[rfc][icedtea-web]encoding test

[Pavel Tisnovsky]

Hi Jiri,

these reproducers seems to be correct. To be honest, it would be quite
difficult to try to "guess" the right encoding when somebody create
jnlp file using encoding different from the XML header (but Window
users with theirs Notepad ... ugh ...).

Ok to commit, but please make sure that Mercurial handle the files
with strange encoding in the right way - diff should be run in binary
mode to check for differences etc etc.

Cheers,
Pavel



----- Original Message -----
> ping?
> 
> 
> updated to latest state
> 
> 2012-04-18 Jiri Vanek <jvanek at redhat.com>
> 
> 	added encodings reproducer (PR1108)
> 	* tests/reproducers/simple/encodingTestsĚŠČŘŽÝÁÍÉĚÉŘŤÝÚŮÍÓÁŠĎŽŹŇ:
> 	new reproducer
> 	*
> 	tests/reproducers/simple/encodingTestsĚŠČŘŽÝÁÍÉĚÉŘŤÝÚŮÍÓÁŠĎŽŹŇ/srcs/encodingTest.java:
> 	main class/main applet class and method of new
> 	encodingTestsĚŠČŘŽÝÁÍÉĚÉŘŤÝÚŮÍÓÁŠĎŽŹŇ.jar
> 	*
> 	tests/reproducers/simple/encodingTestsĚŠČŘŽÝÁÍÉĚÉŘŤÝÚŮÍÓÁŠĎŽŹŇ/testcases/encodingTestTest.java:
> 	testcases for jnlps of this reproducers
> 	*
> 	tests/reproducers/simple/encodingTestsĚŠČŘŽÝÁÍÉĚÉŘŤÝÚŮÍÓÁŠĎŽŹŇ/resources:
> 	four utf-8 and four iso-8859-2 jnlp files and one utf-8 and one
> 	iso-8859-2 html files
> 
> On 05/03/2012 03:45 PM, Jiri Vanek wrote:
> > ping?
> >
> > -------- Original Message --------
> > Subject: [rfc][icedtea-web]encoding test
> > Date: Wed, 18 Apr 2012 11:12:03 +0200
> > From: Jiri Vanek <jvanek at redhat.com>
> > To: IcedTea Distro List <distro-pkg-dev at openjdk.java.net>
> >
> > On 04/17/2012 03:55 PM, Jiri Vanek wrote:
> >> Hi!
> >>
> >> This is reproducer for incorrect behaviour of icedtea-web when
> >> jnlps are in "unexpected" encoding.
> >> We are expecting utf-8 everywhere, but windows users can very
> >> easily expose jnlps in different one.
> >> And we should process them for unix user correctly.
> >> I have used ISO-8859-2 encoding to torture icedtea-web.
> >> All tests to this encoding have failed. I have added also set of
> >> UTF-8 tests above those cases, and
> >> all those are passing (so test should be correct).
> >>
> >> Sorry that diff is not in classical patch format, but diff tools
> >> have not cooperate well with mixed
> >> utf8/non-utf8 content. And it is just adding files so "applying"
> >> should be simple as with patch.
> >>
> >> Please note, that this adds 3 more passing and three more failing
> >> tests - which are representing the
> >> issue. Maybe annotation marking this "representatives" will become
> >> necessary soon?
> >>
> >> The name of the directory is so weird because I wanted to create
> >> resource containing those unhappy
> >> characters. (The engine creates nameOfDirectory.jar)
> >>
> 
>