Re: [Simple-support] Inline list of simple elements with different names

[Jarrod S. <ja...@co...>]

I just found this thread:
http://sourceforge.net/mailarchive/message.php?msg_id=26681126

Cameron's was the exact same problem I am having. So it seems there is
no way around this, without a setName() on the InputNode. Niall I
don't understand your final suggestion in that thread to "use the
Style interface" since that only applies to serialisation to XML, and
not deserialising *from* XML, which is when Cameron and I are needing
to modify the element names. Or am I missing something?

To be clear, what I want to do is rename the <A>, <B>, <C> ... etc.
elements below to all be <ERROR> nodes.

Any suggestions greatly appreciated.

- Jarrod

On Sun, Nov 6, 2011 at 9:37 PM, Jarrod Smith <ja...@co...> wrote:
> I have a doc that has something like the following structure:
>
> <FAIL | OK>
>
>       <!-- In the FAIL Case: -->
>
>       <A>Error relating to A</A>
>       <B>Error relating to B</B>
>       <C>Error relating to C</C>
>        ...
>       <ERROR>General Error Message</ERROR>
>       <ERROR>General Error Message</ERROR>
>       <ERROR>General Error Message</ERROR>
>       ...
> </FAIL | OK>
>
> There could be a lot of different A, B, C, ...., and to make matters
> worse, I haven't even been provided with a specification as to what
> they all could be.
>
> The contents of the elements is always just a string. So the document
> is really just a flat list of elements containing strings. Is there a
> simple way for me to declare a Class to parse this?
>
> Something like:
>
> @Root
> public class MyErrorClass {
>    @ElementList(inline=true)
>    public List<APIError> errors;
> }
>
> @Root                         // <--- Don't want this to care about
> the name of the element.
> public class APIError
> {
>        @Text
>        String error;
> }
>
> Another idea I had was to just use a Visitor to stomp all the element
> names and just make them all ERROR. But it doesn't look like you can
> change the element name of an InputNode. Anyway, this idea would only
> work if the TreeStrategy visits the document Root node before any
> others, so I could set a flag indicating this was a <FAIL> document.
>
> Any suggestions appreciated.
>