SimpleXML Reading node with a hyphenated name

Your assumption is correct. Use

$officeXML->{'document-meta'}

to make it work.

Please note that the above applies to Element nodes. Attribute nodes (those within the @attributes property when dumping the SimpleXmlElement) do not require any special syntax to be accessed when hyphenated. They are regularly accessible via array notation, e.g.

$xml = <<< XML
<root>
    <hyphenated-element hyphenated-attribute="bar">foo</hyphenated-element>
</root>
XML;
$root = new SimpleXMLElement($xml);
echo $root->{'hyphenated-element'}; // prints "foo"
echo $root->{'hyphenated-element'}['hyphenated-attribute']; // prints "bar"

See the SimpleXml Basics in the Manual for further examples.

The elements you're asking about are not part of the default namespace but in a different one. You can see that because they have a prefix in their name separated by the colon:

  ...
  <channel>
    <title>PhotoSpecialist.de</title>
    <!-- title is in the default namespace, no colon in the name -->
    ...
    <g:price>199.00 EUR</g:price>
    ...
    <g:gtin>5022361100576</g:gtin>
    <!-- price and gtin are in a different namespace, colon in the name and prefixed by "g" -->
  ...

The namespace is given with a prefix, here "g" in your case. And the prefix the namespace stands for is defined in the document element here:

<rss xmlns:g="http://base.google.com/ns/1.0" version="2.0">

So the namespace is "http://base.google.com/ns/1.0".

When you access the child-elements by their name with the SimpleXMLElement as you currently do:

$a = $node->title;
$b = $node->price;
$c = $node->gtin;

you're looking only in the default namespace. So only the first element actually contains text, the other two are created on-thy-fly and are yet empty.

To access the namespaced child-elements you need to tell the SimpleXMLElement explicitly with the children() method. It creates a new SimpleXMLElement with all the children in that namespace instead of the default one:

$google = $node->children("http://base.google.com/ns/1.0");

$a = $node->title;
$b = $google->price;
$c = $google->gtin;

So much for the isolated example (yes, that's it already).

A full example then could look like (including node-expansion on the reader, the code you had was a bit rusty):

<?php
/**
 * How to parse an XML node with a colon tag using PHP
 *
 * @link http://.com/q/29876898/367456
 */
const HTTP_BASE_GOOGLE_COM_NS_1_0 = "http://base.google.com/ns/1.0";

$url = 'https://my.datafeedwatch.com/static/files/1248/8222ebd3847fbfdc119abc9ba9d562b2cdb95818.xml';

$reader = new XMLReader;
$reader->open($url);

$doc = new DOMDocument;

// move to first item element
while (($valid = $reader->read()) && $reader->name !== 'item') ;

while ($valid) {
    $default    = simplexml_import_dom($reader->expand($doc));
    $googleBase = $default->children(HTTP_BASE_GOOGLE_COM_NS_1_0);
    printf(
        "%s - %s - %s<br />n"
        , htmlspecialchars($default->title)
        , htmlspecialchars($googleBase->price)
        , htmlspecialchars($googleBase->gtin)
    );

    // move to next item element
    $valid = $reader->next('item');
};

I hope this both gives an explanation and broadens the view a little on XMLReader use as well.

I had the same problem and I'm using Ubuntu 15.10.

In my case, to solve this issue, I installed the package php7.0-xml using the Synaptic package manager, which include SimpleXml. So, after restart my Apache server, my problem was solved. This package came in the Debian version and you can find it here: https://packages.debian.org/sid/php7.0-xml.