[squeak-dev] Re: Magritte Validation Question
Udo Schneider
Udo.Schneider at homeaddress.de
Fri Nov 6 10:28:06 UTC 2009
Thanks for the information. I'll keep my KISS "fix" for the moment until
you did the hard part :-) Currently the fix doesn't seem to break anything.
Best Regards,
Udo
Lukas Renggli schrieb:
> Looks like you've discovered yet another bug in exception handling.
>
> I am looking into removing these resumable exceptions now. We should
> not need to use them. That's too tricky.
>
> Lukas
>
>
>
> 2009/11/5 Udo Schneider <Udo.Schneider at homeaddress.de>:
>> After digging a bit more I think the following patch fixes the issue ... not
>> sure if it breaks anything else though:
>>
>> MAKindError>>#isResumable
>> ^ false
>>
>> CU,
>>
>> Udo
>>
>>
>>
>> Udo Schneider schrieb:
>>> Hi,
>>>
>>> I have a question regarding Magritte Validation. Given the following
>>> description
>>>
>>> descriptionAcquireRetryDelay
>>> ^ MANumberDescription new
>>> selectorAccessor: #acquireRetryDelay
>>> label: 'Aquire Retry Delay';
>>> group: 'C3P0';
>>> priority: 100;
>>> beRequired;
>>> beInteger;
>>> bePositive;
>>> yourself
>>>
>>> I do have a problem if the user enters a non-number. This is IMHO caused
>>> by #bePositive which checks if a given number is positive by sending
>>> #positive.
>>>
>>> However if the user enters "aaa" the object being validated is the String
>>> ... which does not understand #positive.
>>>
>>> If I remove the #bePositive condition everything is find and I do get
>>> (expected) Validation Errors complaining about invalid input and
>>> non-integer.
>>>
>>> So I'm wondering if I need to somehow specify to check the #bePositive
>>> condition (or any other condition for this case) only /after/ the correct
>>> type has been validated/determined.
>>>
>>> Any pointers?
>>>
>>> CU,
>>>
>>> Udo
>>>
>>>
>>>
>>
>>
>
>
>
More information about the Squeak-dev
mailing list
|