What do you mean by "are all test cases covered"?
- are all test cases executed?
- are all test cases ready for execution?
- are all test cases relevant to the test objectives?
- are all test cases associated with/covering some requirement?
- are all requirements covered by/associated with at least 1 test case?
a binder ... how retro [img]/images/graemlins/laugh.gif[/img]
I am just new to this field...am bacically ABAP developer, my top managemnt wants me to write test cases to cover the functionality...they have given Functional Specification for writing test cases for the application which is done in .NET( front end) and accessing the application(interface BAPI-business application programming interface) through SAP( backend)......here my question is how can i know whether all functionality has been covered in the application by simply writing test cases( in between requirements has been changed)..
What you're looking for is a test coverage matrix or a requirements traceability matrix including test cases. It's essentially a map tracing the links between each requirement and the test(s) that verify the requirement has been met. Typically what you want to see is at least 1 test case traced to each and every requirement, and all those test cases executed and (eventually) passed. That would indicate coverage.
As to your objective of 'showing' management that you've covered the requirements, said matrix will do right nicely. [img]/images/graemlins/laugh.gif[/img]
<font color="red">But</font>, as to whether that indicates the product is defect free or that your test cases adequately test the minutiae of the requirement or that they are even correct or that the requirements as stated are testable or even valid ... these are questions that cannot be answered directly by such a coverage matrix. [img]/images/graemlins/frown.gif[/img]
Sounds as if you have been put into a difficult position. The advice Bob gave on the Requirements Traceability Matrix is just what I would do. To find out more on this you might might want to search on www.stickyminds.com, i seem to recall some useful articles there.
And of course you need to think outside of just running tests which show the software works as per documentation, negative testing, boundary value and the like should be used to check for stability etc.