Re: [cf-orm-dev] Entity inheritance - invoking method on SUPER

[Jon Clausen]

Once you create an explicit method in the child class that matches the dynamic accessor method (e.g. getFieldname() ), and that method doesn't exist in the super class, that method becomes the accessor for the package.

For what you're looking to do, I'm not sure inheritance is the best solution. I would suggest mapping ReportCriteria as a many-to-one property, so that your accessor overloads can call it directly. The other option is to ensure that your overload methods exist in the super class.

Jon

[Note: Typo assistance courtesy of iPhone]

> On Feb 4, 2015, at 2:58 PM, Tristan Lee <trista...@gmail.com> wrote:
>
> Trying my first go at entity inheritance...
>
> I have ReportCriteriaAssigned that extends ReportCriteria. For ease on the eyes, here are gists to the source:
>
> ReportCriteriaAssigned - https://gist.github.com/tristanlee85/c22a2f379ced2a621e87
> ReportCriteria - https://gist.github.com/tristanlee85/ce986c199536b38c8508
>
> Between the 2 entities, both have a "fieldname" and "label" property. Idealy, if either of those properties in ReportCriteriaAssigned have a value, I would like to use that. Otherwise, if they are NULL, then I want to use the "fieldname" / "label" property from ReportCriteria since it will always contain a default value.
>
> I created 2 methods getFieldname() and getLabel() which check for the existance of the property and invoke super.get...() if it doesn't exist. However, it appears that the call to super is still referencing the properties of the subclass. It could be that my thought on a superclass for Hibernate has be confused with traditional techniques... thoughts?
>
> --
> You received this message because you are subscribed to the Google Groups "cf-orm-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to cf-orm-dev+...@googlegroups.com.
> To post to this group, send email to cf-or...@googlegroups.com.
> Visit this group at http://groups.google.com/group/cf-orm-dev.
> For more options, visit https://groups.google.com/d/optout.

[Tristan Lee]

Thank you for the reply. I had additional comments to add to this yesterday, but it seems the post was deleted. I'm not sure if that was something I did or if it was still waiting to be moderated...

Regardless, you are correct in how it should have been mapped. After messing with it for a while, this table structure wasn't a good fit for the inheritance model I was going for, and switching to a many-to-one was a better solution.