How to find nodes entirely between two specified nodes

In a XML document such as the following:

<root>
  <fish value="Start"/>
  <pointlessContainer>
    <anotherNode value="Pick me!"/>
    <anotherNode value="Pick me too!"/>
    <fish value="End"/>
  </pointlessContainer>
</root>

How can I use the wonder of LINQ to XML to find any nodes completely contained by the fish nodes? Note that in this example, I have deliberately placed the fish nodes at different levels in the document, as I anticipat开发者_如何学编程e this scenario will occur in the wild.

Obviously, in this example, I would be looking to get the two anotherNode nodes, but not the pointlessContainer node.

NB: the two 'delimiting' nodes may have the same type (e.g. fish) as other non-delimiting nodes in the document, but they would have unique attributes and therefore be easy to identify.

For your sample, the following should do

        XDocument doc = XDocument.Load(@"..\..\XMLFile2.xml");
        XElement start = doc.Descendants("fish").First(f => f.Attribute("value").Value == "Start");
        XElement end = doc.Descendants("fish").First(f => f.Attribute("value").Value == "End");
        foreach (XElement el in 
            doc
            .Descendants()
            .Where(d => 
                XNode.CompareDocumentOrder(d, end) == -1 
                && XNode.CompareDocumentOrder(d, start) == 1 
                && !end.Ancestors().Contains(d)))
        {
            Console.WriteLine(el);
        }

But I haven't tested or thoroughly pondered whether it works for other cases. Maybe you can check for some of your sample data and report back whether it works.