This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
When looking at the latest CVS version of the XQTS, I noticed some issues in the HOF cases: The most import one is that partial-apply() is used in several HOF cases, even though the latest public XQuery 3.0 says that is has been replaced by the use of '?'. Affected test cases include hof-027, hof-040 to -045, hof-912, hof-915 to hof-918 There are also some minor bugs: 1) Many test cases include an XML declaration, even though the result is not valid XML, e.g. hof-001 Expected: <?xml version="1.0" encoding="UTF-8"?>3, should be 3 2) hof-003 uses a location hint to import a module. This should be replaced or amended by a module specification in the catalog 3) hof-033, -034 and -035 expect 'uppercase' (as the function name), but fn:upppercase should also be accepted
Fixed. For item 3, the spec says the prefix of the function name returned by function-name is implementation dependent, so I changed the test to output the local name only.
(In reply to comment #1) > Fixed. Thanks for the extremely fast reaction! I noticed some remaining issues, probably oversights: - in hof-003, there is now a module reference: <module namespace="http://example.com/hof-003">module-hof-003</module> There is, however, no definition of that module in the "sources" section of the catalog. This also introduces an error message in the catalog tests: "cvc-id.1: There is no ID/IDREF binding for IDREF 'module-hof-003'." - partial-apply() is still present in hof-912, hof-915 to -918 > For item 3, the spec says the prefix of the function name returned by > function-name is implementation dependent, so I changed the test to output the > local name only. Looks like a good solution
Another amendment: It seems that my comments in #2 are now fixed in CVS. I encountered some additional issues when revisiting the tests: - What are the correct error codes for hof-915 and -916? The catalog lists FOFU0001, but this code is not defined or referenced in the XQuery 3.0 or F&O 3.0 drafts. - The expected results for hof-046 and -047 use indentation, likely to make them more readable for humans. According to my understanding of the guideline for XQTS, indentation should not be used, and the query does not produce this whitespace by itself.
Bug issue resolved in QT3