On Tue, 10 Jan 2012 07:06:58 -0400, David Bremner <david@tethera.net> wrote:
> On Fri, 30 Dec 2011 09:39:37 +0000, David Edmondson <dme@dme.org> wrote:
> > ---
> > 
> > Added backslash test. UTF8 round-trip still isn't right.
> > 
> 
> Do you mind explaining a bit better in the commit message what the issue
> is here?

I'm not completely sure.

The expected output file is supposed to contain some UTF8 encoded
text. One of the addresses to be tested has a 'human readable' part that
has the same UTF8 text. Testing the functions directly within emacs
retains the text correctly encoded. Pushing the text through the test
suite results in text that does not correctly match the expected
output.

There are many places where things could be wrong:
      - the expected output file doesn't include the correct UTF8 text
        (but it looks fine in emacs),
      - the test wrapper script doesn't correctly pass around the UTF8
        encoded text,
      - the emacs reader is not correctly loading the UTF8 encoded text,
      - <your favourite imagined problem here>.