Bertrand Delacretaz wrote:
On Jan 23, 2008 8:31 AM, Sami Siren <[EMAIL PROTECTED]> wrote:
...we should define the
encoding we use for our .java files with something like the following
and make sure our .java files are properly encoded...
We could do that, but it's IMHO safer to not use any non-ascii chars
in our source files.
UTFf-8 was just an example, if we prefer ascii we can define the
encoding as ascii. By doing this would allow tests to fail equally
everywhere in case someone (like me) accidentally puts something else
but ascii in string literals used in tests.
I fixed the problem in revision 614446, using this:
final String expected = "Archim\u00E8de et Lius \u00E0
Ch\u00E2teauneuf...
assertEquals(expected,metadata.get(Metadata.RIGHTS));
Cool!
Writing these escapes is a bit painful, but that should work
everywhere regardless of encoding.
You can also use native2ascii to convert those.
--
Sami Siren
