1231 2005-04-09 02:17:08 +0000 [DM] Pattern facet for ordered durations can and should be simplified 2007-04-04 19:07:17 +0000 1 1 1 Unclassified XPath / XQuery / XSLT Data Model 1.0 Last Call drafts PC Linux CLOSED FIXED http://www.w3.org/XML/Group/2004/06/xmlschema-2/datatypes.rq20.20050408.html P4 normal --- 1 cmsmcq Norman.Walsh public-qt-comments oldest_to_newest 3845 0 cmsmcq 2005-04-09 02:17:08 +0000 In appendix C (http://www.w3.org/TR/2005/WD-xpath-datamodel-20050404/#xdtschema), declarations are given of the types yearMonthDuration and dayTimeDuration. In each, a pattern facet is given which seeks (as far as I can tell) to provide a regular expression which defines the lexical space of the new types. Since all lexical forms for values in these types must ALSO be legal lexical forms for the parent duration type, it's not really necessary to recapitulate all the constraints of that type. The patterns may be simplified to "[^DT]*" for yearMonthDuration and "[^YM]*(T.*)?" for dayTimeDuration. (The lexical space defined by these is the same as that defined by the current declarations, if my work with Grail is reliable.) If the WGs prefer to use patterns which fully capture all the constraints on lexical forms, the bug in the dayTimeDuration facet which allows forms like "P2HS" and "P2H.S" should be fixed. 3738 1 cmsmcq 2005-04-12 23:22:37 +0000 A simpler pattern for dayTimeDuration is [^YM]*[DT].*" which can readily be glossed as: "You have to have a Day field or one or more of the Time fields, and they must not be preceded by any Year or Month field." It may be slightly easier to see that this is correct than that "[^YM]*(T.*)?" is correct. 3769 2 Norman.Walsh 2005-04-19 15:45:19 +0000 Andrew asked for time to review this issue at meeting 249 3282 3 Norman.Walsh 2005-05-17 14:21:09 +0000 Coordinate with MSM