> Sorry, my statement was wrong:
> If you register your converter for Object and Serializable, everything
> should be fine.
>
> Sven
>
>
> On 09/09/2012 07:21 PM, Alec Swan wrote:
>>
>> In which case can it cause a ClassCastException?
>>
>> On Sun, Sep 9, 2012 at 6:25 AM, Sven Meier <
[hidden email]> wrote:
>>>>
>>>> public Object convertToObject(String value, Locale locale) { return
>>>> value;
>>>> }
>>>
>>> That might fail badly causing a ClassCastException but as long as it
>>> works
>>> for you ;).
>>>
>>> Thanks for creating the issue.
>>> Sven
>>>
>>>
>>>
>>> On 09/09/2012 09:22 AM, Alec Swan wrote:
>>>>
>>>> Here is a new JIRA
https://issues.apache.org/jira/browse/WICKET-4755>>>>
>>>> I added the following code in my application class to implement what
>>>> you recommended as a fix. Maybe this will help other people who run
>>>> into this problem. Does this look good to you?
>>>>
>>>> @Override
>>>> protected IConverterLocator newConverterLocator() {
>>>> ConverterLocator locator = (ConverterLocator)
>>>> super.newConverterLocator();
>>>> locator.set(Object.class, new BasicConverter());
>>>> locator.set(Serializable.class, new BasicConverter());
>>>> return locator;
>>>> }
>>>>
>>>> private static class BasicConverter implements IConverter
>>>> {
>>>> @Override
>>>> public Object convertToObject(String value, Locale locale) {
>>>> return value;
>>>> }
>>>>
>>>> @Override
>>>> public String convertToString(Object value, Locale locale) {
>>>> return value == null ? null : value.toString();
>>>> }
>>>> }
>>>>
>>>>
>>>> Thanks,
>>>>
>>>> Alec
>>>>
>>>> On Sat, Sep 8, 2012 at 3:47 PM, Sven Meier <
[hidden email]> wrote:
>>>>>
>>>>> Hi,
>>>>>
>>>>> Objects#convertValue()'s javadoc states that it returns null, if the
>>>>> value
>>>>> can not be converted, thus your proposed change isn't valid for most
>>>>> cases.
>>>>>
>>>>> You could easily fix this issue in your application by registering a
>>>>> custom
>>>>> converter for Serializable.
>>>>>
>>>>> Since Objects#convertValue() does more actually than "convert[ing]
>>>>> numerically" (as stated in the javadoc, it handles booleans and Strings
>>>>> too), I think the method could be improved, e.g by adding another
>>>>> if-statement to it:
>>>>>
>>>>> if (toType.isInstance(value))
>>>>> {
>>>>> result = toType.cast(value);
>>>>> }
>>>>>
>>>>> Please open an issue in Jira.
>>>>>
>>>>> Sven
>>>>>
>>>>>
>>>>>
>>>>> On 09/08/2012 09:11 PM, Alec Swan wrote:
>>>>>>
>>>>>> An easy fix for this bug is to change the first line in
>>>>>> org.apache.wicket.util.lang.Objects#convertValue from
>>>>>> Object result = null;
>>>>>> to
>>>>>> Object result = value;
>>>>>>
>>>>>> Thanks,
>>>>>>
>>>>>> Alec
>>>>>>
>>>>>> On Sat, Sep 8, 2012 at 1:02 PM, Alec Swan <
[hidden email]> wrote:
>>>>>>>
>>>>>>> I attached a quickstart with a test in
>>>>>>> TestHomePage#formSubmitsSuccessfully.
>>>>>>>
>>>>>>> The test throws "'NEW VALUE' is not a valid Serializable" error when
>>>>>>> "NEW VALUE" string in "value" textField is submitted as a part of
>>>>>>> myForm ajax submission.
>>>>>>>
>>>>>>> This bug should have a high priority in 1.5 because this used to work
>>>>>>> in 1.4 and there is no clear way to track down all places where this
>>>>>>> needs to be fixed in the existing code base.
>>>>>>>
>>>>>>> Thanks,
>>>>>>>
>>>>>>> Alec
>>>>>>>
>>>>>>> On Sat, Sep 8, 2012 at 12:16 AM, Paul Bors <
[hidden email]> wrote:
>>>>>>>>
>>>>>>>> Can you please put together a quick start for us?
>>>>>>>>
>>>>>>>> ~ Thank you,
>>>>>>>> Paul C Bors
>>>>>>>>
>>>>>>>> On Sep 7, 2012, at 14:39, Alec Swan <
[hidden email]> wrote:
>>>>>>>>
>>>>>>>>> Hello,
>>>>>>>>>
>>>>>>>>> I decided to extract this issue into a separate thread.
>>>>>>>>>
>>>>>>>>> I keep running into issues with default type converters in 1.5. I
>>>>>>>>> also
>>>>>>>>> think I found a bug in
>>>>>>>>> org.apache.wicket.util.lang.Objects#convertValue. A call to
>>>>>>>>> convertValue(nonNullNonArrayValue, Object.class) will always return
>>>>>>>>> null if nonNullNonArrayValue is a value that is not null and not an
>>>>>>>>> array! Shouldn't it always return the first parameter when the
>>>>>>>>> second
>>>>>>>>> parameter is Object.class?
>>>>>>>>>
>>>>>>>>> Is this a legitimate bug?
>>>>>>>>>
>>>>>>>>> Thanks,
>>>>>>>>>
>>>>>>>>> Alec
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> ---------------------------------------------------------------------
>>>>>>>>> To unsubscribe, e-mail:
[hidden email]
>>>>>>>>> For additional commands, e-mail:
[hidden email]
>>>>>>>>>
>>>>>>>>
>>>>>>>> ---------------------------------------------------------------------
>>>>>>>> To unsubscribe, e-mail:
[hidden email]
>>>>>>>> For additional commands, e-mail:
[hidden email]
>>>>>>>>
>>>>>> ---------------------------------------------------------------------
>>>>>> To unsubscribe, e-mail:
[hidden email]
>>>>>> For additional commands, e-mail:
[hidden email]
>>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail:
[hidden email]
>>>>> For additional commands, e-mail:
[hidden email]
>>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail:
[hidden email]
>>>> For additional commands, e-mail:
[hidden email]
>>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail:
[hidden email]
>>> For additional commands, e-mail:
[hidden email]
>>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
[hidden email]
>> For additional commands, e-mail:
[hidden email]
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
[hidden email]
> For additional commands, e-mail:
[hidden email]
>
Locked
invalidserializable.zip (55K) 