[CDBI] Re: how to debug broken inheritance?

Edward J. Sabol sabol at alderaan.gsfc.nasa.gov
Sun Nov 20 02:24:56 GMT 2005


> Generally this was working fine, except for one class which "refused"
> to inherit it's triggers and r'ships. Previously I thought it had
> something to do with that subclass defining its own triggers, but now I
> have an example of a class with *no* triggers of its own which also
> "refuses" to inherit from Page. In fact, this new class is very simple
> -- it adds only one has_a rship (to a DateTime object). However, this
> class does get used in a lot of other has_a rships in other classes.
>
> This makes me think the problem is not internal to this class... and
> I'm at a loss for how to debug. I don't have much knowledge of CDBI
> internals -- can anyone suggest where to look to start debugging this?
> Why would the inheritance rship get clobbered?

Which version of CDBI are you using? I was under the impression that recent
versions handle inheritance better, notably for has_a relationships.

If you are using the latest CDBI, then your course of action is to pare down
your classes into a failing test case, preferably using one of the t/*.t test
files from the distribution as a starting point so that Tony can just plop
your test case into his "t" directory and run "make test". Once you have the
failing test case, file a bug report at http://rt.cpan.org/.

Hope this helps,
Ed




More information about the ClassDBI mailing list