| 摘要 |
Defining a software test is disclosed. A benchmark complexity for a test case including at least one test step having a check condition and a functionality is defined. The test case is represented as a flow graph in which the check condition of each test step of the test case is represented as an edge of the flow graph and the functionality of each test step is represented as a vertex in the flow graph. A test case complexity is determined by determining a number of independent paths in the flow graph. If the test case complexity as determined based on the number of independent paths exceeds the benchmark complexity, the number of test steps in the test case is reduced. |
