StringWriter writes to strings, whose underlying memory representation is always UTF-16 in the CLR (let's not get into the surrogate pair issue here...).
Thus, if you have StringWriter write to a string, it *will* be UTF-16 whether the first line in the string claims so or not. This is also how MSXML behaves, by the way.
Of course, when you take the string and convert it into a memory buffer in another encoding you need to take care of converting the encoding-providing header, which means extra work to you. Whether this is actually a good design or not is another issue...
I would expect the fact that System.String is unicode has something to do with it :) So, it just won't support another format.
If you want another encoding you should probably use something along the lines of a binary stream.
An UTF8 encoded xml document stored in a UTF16 string doesn't really make much sence now does it :)
Just my 2c!
Thanks for all the input guys. Like I said, It seems to just reaffirm what I wrote - You can't make you XML document's header UTF-8 encoded without going through some hoops.
Perhaps this is something that would best be solved using a method of the XMLTextWriter, something like 'SetEncodingHeader(Encoding encoding)' that would allow this without too much hassle..
"I would expect the fact that System.String is unicode has something to do with it :)"
Technically, UTF-8 is also Unicode--just a different encoding for it ;).
I know that UTF-8 is unicode as well.
It's just that in the BCL UTF-16 is referred to as Encoding.Unicode... which is why I wrote that statement.
Anyways, the System.String class is an UTF16 string, so an XmlTextWriter with SetEncodingHeader would only enable you to have an encoding mismatch...
Morty :)
First of all thanks a lot for this piece of code, as its just what I was looking for.
Unfortunately I am able to get only 4096 bytes of data from the MemoryBuffer object.