|
As per my experience
with IBatis and as per my understanding of your problem, IBatis calls the setter
properties on an object only once when it is populating the object with data. I
have debugged the objects passed to IBatis many times and the setter properties
are called only once for an object population. The behaviour as per my
experience is that when IBatis creates an object via reflection, all the member
variables will be initialized to their default values. You can override this by
setting the values of the member variables to your business-domain related sentinel
values in default constructor. I hope this helps. Can you provide the
snippet of your code that is throwing the exception?
From: Arjen Smits
[mailto:[EMAIL PROTECTED] I have an question concerning how ibatis fills the
object it maps its data to. I know it uses the properties of an object to set its
data. However in what sequence does it do this? Does it first initialises the field
by setting it to null? Or does it do some other form of initialisation first
before filling the object with the actual data? The reason I am asking is because currently I
implemented an sort of data integrity checking in the data object. However ibatis triggers the error mechanism when it
try’s to set the data in the object. For example: I got an Class User. This class has the String
Username property. In the set procedure of this property I check if the
value being used is not null and the string is longer then 0. If it is I throw
an specialised exception. So what happens now is that when I query the database
with an select and ibatis tries to fill the user object it throws this
specialised exception, indicating that the value being used is either null or
an empty string. After some debugging I found out it actually fills it with an
empty value first, afore putting in the “real” value. So it seems to initialise the object first with empty
values. Can I get confirmation of this behaviour ? | ||||||||||||
- Confirmation about how ibatis fills the objects it maps to... Arjen Smits
- RE: Confirmation about how ibatis fills the objects i... Jehanzeb Musani
- RE: Confirmation about how ibatis fills the objects i... Arjen Smits


