Hello,
I'm currently using XMLBeans 1.0.4 to parse a rather complex schema with
lots of abstract types, type-inheritance and use of substitution-groups.
After the parsing process all XmlObject instances are then tranformed
into POJOs for serialization reasons.
Because of a complex type-hierarchy the transformation code has a lot
of code like this:
if (Y.type.isAssignableFrom(x.schemaType())) {
Y y = (Y)x.changeType(Y.type)
...
} else if (Z.type.isAssignableFrom(x.schemaType())) {
Z z = (Z)x.changeType(Z.type)
...
} else if { ...
Why is changeType() necessary at all? If I print the class of 'x' in
the example above, I get 'ZImpl' in the second case.
So when XMLBeans already creates the right class instance *before*
calling changeType(), where is the difference between
Z z = (Z)x.changeType(T.type);
and a simple
if (x instanceof Z) {
Z z = (Z)x;
...
}
I don't understand the javadoc comment on changeType() so well,
especially this part:
"..., but there are no constraints at the roottype changes are never
prohibited at the root of an xml tree."
What does this mean? Under which circumstances can I rely on using a
direct cast delivering the same result as a call to changeType()?
Thanks in advance,
Robin.
PS: I will use XMLBeans 2.0 soon (substitution-groups!), so any answer
based on 2.0 would be equally appreciated!
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
