this is indeed an issue with the tests. I've fixed three test cases:
testGetPropertiesNamePatter(), testHasNode() and testHasProperty()
svn revision is: 291168
regards
marcel
Stefan Guggisberg wrote:
On 9/23/05, David Caruana <[EMAIL PROTECTED]> wrote:
What is the expected behaviour if Node.hasProperty is called with a
malformed path? I believe it can throw a RepositoryException, but the
spec isn't explicit.
i agree. passing a syntactically invalid relative path should result in a
RepositoryException (malformed path/name). this would be consistent
with the rest of the api. and the spec should be more explicit in this case.
If that is the case, there may be an issue with the TCK
NodeReadMethodsTest.testHasProperty(). It builds a non-existent
property name by cycling through the existing property names and
concatenating them. In our test data, we have namespaced property names
resulting in a path that contains many ':' seperated elements. Our
hasProperty() implementation is throwing an invalid name exception and
thus the test fails.
Is that an issue with the test or should we return false in the case of
malformed paths?
i guess this is an issue with the test and should be fixed.
cheers
stefan
Regards,
David Caruana
www.alfresco.org
