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 test uses a strings with spaces and newlines as the value for an xml:id attribute, assuming that the normalization will remove them. But the specs say 3.7.1.1 Attributes ... 5) If the attribute name is xml:id, the string value and typed value of the attribute are further normalized by discarding any leading and trailing space (#x20) characters, and by replacing sequences of space (#x20) characters by a single space (#x20) character. But "sequence of space are replaced by a single space", so at the end the xml:id attribute is assigned the string "ab c d" that is an invalid QName and causes the error XQST0082
The XML Query WG considered this during our Last Call WD and made an explicit decision not to require checking the value of xml:id attribute being constructed. Please see: http://www.w3.org/Bugs/Public/show_bug.cgi?id=1570 As a result, the rule that you cite constructs the value by replacing a sequence of spaces with a single space and then ends. I believe that this test case is not in error. Please close this bug of you agree with this resolution.
Hi Andrew, you are right; I missed that the NCName cosntraint had been removed. Thanks, Alberto