XmlDeserization with encoding

Question

I am de-serializing string to My classes. for that i have written a generic method

private T ExtractResponse<T>(string response) where T : class
        {
            T obj = null;               
            if (!String.IsNullOrEmpty(response))
            {                
                var serializer = new XmlSerializer(typeof(T));
                using (TextReader reader = new StringReader(response))
                {
                    obj = (T)serializer.Deserialize(reader);
                }
            }
            return obj;
        }

Previously i were just having ascii character so no Problem in deseriailizing. Now i am getting some non ascii characters in my responses like [2J2J] Due to that my xml is not deseriailizing into object. My xml document encoding is "ISO-8859-1" What change should i do in above logic method so that it should work correctly with my encoding.

Answer 1

I got a close answer to my question from my similar msdn forum post.Answer is

Your method will always use Unicode (UTF-16) as a String .NET is encoded in Unicode - and the StringReader has no possibility to pass any other encoding for that reason.

If you want to use different encodings your method should accept a MemoryStream or a byte array. The method that creates the response will have to rewritten that it writes the data into a MemoryStream with the correct encoding, and returns resulting bytes.

Then you can use a StreamReader and pass a different encoding, for example:

private T ExtractResponse<T>(byte[] response) where T : class
{
    T obj = null;
    if (response != null && response.Length > 0)
    {
        var serializer = new XmlSerializer(typeof(T));
        using (TextReader reader = new StreamReader(
            new MemoryStream(response), 
            System.Text.Encoding.GetEncoding("iso-8859-1")))
        {
            obj = (T)serializer.Deserialize(reader);
        }
    }
    return obj;
}

You could also add the encoding as an (optional) second parameter.

Note: For a Xml Document you can use a XmlReader for the XmlSerializer. That can detect the encoding from the document declaration.

Answer 2

Looks like you got your answer over at the MSDN Forums already. But in case someone else stumbles on this...

Specify the encoding you want in your StringReader, like so:

    using (TextReader reader = new StringReader(response, System.Text.Encoding.GetEncoding("iso-8859-1")))