5684 2008-05-13 11:17:11 +0000 [XQTS] some FunctX timezone test depend on context timezone 2008-06-26 14:09:59 +0000 1 1 1 Unclassified XML Query Test Suite XML Query Test Suite unspecified PC Windows NT RESOLVED FIXED P2 normal --- 1 oliver frans.englich public-qt-comments oldest_to_newest 20083 0 oliver 2008-05-13 11:17:11 +0000 The following FunctX tests only work when the static timezone is +02:00 (although the results on the FunctX website are given for an implicit timezone of -05:00) functx-fn-adjust-dateTime-to-timezone-6 functx-fn-adjust-dateTime-to-timezone-7 functx-fn-adjust-dateTime-to-timezone-all functx-fn-adjust-date-to-timezone-5 functx-fn-adjust-date-to-timezone-6 functx-fn-adjust-date-to-timezone-all functx-fn-adjust-time-to-timezone-5 functx-fn-adjust-time-to-timezone-6 functx-fn-adjust-time-to-timezone-7 functx-fn-adjust-time-to-timezone-8 functx-fn-adjust-time-to-timezone-all functx-fn-implicit-timezone-1 I can think of 4 possible solutions: 1) Rewrite the adjust tests to compare the timezone with implicit-timezone(), and drop the implicit-timezone-1 test; although this will be quite messy in some cases and loose the succinctness of the tests. 2) Specify that the entire test suite should be run in a particular timezone (eg. +02:00). 3) Add another element/attribute/role to the test-cases to specify the timezone in which a test must be run. This is my preferred solution. 4) Remove these tests entirely. 20121 1 andrew.eisenberg 2008-05-15 20:02:32 +0000 We ran into this issue long ago. My recollection is that agreed to use explicit timezones in most of our test cases, and to use Z as the expected value of the implicit timezone in just a small number of the test cases. Perhaps I recall wrongly, or perhaps we changed this over time. I went looking for precedent, and found that in fn-implicit-timezone-2 that we expect "-PT5H" and the comparator is Inspect. 20125 2 frans.englich 2008-05-16 07:37:50 +0000 Anyhow, it needs fixing. I didn't have the measures for reviewing each test individually -- things like this will crop, but I believe it will be relatively few issues. 20860 3 frans.englich 2008-06-25 16:35:20 +0000 I've tried to fix this in CVS. These files are generated and it's not the easiest thing to get an overview of so it's very likely I've missed something here. 20927 4 oliver 2008-06-26 13:50:48 +0000 These tests have all been removed/fixed correctly, with the exception of the expected result for functx-fn-adjust-time-to-timezone-all. 20930 5 frans.englich 2008-06-26 14:09:59 +0000 Yeah, updated version in CVS, I hope it's right.