How do I create a XmlTextReader that ignores Namespaces and does not check characters

Question

I want to use a XmlTextReader that ignores Namespaces and does not check characters. To ignore namespaces I can set the property Namespaces=false a

Answer 1

In my case, as part of a provider's api versioning strategy, I came accross the need of supporting different response namespaces while deserialising with XmlSerializer to the same C# class.

The responses are basically the same, but the root namespace in the response is different depending on the credentials passed to the service.

So, I have used XmlWrappingReader approach to rename the namespace returned by the service to the namespace that the C# classes have attached through the XmlTypeAttribute.

But just one caveat: use string intering as I suspect the XmlSerializer is using Object.ReferenceEquals while reading from the XmlReader to match against the target namespace in the c# class.

In my case, the replacementNs parameter was obtained from the type's XmlTypeAttribute through reflection and apparently was not the exact same object reference as the string internally used by the XmlSerializer.

I almost lost my hair with this silly thing…

public class ReplaceNsXmlReader : XmlWrappingReader
{
    private readonly string replacementNs;

    public ReplaceNsXmlReader(XmlReader reader, string replacementNs)
        : base(reader)
    {
        //
        // NOTE: String.Intern is here needed for the XmlSerializer
        // that will be using this reader to deserialize correctly
        //

        this.replacementNs = String.Intern(replacementNs);
    }

    public override string NamespaceURI
    {
        get => replacementNs;
    }
}