Question

I’m working with a RESTful WCF service. One of the service methods returns byte[] (which contains a file).

On the client side, I have some code that uses the WebRequest class to invoke that particular service method. Using the WebRequest, I’m getting the response stream like:

Stream stream = webReq.GetResponse().GetResponseStream();

From this stream, I am then reconstructing a byte[] and then outputting the File locally. The problem is that the reconstructed file on the client-side doesn’t resemble the file that was returned from the service side (I get a corrupt PDF file that is much larger in size than the one sent from the service side). Just before the service method returns the byte[], I outputted that byte[] to disk on the service side and it created the file fine… This points to something going wrong between that method returning the byte[] and my client side reconstructing the byte[] from a Stream on the client side… to reconstruct the byte[] from the Stream, I’m using the following method which someone posted in the past on stackoverflow:

public static byte[] ReadFully(Stream input) 
{     
    byte[] buffer = new byte[16*1024];     
    using (MemoryStream ms = new MemoryStream())     
    {         
        int read;         
        while ((read = input.Read(buffer, 0, buffer.Length)) > 0)         
        {
            ms.Write(buffer, 0, read);         
        }         
        return ms.ToArray();     
     } 
}

Any ideas what could be going wrong?

Answer 1

I guess that the response from the server contains some envelope in addition to the raw bytes. Like a XML envelope or something. Which would of course suppose that the bytes are base64 encoded string in the response because you cannot store binary data into XML. It would also explain why you are getting a bigger buffer on the client than the actual PDF that the server has sent.

This will of course depend on what binding your WCF service uses and how is it configured. When you dump the contents of the MemoryStream you are reading on the client what exactly do you see? This should give you further hints on how the actual PDF file is encoded in the HTTP response body.