This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
The fn:nilled function does not test for the presence of the xsi:nil attribute set to true, but for elements with the nilled PSVI propery (i.e., elements with xsi:nil set to true and which are declared in the schema as nillable). This means non-validated elements will always either return () or false. Many of those tests currently have true as expected result. Those tests are: fn-nilled-5 fn-nilled-7 fn-nilled-9 fn-nilled-11 fn-nilled-12 fn-nilled-15 fn-nilled-16 fn-nilled-17 It would probably be preferable to have the tests for which the expected result is true to be migrated to the part of the test suite which cover the validation feature. Keeping some of the test using fn:nilled over non-validated elements with the xsi:nil attribute, and returning false would probably be another good idea. - Jerome
Jerome: Thanks. Changed expected outcome to "false". Will atempt to write a test that uses "validate lax" to find a "true" outcome. Please close bug if in agreement and when able to verify. Thanks, Carmelo
In the case of fn-nilled-11 the expected result should be true, because it tests for not(nilled()). The result appears to have been changed mechanically to "false". And as Jerome says, it would be useful to have some tests for fn:nilled in a schema-aware environment, as the current tests are really only testing a trivial case.
Very True. Changed the value to "true". I will attempt to add extra test for a schema aware environment. Thanks, Carmelo
The expected results for fn-nilled=11 in 0.9.4 still say "false".
Michael: No idea how this one slipped byy. Corrected results. Thanks, Carmelo