On Mon, Jun 2, 2008 at 8:14 AM, Ramkumar R <[EMAIL PROTECTED]> wrote:
> Hi Simon,
> After downloading the complete latest code from the repository, i noticed
> that the reason for the failure in CouldNotResolveLocation for
> implementation.resource and implementation.widget validation is due to the
> missed code while applying the patch.
>
> The changes suggested in the patch does not seem to appear in the committed
> code. For instance TUSCANY-2344 suggested a change in
> WidgetImplementationProcessor resolve method as shown below, which is
> required for the tests to be sucessfull.
>
> while (reader.hasNext()) {
> @@ -128,8 +149,11 @@
> } catch (IOException e) {
> ContributionResolveException ce = new
> ContributionResolveException(e);
> error("ContributionResolveException", resolver, ce);
> - throw ce;
> + //throw ce;
> }
> + } else {
> + error("CouldNotResolveLocation", resolver,
> implementation.getLocation());
> + //throw new ContributionResolveException("Could not resolve
> implementation.widget location: " + implementation.getLocation());
> }
> Not sure, if i should open a new JIRA OR reopen the older ones to apply the
> patch again. Please suggest.
>
> Also would be helpful if you could elobrate more about the conversion of
> tests to JUnit4. Thakns.
>
> On 5/29/08, Simon Laws <[EMAIL PROTECTED]> wrote:
> >
> > Hi
> >
> > FYI. I've seen a couple of problems with the widget and resource
> validation
> > testing during may latest build. CouldNotResolveLocation doesn't seem to
> be
> > raise. I've @Ignored these tests for now just in case it's going to
> affect
> > others (I changed the test to JUnit4 to make this easy) .
> >
> >
> > As an aside we should probably go through these tests and convert to
> Junit4
> >
> > Also I notice that the original tests I added don't fit into the neat
> > categorization scheme that has been used subsequently so I'll endeavor to
> > move the original tests into the new scheme to tidy things up.
> >
> > Simon
> >
>
>
>
> --
> Thanks & Regards,
> Ramkumar Ramalingam
>
Hi Ram
Can you identify which parts of the patch are missing and create a new patch
based on just these. As they didn't apply properly in the first place I
don't think that trying to apply the existing patch again will have the
desired effect.
Re. Junit4. Some of our tests in Tuscany use JUnit4 and some of them use
older versions of JUnit. As we are creating new tests here it would be
convenient to use the latest version of JUnit.
Regards
Simon
