Re: Combining float/string to specify value/units

[Rick Jelliffe <ricko@allette.com.au>]

From: Steve Rosenberry <steve.rosenberry@verizon.net>

> Is it possible to derive an attribute type from a float primitive with
> an appended string indicating the physical units associated with the
> value

Try this (I think):

1) Derive an atomic simple-type, an enumeration of the physical units.
2) Derive an atomic simple-type, with the particular floating restrictions
you want.
3) Derive a union simple-type from both those.
4) Derive a list simple-type made up of that union type, of length 2.

This type should validate your data successfully. The cases that it won't
validate are the unlikely cases that someone has two numbers or two units or
the unit ahead of the number.  Note that there is no way to impose a
constraint on the value space of the number based on the unit, so you may
not be gaining much (except better error messages?)  compared to using a
regular expression.

Cheers
Rick Jelliffe