On 21.04.2004 13:53, Marc Portier wrote:

so I'm going for getId on form to return "" (which it was)
and removing getId() == null checks since I favour the NPE here

Please not waiting for the NPE somewhere latter in the code for the reasons Ugo pointed out.


7/ should validation stop as soon as possible or continue to allow all validation errors to be set?

Don't know which is best. *If* we cannot decide (due to different needs in different usecases) then we will need to add an attribute to let the form designer control the behaviour.

hm, we should at least be consistent

clarifying a bit more:


we seem to have two distinct cases:
widgets with child-widgets
- i've seen code that doesn't evaluate the parent if one of the kids is
invalid
- but all kids seem to be evaluated even if a previous brother failed

widgets with validation on definition and widget level
- widget level rules stop evaulation if the definition level was not ok,
and as soon as one of the widget-level validations fails

I haven't given it much thought yet, if there are clear positions out
there I'm quite happy to just enforce code wise (and add some javadocs
removing any possible future surprise)

so comments welcome

IMO validation should always run on all widgets as it must be possible for the user to get a whole form fixed in one step:
http://marc.theaimsgroup.com/?l=xml-cocoon-dev&m=107962387820282&w=4
http://marc.theaimsgroup.com/?l=xml-cocoon-dev&m=107963156530885&w=4


Joerg

Reply via email to