SimpleXMLElement and encoding

PeterPeiGuo · March 22nd, 2010, 10:13 AM

I used the following code to parse RSS documents:

Code:

		$page = file_get_contents($rss);
		$feed = new SimpleXMLElement($page);

But it failed to parse some BIG5 encoding RSS and gave this warning:

Code:

Warning: SimpleXMLElement::__construct(): input conversion failed due to input error

here is one example document:
http://rss.chinatimes.com/rss/latestnews.rss

The issue has some inconsistency. Obviously, being an RSS document, the content of the above link changes over time. SimpleXMLElement can parse some instances with no issue at all, but not all instances.

PeterPeiGuo · March 22nd, 2010, 11:55 PM

Apparently PHP has more limited encoding support comparing to some other languages. Other than the BIG5 encoded document that I mentioned in my original post, SimpleXMLElement also had issues with a gb2312 encoded document. However I found a workaround for that gb2312 encoded document, by simply do the following pre-process, before I pass the document into SimpleXMLElement:

Code:

$page = str_replace('encoding="gb2312"', 'encoding="GBK"', $page);