Hi,
A newbie question on struts.
The way a view is built using struts is as follows (I think!). After
the action class processes the input and makes business logic method
invocations on the model layer (say session beans or business delegates), it
needs to fetch the data required for the next view and populate some value
objects . The view then accesses these objects (thru html:text, bean:write
tags etc) to create the html output.
The problems I see with this is that the action class becomes
extremely cluttered with the logic for fetching and populating the value
objects, especially when there are multiple 'commands' the action class is
designed to process ('sort', 'edit', 'save' etc) and there are several
forwards out of the action with different views.
(I can see that the question of where to take the data from becomes
a problem when there are validation errors. In this case, the view has to
take the data from the ActionForm and not from the model. But this is a
lesser problem than the above mentioned problem of action class complexity,
I think.)
Isn't it better that the view populates itself from the model via
system state beans.These can be simple beans which query the model layer for
the values required by the view. This way, the action class need not be
modified, if the view is changed to add a new display-only field, for
example. Struts does not seem to support this since we cannot take the data
from one bean (system state beans) and submit via another (ActionForm bean).
Cheers,
Krishna
--
To unsubscribe, e-mail: <mailto:struts-user-unsubscribe@;jakarta.apache.org>
For additional commands, e-mail: <mailto:struts-user-help@;jakarta.apache.org>
