This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
- well-formed-feature-2: This test case should also allow XQST0123 as result, because "qname" is no valid feature. As an alternative, the feature string could be reduced to "not:a:qname" - well-formed-feature-4: I would have expected XPST0081 as error - require-prohibit-1: XQST0120 should also be accepted here - require-prohibit-3: The current specification says that the "Serialization Feature does not have a feature name, and cannot be required or prohibited in a module.", so this test should probably return XQST0123 instead. - require-prohibit-5: XQST0120 should also be accepted here - require-all-optional-features-5-s: I would suggest to also add XQST0128 as correct result, because the spec says that a) "it is a static error [err:XQST0128] if a feature name that an implementation supports appears in a prohibit-feature option declaration and the implementation is unable to disable the feature" and b) "all-optional-features corresponds to the set of all named features that correspond to features listed in 5.2 Optional Features"
Bugs fixed and committed to cvs. Please see below for comments on the test cases. (In reply to comment #0) > - well-formed-feature-2: > This test case should also allow XQST0123 as result, because "qname" is no > valid feature. As an alternative, the feature string could be reduced to > "not:a:qname" Fixed > > - well-formed-feature-4: > I would have expected XPST0081 as error Fixed already by Tim some time ago > > - require-prohibit-1: > XQST0120 should also be accepted here I am rejecting this proposed change. I would think that during the static checking phase duplicates and clashing features would be caught. If you disagree please feel free to re-open the bug. > > - require-prohibit-3: > The current specification says that the "Serialization Feature does > not have a feature name, and cannot be required or prohibited in a > module.", > so this test should probably return XQST0123 instead. Test has been removed > > - require-prohibit-5: > XQST0120 should also be accepted here I am rejecting this proposed change. I would think that during the static checking phase duplicates and clashing features would be caught. If you disagree please feel free to re-open the bug. > > - require-all-optional-features-5-s: > I would suggest to also add XQST0128 as correct result, because the > spec says that a) "it is a static error [err:XQST0128] if a feature name > that an implementation supports appears in a prohibit-feature option > declaration and the implementation is unable to disable the feature" and > b) "all-optional-features corresponds to the set of all named features that > correspond to features listed in 5.2 Optional Features" This test looks to have been fixed already
Yes, I agree with your reasoning; thanks.
*** Bug 21839 has been marked as a duplicate of this bug. ***