Re[2]: [lmi] Calculation summary XML resources structure (with some examples)

[Vadim Zeitlin]

On Mon, 25 Sep 2006 17:14:16 +0000 Greg Chicares <address@hidden> wrote:

GC> I am not opposed to distinguishing numeric and string data. Maybe
GC> it would be less useful to maintain that distinction if we choose
GC> to do all number formatting in C++, but I wouldn't oppose keeping
GC> the types distinct even in that case if you still think it's a
GC> good idea.

 I think it's a useful idea as long as we want to do type checking. If we
don't care about this, then having a single name would be easier and more
readable, but I thought that verifying that the values of double_scalars
were really doubles could help us catch some errors.

GC> > Say we want to reference a 17th of 'AcctVal' vector. The corresponding
GC> > XPath expression would be:
GC> > If we use the current XML structure with "number" attribute:
GC> > /illustration/address@hidden'AcctVal']/address@hidden
GC> > For the proposed XML structure:
GC> > /illustration/address@hidden'AcctVal']/duration[position()=17]
GC> 
GC> The second expression uses only three more characters, but lets
GC> us save many redundant characters in the data file.

 Which makes it more readable and allows to edit it easier because you
don't have to manually update number attribute values after e.g. moving
them around.

GC> Yes, I agree: 'position()' is a win.

 I'm glad we agree about this.

 Thanks,
VZ