(Didn't realize that I was cc'ed.)
We do exactly what the spec requires. (If not, either there is a bug, or we
have very good reasons. :-) )
- For simple types, the spec says that the facets are inherited from the
base type, so we "collect" them into the derived type.
- For complex types,
- For restriction, the spec says that the {content type} is what's in the
derived type, so we don't include anything from the base type into the
{content type} of the derived type.
- For extension, the specs says the {content type} is a sequence
containing things from both the base type and the derived type, so we have
that sequence in the derived type.
So getDefinedFacets on XSSimpleTypeDefinition returns all the facets
(including those inherited from base types). And getParticle on
XSComplexTypeDefinition gives you the particle as required by the spec.
Note that getContentModel() is used to get another implementation class
that facilitates the schema assessment, and is not intended to be used by
the users (e.g. you :-) ).
HTH,
Sandy Gao
Software Developer, IBM Canada
(1-905) 413-3255
[EMAIL PROTECTED]
"Bob Foster"
<[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]>
cc: Sandy
Gao/Toronto/[EMAIL PROTECTED]
07/26/2003 02:10 Subject: Re: examining XML
schema attributes
AM
Please respond to
xerces-j-user
> "negativeInteger" is *not* a primitive type. It's derived from "integer",
> which is in turn derived from "decimal". And "decimal" is a primitive.
But
> you are right in that, to determine whether a type is derived from
> negativeInteger, one needs to call getBaseType recursively.
> getPrimitiveType won't help here, because it would return "decimal".
If one wants to build up the set of all facets defined in user-derived
types, do you have to recursively trek through the base types or are they
collected into the attribute type?
An earlier response having to do with complex types led me to think they
might be (collected), so perhaps I should verify that answer, as well. Does
the content model/particle of a complex type incorporate the net of all the
base complex types, or does one have to recursively construct the net
model?
Or is this the wrong question? I see the validator casts the
XSComplexTypeDeclaration to an XSComplexTypeDecl in order to call
getContentModel(). Is it the content model that is aggregated, while the
particle(s) only reflect the current type?
Sorry to belabor these points, but even though "XSSimpleTypeDefinition is
preferred over XSSimpleTypeDecl" it is really not obvious when one must use
an "implementation type" in order to avoid re-doing work the parser has
already done.
Thanks.
Bob Foster
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
