[
Lists Home |
Date Index |
Thread Index
]
Hi,
> This exchange came up on a different newsgroup. Anyone
> know what the correct answer is?
>
> > > > MSXML4 is complaining about the schema below, saying that it is
> > > > non-deterministic - i cant work out precisely why.
> > >
> > > > <xsd:sequence>
> > > > <xsd:element name="characteristic" type="chartype1"
> > > > maxOccurs="unbounded"/>
> > > > <xsd:element name="characteristic" type="chartype2"
> > > > maxOccurs="unbounded"/>
> > > > <xsd:element name="characteristic" type="chartype3" minOccurs="0"
> > > > maxOccurs="unbounded"/>
> > > > <xsd:element name="characteristic" type="chartype4"
> > > > maxOccurs="unbounded"/>
> > > > </xsd:sequence>
> > >
This content model is infact invalid. In XML Schema you're not allowed to declare
multiple elements with the same name and type in the same content model. This is
the reason for MSXML4's error message. If the different types can be derived from
a common base type you can solve this by using the xsi:type [1] attribute in the
instance document but otherwise you have to declare the elements with different
names.
Cheers,
/Eddie
[1] http://www.w3.org/TR/xmlschema-0/#UseDerivInInstDocs
>
> > > The problem is probably because of the minOccurs=0 on the third
> > > element. There is no way to know if the third element in such a
> > > sequence is of type 3 or 4 without doing a read-ahead of it's child
> > > elements. I don't think that Parsers read-ahead typically.
> >
> > But the minOccurs shouldnt be causing a problem. For starters, if we get rid
> > of the minOccurs, the parser still complains. Are you saying that a schema
> > enforcing the instance below would be non-deterministic because all the
> > elements in the sequence have the same name?
> >
> > <library>
> > <publication type="journal">...</publication>
> > <publication type="journal">...</publication>
> > <publication type="journal">...</publication>
> > <publication type="hardback"/>...</publication>
> > <publication type="book"/>...</publication>
> > <publication type="book"/>...</publication>
> > <publication type="book"/>...</publication>
> > <publication type="book"/>...</publication>
> > <publication type="reference"/>...</publication>
> > <publication type="misc"/>...</publication>
> > <publication type="misc"/>...</publication>
> > </library>
> >
> >
> > How else do you get around this? Giving each element a unique name would be
> > a little silly, especially when there could be hundreds of them. besides..
> > if you then wanted to transform the XML into some HTML (as i do).. it would
> > be much trickier.
>
> Andrzej Jan Taramina
> Chaeron Corporation: Enterprise System Solutions
> http://www.chaeron.com
>
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
>
> The list archives are at http://lists.xml.org/archives/xml-dev/
>
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://lists.xml.org/ob/adm.pl>
|
