RE: MS-FO-LC1-100: Promotion of untyped values in fn:distinct-val ues?

I think the rules for fn:distinct-values should be:

* untypedAtomic values are cast to string
* non-comparable values are allowed, and treated as distinct

Michael Kay

> -----Original Message-----
> From: Michael Rys [mailto:mrys@microsoft.com] 
> Sent: 14 August 2003 20:06
> To: Ashok Malhotra; public-qt-comments@w3.org
> Subject: RE: MS-FO-LC1-100: Promotion of untyped values in 
> fn:distinct-values?
> 
> 
> 
> This seems incorrect. We should not change the type of the 
> values in the result. We should only use a certain type for 
> the comparison. Also, with the wording below, what would the 
> semantics be, if you have the static type
> 
> (Xdt:untypedAtomic | xs:string | xs:decimal)*
> 
> and a sequence of values of each of the member types?
> 
> Best regards
> Michael
> 
> > -----Original Message-----
> > From: Ashok Malhotra
> > Sent: Sunday, August 10, 2003 8:30 AM
> > To: Michael Rys; public-qt-comments@w3.org
> > Subject: RE: MS-FO-LC1-100: Promotion of untyped values in
> fn:distinct-
> > values?
> > 
> > fn:distinct-values does not say how values of type xdt:untypedAtomic
> are
> > handled.  I suggest we use the same rule that we use for aggregate 
> > functions, namely  "Values of type xdt:untypedAtomic are cast to the
> type
> > of the other items in $arg. If all values in $arg are of type 
> > xdt:untypedAtomic they are converted to xs:string."
> > 
> > All the best, Ashok
> > 
> > > -----Original Message-----
> > > From: public-qt-comments-request@w3.org 
> [mailto:public-qt-comments- 
> > > request@w3.org] On Behalf Of 
> Michael Rys
> > > Sent: Wednesday, July 09, 2003 2:33 AM
> > > To: public-qt-comments@w3.org
> > > Subject: MS-FO-LC1-100: Promotion of untyped values in 
> fn:distinct-
> > values?
> > >
> > >
> > > Class: Technical
> > >
> > > Section 15.1.11: Are values of type xdt:untypedAtomic promoted? If
> so
> > > how?
> > >
> 
> 
>