> Thanks for the other fixes. The basic issue with this comma problem > (and the whitespace mentioned elsewhere), is these oddities make > extraction from the use cases for testing much more difficult. I > understand the comma is not text. It seems it should still be there > because the output is still a sequence; it is not inside another > element which would make the elements be children. It should be pointed out that there is no normative serialization of the results of a query. In cases where the result of the query is a single document node or element node (which is true of most of the use cases), results can be compared by serializing and then converting to a canonical XML form. But if the result is say a sequence of integers, there is no definitive way of serializing it that all products will adhere to. Michael Kay
Received on Tuesday, 28 October 2003 17:26:17 UTC