <images>
      <image>A</image>
      <image>B</image>
      <image>C</image>
    </images>
  </item>
</items>');

// Loop over items in original xml
foreach ($xml->item as $item) {
    if (count($item->images->image) > 1) {
        // Clone entry
        $images = clone $item->images;

        // Remove entry and replace with empty
        unset($item->images);
        $item->addChild('images');

        for ($i = 0; $i < count($images->image); $i++) {
            $num = ($i > 0) ? $i : '';
            $item->images->addChild('image'.$num, $images->image[$i]);
        }
    }
}

//$xml->asXML('POTO.xml');
echo $xml->asXML();

Looks like this doesn't work prior to version 5.2.1 (http://3v4l.org/m0VoD), but that's going back pretty far, really. Rasmus doesn't want anyone running less than 5.4 (http://www.youtube.com/watch?v=anr7DQnMMs0&t=10m27s), and I'm inclined to agree.

For the record, this behavior, of copying only the reference is not a quirk of PHP. It's a common practice. Further reading: http://en.wikipedia.org/wiki/Cloning_(programming)

Collectives™ on Stack Overflow

Restructure XML in PHP

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related