Hi,
using:
Sends the following to my action:
submit => [ Submit ] submit.x => [ 40 ] submit.y => [ 7 ]
x / y are coordinates of my click, defined by w3c. Now they are sent
to my action, and OGNL tries to set it. Which leads to:
> Error setting expression 'submit.x' with value '[Ljava.lang.Stri
Hi Christian,
unfortunately IIRC there is no way to avoid that OGNL tries to find the
property accessor.
Adding this row in your log4j.xml (properties) should mute this log message
(which it should not be a warning message)
On 26 July 2011 11:27, Christian Grobmeier wrote:
Thanks Maurizio.
I don't want to reduce my error level, now I created something weird.
Not nice, but it works:
private Submit submitCoord = null;
public Submit getSubmit() {
return submitCoord;
}
public void setSubmit(String submit) {
You may be able to write an interceptor that strips X and Y from your cgi
parameters. You would need to set the interceptor prior to the default
interceptor that assigns params to your action set methods.
On Jul 26, 2011 5:50 AM, "Maurizio Cucchiara"
wrote:
> Hi Christian,
> unfortunately IIRC th
@Christian
Good workaround,
in my defence I can say that I have no problem to change the log level,
thank to maven I can have different profile with different log level.
Maurizio Cucchiara
Il giorno 26/lug/2011 12.26, "Christian Grobmeier" ha
scritto:
> Thanks Maurizio.
>
> I don't want to reduc
Aaron's advice suggested me another way to obtain what you are looking for:
you could simple blacklist x & y parameters through the parameters
interceptor.
Maurizio Cucchiara
Il giorno 26/lug/2011 12.30, "Aaron Brown" ha
scritto:
> You may be able to write an interceptor that strips X and Y from
> @Christian
> Good workaround,
> in my defence I can say that I have no problem to change the log level,
> thank to maven I can have different profile with different log level.
hehe yes I have a similar setup I guess. But "expected exceptions"
make me always nervous, even in development mode ,-)
There are different schools of thought, this is kind of question that has
multiple answers: please, take a look at
http://struts.1045723.n5.nabble.com/S2-OgnlValueStack-Error-setting-expression-warnings-after-upgrade-from-struts-2-to-struts-2-1-7-td3495667.html
On 26 July 2011 14:58, Christian Gro
Just a question, if I may :-) Doesn't Struts 2 have the ability to
only set the properties that are defined on the bean?
On Tue, Jul 26, 2011 at 8:46 AM, Maurizio Cucchiara
wrote:
> There are different schools of thought, this is kind of question that has
> multiple answers: please, take a look a
That's exactly what's happening, but it's also warning (in devmode only)
when it can't set one so that you, the developer, can make the right
decision for your program.
(*Chris*)
On Tue, Jul 26, 2011 at 6:50 AM, Paul Benedict wrote:
> Just a question, if I may :-) Doesn't Struts 2 have the abi
I don't understand that. Perhaps I haven't programmed Struts 2 enough,
but I thought Struts (1 or 2) only captured what's defined in the
model and ignores the other request parameters. Is it acting
differently because of nested properties?
On Tue, Jul 26, 2011 at 9:54 AM, Chris Pratt wrote:
> Tha
I see, I am one of those who always want to see whats going on
Interesting read. Might be worth to compile a wiki entry?
On Tue, Jul 26, 2011 at 3:46 PM, Maurizio Cucchiara
wrote:
> There are different schools of thought, this is kind of question that has
> multiple answers: please, take a look
I think you're thinking about it backwards. It doesn't start with the bean,
inspect every property, attempt to find data and convert that data to the
proper format, then inject it into the bean.
It starts with the data on hand (in this case all the GET/POST parameters)
and attempts to find a home
Hi Paul,
you hit the nail on the head: if you try to get
http://localhost:8080/struts.action?whatever=whaever_value you won't see any
exception (in no-dev mode). On the contrary, when you use a nested property,
ognl change the target object.
For example if you use a "a" property, Ognl uses as a tar
>
> I think you're thinking about it backwards. It doesn't start with the
> bean, inspect every property, attempt to find data and convert that data to
> the proper format, then inject it into the bean.
Chris is right: take a look at row 292 of [1]
[1]
http://svn.apache.org/viewvc/struts/struts2
@Christian
I don't know if it does or not, IMHO there is something that is not proper
consistent between the nested property and the plain property (looking at
the code, I realize that theoretically only when the user runs s2 in dev
mode, this kind of exception should be throwned). My guess is tha
16 matches
Mail list logo
