Keyword driven test, how and where to derive keyword ?
We have a pilot automation project using keyword driven framework.
The concept and the engine / driver implementation are easy to understand but when come to actual test case design and implementation, automation engineer often has difficulty defining the right encapsulate high level business keyword.
Is there any guideline or "tips" that can help tester do some form of maping from Use Cases, Scenarios (stories) to keywords ?
Re: Keyword driven test, how and where to derive keyword ?
We generally sit down with a manual tester who is very familiar with the application and have them go through some of their tests. We use a three tier keyword with the level 1 doing the grunt work like pushing a button, selecting from a drop-down menu, etc. The highest level is probably the most difficult to isolate. It has to be high enough to handle maintenance changes quickly, but low enough to provide a useful function. Many times our high level, (level 3), keywords do complete sections of the functional requirement which has already been tested in detail, to get to the point where the lower levels can take over. An easy example is a data-driven test for logging in. A random ID and a random PW are used to test this function, then the correct inputs are used to validate it works as designed. In another test, the log-in test with the desired inputs is turned into a keyword at a high level so that the log-in function is used but not really tested. It is used as a tool to get to a specific screen to test it.