Getting XPath for node with XmlReader

Question

How to get XPath for the current node with XMLReader?

E.g.:

<Employee>
    <Entity>
        <Id>1</Id>
    </Entity>
</Employee>

So I need to get XPath for 1 which is Employee/Entity/Id. Any ideas?

using (var reader = XmlReader.Create(basePath, settings))
{
    while (reader.Read())
    {                   
        if (reader.NodeType == XmlNodeType.Text)
        {
            // need to get xpath of the text node
        }
        else if (reader.NodeType == XmlNodeType.Element)
        {
            // need to get xpath of the current node
        }
     }
 }

Answer 1

My first suggestion would be to use a higher level API like LINQ to XML. The only reason to use a low level API like XmlReader is for extremely large files. With LINQ to XML, a naive implementation is fairly trivial:

var doc = XDocument.Parse(xml);

foreach (var element in doc.Descendants())
{
    var path = element.AncestorsAndSelf().Select(e => e.Name.LocalName).Reverse();
    var xPath = string.Join("/", path);
}

Using XmlReader is a bit more involved as you have to track the element path as you go:

using (var reader = XmlReader.Create(basePath, settings))
{
    var elements = new Stack<string>();

    while (reader.Read())
    {
        switch (reader.NodeType)
        {
            case XmlNodeType.Element:
                if(!reader.IsEmptyElement)
                    elements.Push(reader.LocalName);
                break;
            case XmlNodeType.EndElement:
                elements.Pop();
                break;
            case XmlNodeType.Text:
                path = string.Join("/", elements.Reverse());
                break;
        }
    }
}

Here's a working demo: https://dotnetfiddle.net/dpOzuL

Note that while this works for your trivial example, this is a very naive creation of an XPath expression and won't work in all cases (for example, when you have multiple siblings with the same name or when namespaces are involved).