Re: [office-formula] Data types

[robert_weir@us.ibm.com]

Patrick Durusau <patrick@durusau.net> wrote on 01/19/2010 11:05:58 AM:


> Eike Rathke wrote:
> > Hi Patrick,
> >
> > On Monday, 2010-01-18 15:29:48 -0500, Patrick Durusau wrote:
> >
> > 
> >> For the agenda tomorrow, can the data types in chapter 3 be on the 
agenda?
> >>
> >> Question: For datatypes defined in XML Schema part 2, is the a reason 
we 
> >> can't use that as a reference?
> >> 
> >
> > As I mentioned in my previous mail in the reply to the document 
upload,
> > we (you and me and Michael and David, IIRC) discussed that some months
> > ago. There are differences between the XML datatypes and the types we
> > define. OpenFormula also does not define an XML model, I don't see
> > a reason why we should reference XML datatypes if those types are not
> > identical.
> >
> > 


Another example:  some of the character processing functions, the ones 
that convert characters from legacy codepages, can return Unicode 
characters that our outside of the character repertoire permitted by 
xsd:string.

Now you might think this is a horror, but it has legitimate uses, such as 
in a formula expression that converts a string, does an IF() test to see 
if the character is in a permitted range, and if not, then returns a 
permitted value.

For example:

IF([.A1]>9;CHAR([.A1]);" ")

Although CHAR([.A1]) may return a Unicode character that is not permitted 
in XML (not even via a numerical character entity), there is nothing that 
prevents it from being used in a calculation.  Thus, the runtime types of 
OpenFormula are broader than the XSD datatypes. 


-Rob