1802 2005-07-22 03:44:08 +0000 [FS] editorial: E.1.3 Attribute filtering 2007-01-16 17:47:44 +0000 1 1 1 Unclassified XPath / XQuery / XSLT Formal Semantics 1.0 Last Call drafts All All CLOSED FIXED P2 minor --- 1 jmdyck simeon public-qt-comments oldest_to_newest 5127 0 jmdyck 2005-07-22 03:44:08 +0000 E.1.3 Attribute filtering Notation "Value filter @QName => ()" "Value filter @QName => SimpleValue" But SimpleValue derives (), so in a judgment where the RHS value is (), it could be an instance of either form. And they appear to mean slightly different things. "Value filter @QName => () or SimpleValue" This seems fairly ad hoc. Instead of: Value1 filter @xsi:nil => () or false why not use existing notation and say: Value1 filter @xsi:nil => SimpleValue SimpleValue in { (), false } Sem / rule (1|2) / premise 1 "dynEnv |- Value1 of attribute:: => Value2" What kind of judgment is this? Maybe you mean: dynEnv |- axis attribute:: of Value1 => Value2 (leftover from last year, comment #236) Sem / rule (1|2) / premise 2 "dynEnv |- Value2 of attribute, QName => ..." Again, what kind of judgment is it? I don't even have a guess. (leftover from last year, comment #237) 8908 1 simeon 2006-03-27 21:31:03 +0000 Applied the following fixes: (1) distinguished when the attribute content is empty: Value filter @QName => () from then the attribute is absent: Value filter @QName => absent (2) As suggested, used the standard notation when multiple values are possible for a judgment, as in: Value1 filter @xsi:nil => SimpleValue SimpleValue in { (), false } (3) Fixed the definition for the filter judgment to use the axis and nodetest judgment properly. - Jerome