Cart
[Home] [By Thread] [By Date] [Recent Entries]
At 10:28 AM 5/9/2003 -0400, Mike Champion wrote:
>On Fri, 09 May 2003 09:47:15 -0400, Jonathan Robie
><jonathan.robie@d...> wrote:
>>Let's be clear: the PSVI is not in the XQuery Data Model. Nobody wanted
>>the PSVI in the data model, there's a lot of stuff in the PSVI that we
>>don't need, and it's defined at the wrong level.
>
>You will perhaps perhaps forgive people who read section 3.3:
>
>"Some aspects of the data model are dependent upon XML Schema validity
>assessment; this document describes how to determine those aspects of the
>data model from a Post Schema Validation Infoset. [Definition: A Post
>Schema Validation Infoset, or PSVI, is the augmented infoset produced by
>an XML Schema validation episode.]."
>
>and remain unclear on this point :-) You're right of course that there
>are many aspects of PSVI that are not there, but "PSVI" is a shorthand for
>"validation time augmentation of nodes with XSDL type information" to many
>people.
Right - some aspects of the data model depend on XML Schema validity
assessment - this refers largely to the type names of W3C XML Schema simple
and complex types. Here's the crucial principle:
>If a PSVI is not available, then the data model is constructed from the
>Infoset in a manner that is compatible with the expectations of
>well-formed or DTD-validated parsing of an XML document.
I think that's important. Look at 3.6 for an example showing how that works
out:
If the [validity] property exists and is "valid":
( This says how to extract the type names from the PSVI of a schema )
If the [validity] property does not exist on this node or any of its
ancestors, Infoset-only processing is applied:
( This says how to maintain the types ID, IDREF, IDREFS, ENTITY,
ENTITIES, NMTOKEN, or NMTOKENS
for DTD-validated Infosets )
Otherwise, xs:anyType for elements or xs:anySimpleType for attributes.
( In other words, the stuff is untyped. )
One of these three branches tells how to create the information from a PSVI
- and you can only do that if there is a PSVI. There are a bunch of
specific mappings for nils, specific types etc. If you aren't working from
a PSVI, they aren't there.
>And of course the Right Thing is to read the freakin' specs and make
>specific comments to the appropriate w3c comments list!
Absolutely!
Jonathan
|
