Well, here’s how I would do it:
Your tpl =
<content>
<text1>[~[+id+]~]</text1>
<text2>[[cdata? &txt=`[+pagetitle+]`]]</text2>
<image />
<link>http://www.google.com</link>
<target>_blank</target>
</content>
Why do I use that cdata stuff for the text2 node? To avoid character problems when parsing xml (stuff like ampersands etc.).
The snippet cdata looks like this:
if($txt) {
$output = '<![CDATA[' . $txt . ']]>';
return($output);
}
In my document, I simply use:
<docs>
[[Ditto? &startID=`58` &tpl=`DittoTemplate` &total=`20` &display=`20`]]
</docs>
It’s good practise to enclose the nodes in one outer node, imo. Hence, the <docs></docs> stuff.
Note: In the document edit-window, tab #2, choose XML as document-type. Also, don’t use a template (=blank).