|
This post was updated on .
Hi, David
I come up with two ideas, one is simple check, one is complex check.
Simple check just sets a flag to false before running the example, and reset it to true in the end of the example. If all test function is OK, the flag is true, so just check the flag. Besides, we can check the example log, if the flag is true and no "error" in its log file, then we can assert that this example is successful.
Complex check means we need match file for each test point, this will be more effective to know whether the example is successful or not. But the disadvantage is many match file in the project which is not a good thing for carbondata I think.
So I suggest simple check, any idea ?
|