OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index]
RE: [ebxml-dev] ccts question

Michel,

<Michel>I am currently trying to implement the CCTs that have been
defined in the CCTS. I am trying to make these usable by our analysts
when they model their requirements. However, I have come accross a
problem. The CCTS has defined a type called NumericType. This type  is
supposed to hold "numeric information that is assigned or is determined
by calculation, counting or sequencing. It does not require a unit of
quentity or a unit of measure.". When an analyst is looking at a value
he is used to considering a value as an integer or a decimal value and
in certain circumstances he wants to ensure that the value is only
integer. He may also want to ensure that the value is unsigned.
My question is, how does one describe such a common type as integer or
positive integer which are in prinicple merely restrictions of the type
numericType?</Michel>

The Core Component Type "Numeric. Type" is of the primitive type
"Decimal". According to Table 7-1 for Decimals you can specify six
Content Component Restrictions or Facets: Total Digits, Fractional
Digits, Minimum Inclusive, Maximum Inclusive, Minimum Exclusive and
Maximum Exclusive. You can in addition use the Supplementary Component
"Numeric. Format. Text", which unfortunately is not further defined in
CCTS.

<Michel>Is it possible to create a specific CCT that inherits the same
properties as numericType but restricts it to integer values?</Michel>

Whenever you restrict values you should define a "Data Type", that
inherits the properties of the Core Component Type it is based on, but
with more restricted values (defined by means of Facets on Content and
Supplementary Components).

<Michel>I also have a socond problem with the CCTS  that concerns the
coherence with the naming convention. In looking at the CCTs and in
particular table 8.1. I find that the attributes for a given object
class do not all have the object class as the primary representation
term. For example Amount Currency.Code List Version. Identifier and
Amount Currency. Identifier are both a part of the object class Amount.
According to the CCTS they should have the object class term "Amount".
The sema is true for the object class "Code" where all its attributes
are associated with the object class "Code List" except for "Content"
and "Name. Text". Again we can see the same inconsistancy for the
"Identifier" "Quantity" and "Measure" object classes. There also appears
to be an inconsistancy with the use of the "Language" attribute as it
appears in several CCTs (Code, Text) but does not take the object class
to which it belongs. Are these inconsistancies normal? And if so, where
are rules that allow one to use attributes with different object classes
in a given class described?</Michel>

You are referring to Supplementary Components which, according to figure
7-1, are no Core Components and consequently do not need to conform to
the Core Component Naming Conventions (although I agree it would have
been less confusing and more consistent if they did).

Hope this helps.

Fred



<<attachment: winmail.dat>>



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index]