PHP library for parsing XML with a colons in tag names?

Question

I've been trying to use SimpleXML, but it doesn't seem to like XML that looks like this:

<xhtml:div>sample <xhtml:em>italic</xhtml:em> text</xhtml:div>

So what library will handle tags that look like that (have a colon in them)?

Please post the relevant code snippet. SimpleXML shouldn't have any problems with namespaces. — Josh Davis, Oct 16 '09 at 01:58
I've also experienced the same issue with `SimpleXMLElement($data)` returning null if the xml data has colons in the tags. — Jake Wilson, Oct 12 '12 at 19:11

Nathan · Accepted Answer · 2014-11-24T02:55:02.287

88

Say you have some xml like this.

<xhtml:div>
  <xhtml:em>italic</xhtml:em>
  <date>2010-02-01 06:00</date>
</xhtml:div>

You can access 'em' like this: $xml->children('xhtml', true)->div->em;

however, if you want the date field, this: $xml->children('xhtml', true)->div->date; wont work, because you are stuck in the xhtml namespace.

you must execute 'children' again to get back to the default namespace:

$xml->children('xhtml', true)->div->children()->date;

edited Nov 24 '14 at 02:55

answered Apr 23 '10 at 06:50

Nathan

11,938
12
55
62

1

not sure why this isn't the selected answer. But for anyone in the future this is the one that solved my questions/problem! :) – daveomcd May 18 '11 at 16:22
What about multiple tags with ':'? This is how I were trying to access it: `$array['body']['Order']['Extensions']['data:AdditionalReferences']['data:YourRef']` – user2924019 Mar 10 '20 at 17:45

score 28 · Answer 2 · answered Aug 28 '11 at 01:07

28

If you want to fix it quickly do this (I do when I feel lazy):

// Will replace : in tags and attributes names with _ allowing easy access
$xml = preg_replace('~(</?|\s)([a-z0-9_]+):~is', '$1$2_', $xml);

This will convert <xhtml: to <xhtml_ and </xhtml: to </xhtml_. Kind of hacky and can fail if CDATA NameSpaced XML container blocks are involved or UNICODE tag names but I'd say you are usually safe using it (hasn't failed me yet).

answered Aug 28 '11 at 01:07

EarnestoDev

507
4
7

Works but it is a bit hacky.. screwed the contents of my `` tags and changed time – supersan Jun 22 '17 at 04:47

score 6 · Answer 3 · answered Oct 16 '09 at 00:10

6

Colon denotes an XML namespace. The DOM has good support for namespaces.

answered Oct 16 '09 at 00:10

Ollie Saunders

7,787
3
29
37

SimpleXML probably does too, but the OP is looking for a tag "xhtml:div" instead of just "div". – Matthew Scharley Oct 16 '09 at 00:15
SimpleXML has *some* stuff for dealing with it, but I still can't get it to work right. – mpen Oct 16 '09 at 01:44

score 2 · Answer 4 · answered Feb 22 '13 at 14:47

2

I don't think it's a good idea to get rid of the colon or to replace it with something else as some people suggested. You can easily access elements that have a namespace prefix. You can either pass the URL that identifies the namespace as an argument to the children() method or pass the namespace prefix and "true" to the children() method. The second approach requires PHP 5.2 and up.

SimpleXMLElement::children

answered Feb 22 '13 at 14:47

Patryk K

57
3

Isn't this exactly what Nathan Reed suggested in the answer I accepted? I agree that regex-fu is a dirty hack, but having to go through the `children()` selector isn't very fun either. – mpen Feb 22 '13 at 17:06
Yes, the same. I just wanted to indicate the you can also pass the URL that identifies the namespace to the children() method which works with PHP 5 and up. IMHO, there is no need to do a dirty hack, when there is a core method available. – Patryk K Feb 22 '13 at 17:22

PHP library for parsing XML with a colons in tag names?

4 Answers4

Linked

Related