1. Things which I have done this week
(1) Read the source code about all domain classes and some code about the report module.
(2) Read the document about unit test framework used in OpenMRS.
2. Things which I am doing now
(1) Reading the API doc of XStream and some specific source code about those classes which will be used in this project.
3. Question List
(1) The form of writing the unit test for Serialization
As i see the document and some codes about unit test in Openmrs. I have a problem, i think if i should write a new Testcase for every domain object? or i can just write a new method with "@should" in the class "SerializatoinServiceTest" for every domain object?
Ben, can you tell me which test class i can refer to?
(2) About implementing the feature of "@Element(data=true)"
I think if i extends some classes of XStream and override methods in them in order to implement the feature "@Element(data=true)", then I don't know this manner whether will break some licences in XStream?
(3) The form through which I will do my project
Some projects will be done in a branch or module. And I want to know if I can directly code for my project in current core codes in Openmrs, Or also need make a new branch for my project?
Subscribe to:
Post Comments (Atom)
:-)
ReplyDelete1) The @should annotations won't be necessary. There will be so many tests that it would look silly. The tests will go int the serialization package and/or test
ReplyDelete2) I would prefer not to have to have to modify the xstream jar file. A quick google search for cdata xstream shows that there is a way to add CDATA with the pretty print class.
3) All projects are done in either branches or modules. Only very small patches just go against trunk. This is core work, so it will be done in a branch. See http://openmrs.org/wiki/Developers#Requesting_Subversion_Repository_Access for how to request your branch.