In a recent project that I almost completed, we used an architecture that uses the XXXManager classes as the top layer for interacting with the web / services layers.
For example, there are Windows services that run on a schedule that import data from several different data sources into our system. Within this service, several "Manager" classes are called ie CPImportScheduleManager, CPImportProcessManager, etc.
Now these Manager classes are much more than just a method of passing up the chain for use in web / service layers. For example, my UserManager.Register () method is not only saved by the user through lower-level nodes, but also sends a WAP-push to the user and determines the mobile phone used, etc.
I was suggested that this type of architecture be the usual way to make OOP fit into the procedural model. I can see them here, but what interests me is a set of top-level classes, any level of web services can simply call the same general method without rewriting the code. Thus, if I wanted to write a web service that at some point registered the user, I could again simply call the UserManager.Register () method without overwriting all the logic.
I have never been the best person to explain myself, but if my mistakes make sense, feel free to advise your alternatives.
Hi Chris.
source
share