June 3, 2015 at 5:51 pm #8318NicholasParticipant@shicky
An improvement I’m targeting in my new position is the measurement of test code coverage. For example I have just finished refactoring a suite of functional tests and would be interested to know if I’ve dropped the level of coverage we previously had.
We’re using robot framework, with the underlying application written in c# if that impacts any recommendations.
An additional question for those that have implemented this….was it worthwhile or even a negative? It works well in my scenario but I do have concerns it could very easily be abused and the test team held to a random ‘70%’ coverage across all functional test suites.June 15, 2015 at 10:29 am #8426HuwParticipant@huwontour
Take a look at this:
Which is a really good set of videos explaining what coverage actually is.
This is good as well:
LLyr is a bit of a whiz on this topic, feel free to contact LLyr directly.June 18, 2015 at 5:32 pm #8511JasonParticipant@jasoncraig
One approach I’ve been pondering, as I too use Robotframework, is using Selenium library and use the highlight element keyword in conjuncture with capture screenshot keyword to highlight the html that is being asserted or acted upon.
This would give me a UI coverage map.
For back end I’m not sure what is recommended for C#.
JasonJuly 19, 2015 at 8:21 am #8810ThanhParticipant@rocky
Test Coverage more depends on how you count the test? You count tests based on number of test cases, test ideas, test environment, etc
- You must be logged in to reply to this topic.