Talk:WKHRDY

SNPWG 8 agreed

jens 19:48, 14 February 2009 (CET)

added a comment to clarify the multi working time option

raphael 20:12, 1 September 2009 (UTC): Propose changing cardinality of sub-attributes TIMSTW and TIMENW to 1..* to handle this.

raphael 20:24, 1 September 2009 (UTC): Also, addition of the following constraint to the specification for WKHRDY:

Other

Pair-wise correspondence between timstw and timenw is required. For each timstw/timenw pair, timstw must precede timenw.

And the following definition of "pair-wise" correspondence, in, say, the product specification:

Pair-wise correspondence

Let a1…am be the sequence of values of attribute A of an instance of object class O and b1...bn the sequence of values of attribute B of the same instance, each in the same order that the values occur in the data set. (Null values are permitted.) Attributes A and B have pair-wise correspondence when: (i) m = n; (ii) the encoding guide or product specification defines a relationship, or assigns special significance, for pairs (ai, bj) if and only if i = j. Informally, the attributes are pair-wise correspondent if and only each value is associated with its opposite number for the other attribute. For example, each value of attribute “day-of-the-week” may be associated with a value of attribute “office-hours”, signifying that the office in question is open during those hours on that day of the week.

This may seem complicated, but the S-57 encoding guide tries to achieve a similar goal with NATQUA/NATSUR in clause 7.1 of Appendix B.1. If list types (e.g., "1,2,4") are not allowed, the proposed solution or an alternative is needed.