Joerg Heinicke dijo: > On 19.02.2004 12:57, Antonio Gallardo wrote: > >>>That's correct, but we also should have a future validation in mind. A >>>form definition like Daniel's sample below must raise an exception then. >>>Furthermore what happens if you have an additional widget (outside of >>>repeater) with @id="foo.1.foo". This must raise an exception too of >>>course thiugh the definition itself can be valid from XML point of view. >> >> >> You are right, but I see a potential slowdown of the CForm transformer. >> AFAIK, these type of validation is interesting just at development time, >> but not in a production environment when the same definition (no >> changed) >> will be validated over and over. >> >> WDYT? > > This can not be done statically (i.e. only during development) as a > single widget could have an id like "foo.25.foo". As long as the > repeater stays below 25 rows it's ok, with more than 25 not. > Another possibility would be to exclude "foo.x.foo" in general ...
I will prefer to exclude "foo.x.foo" in general. If not, we will need to check that will be a slowdown at all. Best Regards, Antonio Gallardo
