How can EncoderFallbackExceptions caused by invalid utf16 surrogates be ignored when Console.Out.Encoding is UTF8?

Question

I've observed, using Visual studio 2008 (with .NET 3.5), that the the value of Console.Out.Encoding changes based upon if running via a debugger and Console App/Windows App in the following way:

• GUI App - using debugger
  • System.Text.SBCSCodePageEncoding
• GUI App - not using a debugger
  • System.Text.UTF8Encoding
• Console App - using a debugger
  • System.Text.SBCSCodePageEncoding
• Console App - not using a debugger
  • System.Text.SBCSCodePageEncoding

Now when Console.Out.Encoding is set to SBCSCodePageEncoding it doesn't throw exceptions when invalid utf16 surrogates written to it. For example:

string invalidStringContiaingHighOrderSurrogateWithOutMatchingLowOrderSurrogate = '\uD81B'.ToString() + ";";
Console.WriteLine(invalidStringContiaingHighOrderSurrogateWithOutMatchingLowOrderSurrogate);
Console.Out.Flush();

However when Console.Out.Encoding is set to System.Text.UTF8Encoding writing invalid utf16 surrogates to it throws System.Text.EncoderFallbackException.

I want to be a able to ignore EncoderFallbackExceptions without having to add a try catch block to each use of Console.Out in the large windows application I working with. How can I do this?

1. Changing to a console app is NOT an acceptable solution.
2. Somehow changing Console.Out.Encoding to SBCSCodePageEncoding IS a acceptable solution.
3. I can't change Console.Out.Encoding.EncoderFallback as I get the following exception System.InvalidOperationException "Instance is readonly"
4. The win32 function SetConsoleOutputCP seemed to have no effect on my GUI/Windows app.

Answer 1

Use the Console.OutputEncoding property to set the encoding however you want.

(Personally I'd stick with UTF-8, but create an UTF8Encoding, setting an appropriate EncoderFallback value, e.g. ReplacementFallback.)