Allow PDFDocumentBuilder to save PDF into a stream

[ghost]

For now you can not get PDF with a size more than 2 gigs. And it takes a lot of memory to obtain the raw bytes in the end.
I suggest instead of Build method to add a SaveTo method to save the results into a stream. Which can be a MemoryStream, a FileStream, etc ...

        public void SaveTo(Stream destination)
        {
            CompleteDocument();

            if (!context.Stream.CanSeek)
            {
                throw new InvalidOperationException("PdfDocument.Build() called with non-seekable stream.");
            }

            context.Stream.Seek(0, SeekOrigin.Begin);
            context.Stream.CopyTo(destination);
        }

The Build method is better to rename to something like ToByteArray, because it does not do anything except converting the result stream to bytes.

[cremor]

PdfDocumentBuilder offers a constructor overload that accepts a Stream parameter. There you can pass your target stream.
PdfDocumentBuilder.Dispose() will call CompleteDocument(). So just put your builder in a using statement and you will have a finished PDF after the using block in your target stream without having to build any byte array.

I agree that the API is not very nice since you need to know what PdfDocumentBuilder.Dispose() does, but at least it should work.