This is now been published on the MSDN XML developer center as I mentioned in an earlier post. A very useful component. The main points to remember about its usage are;
- It is a streaming XmlReader implementation. That is why the XPath subset is limited. There is no intelligent caching or bookmarking to walk back up the tree of nodes.
- It abstracts away details of the XmlReader API by using an XPath expression, hence this is very much a usability driven component.
- Combining this with an XmlWriter makes for efficient, streaming XML transformations without having to resort to XSLT in many simpler transformation scenarios.