[xml] Starting with 2.6.10 umlauts are converted to entities?

[Tobias Grimm <tobias grimm cas-soft de>]

Hi!

I ran some unit tests on my xmlreader wrapper today, which I haven't 
done for a while. Some tests break, because umlauts are converted to 
entities instead to proper utf8.

This is the xml:

<?xml version="1.0" encoding="iso-8859-1"?>
<root><first>äöüß</first></root>

And if I do i.e. this:

xmlNodeDump(buffer, node.doc, node, 0, 0)

I get:

buffer.content='<root><first>&#xE4;&#xF6;&#xFC;&#xDF;</first></root>'

All versions before 2.6.10 worked just fine an returned  a proper 
utf8-String.

What has changed in 2.6.10? I couldn't find anything useful in the 
changelog.

bye,

Tobias