This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
This test does fn:filter( fn:string#1, () ) and expects the result (). I think it is legitimate to raise XPTY0004, on the grounds that the supplied function is the wrong type: it returns a string rather than a boolean. Under Errors and Optimization it is also legitimate to return ()
Hi Tim, this seems right to me, however would you be able to resolve this bug issue. Thanks.
Agreed. For the record, it is the text "If the result of invoking the new function would necessarily result in a type error, that error may be raised during function coercion. It is implementation dependent whether this happens or not." that permits XPTY0004.
Fixed in CVS. Please mark the bug as CLOSED if you agree with the resolution. Otherwise, REOPEN.