Re: [xml] Validating Schemata



Hi,

Frans Englich wrote:
On Thursday 11 November 2004 15:49, Daniel Veillard wrote:
[...]
 all Schema for Schema published I know of are broken anyway I don't
think they ever fixed the bug I reported [1]
[...]
 Which version of the Schema ?
[...]
This one:
http://www.w3.org/2001/XMLSchema.xsd

The current Schema for Schemata cannot be constructed with a machanism
which would process a user created schema. The schema for schemata is
a special case schema, violating some of the rules for valid schemata.
One example is the derivation from 'anySimpleType', which is otherwise
forbidden [1]:

"The ·simple ur-type definition· must not be named as the ·base type
definition· of any user-defined atomic simple type definitions: as it
has no constraining facets, this would be incoherent."

[1] http://www.w3.org/TR/xmlschema-1/#Simple_Type_Definition_details

In a user written schema this won't work, since 'anySimpleType' has
a 'variety' of 'absent'; thus a derived type would have no 'variety'
as well, which is not allowed. Additionally the derivation
mechanism for facets won't work.


With the output:

bash-2.05b$ xmllint --schema /tmp/XMLSchema.xsd Severities.xsd --noout
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586
Unimplemented block at xmlschemas.c:5586

Yeah, identity-constraints - as noticed by Daniel - are not implemented
yet; but I'll get a laptop in a couple of days, so I can work on it in
the train ;-)

/tmp/XMLSchema.xsd:761: element element: Schemas parser error : Element 'element': A global element declaration with the name 'element' does already exist. /tmp/XMLSchema.xsd:974: element element: Schemas parser error : Element 'element': A global element declaration with the name 'group' does already exist.
WXS schema /tmp/XMLSchema.xsd failed to compile
bash-2.05b$

It's a bug. Sub-components of 'all' and 'choice' were put into the
top-level symbol space. Will commit the fix soon.

I'm upgrading to 2.6.16.

Won't help here. Addionally you'll get another bug, indicating the
'version' attribute to be not a valid 'token'. Will commit the fix soon.

Having fixed those bugs I get about 60 error messages, which say that
the types don't have a variety - see above explanation.
Additionally:
"XMLSchema.xsd:188: element complexType: Schemas parser error : CT local: The content model is not determinist."

At this point the process seems to hang. Don't know why.

Greetings,

Kasimier





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]