W3C home > Mailing lists > Public > public-qt-comments@w3.org > February 2004

ORA-XQ-233-C: Orderspec should specify that empty sequence is considered as the same type as other tuples

From: Stephen Buxton <Stephen.Buxton@oracle.com>
Date: 16 Feb 04 14:04:35
Message-Id: <200402162204.i1GM4aG12541@rgmgw5.us.oracle.com>
To: public-qt-comments@w3.org

SECTION 3.8.3: Order by and Return Clauses

In 3.8.3, it has a rule stated as:
"Each orderspec must return values of the same type for all 
tuples in the tuple stream".
Later, the orderspec considers the special values in the orderspec,
such as empty sequence. 
Then, empty sequence should be considered the same type as
for all other tuples in the tuple stream. 

Suggested rewording:
"For each position in the tuple, there must exist an atomic type for which the gt operator is defined, such that the value of the orderspec in that position is either the empty sequence or can be promoted to that atomic type".

Consider a tuple stream of 2 tuples, the first tuple is empty sequence and the second tuple is an atomic
value with type "xs:string" and typed value is "cat". From the spec,
it is not clear if this tuple stream is legal for order unless we
state that empty sequence is considered the same type as
'xs:string' in this case.

- Steve B.
Received on Monday, 16 February 2004 17:04:39 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:45:17 UTC