The SCT binary will traverse through a given directory, loads all available test files (defineable) and their configurations. Afterwards the tests are run (executed) and the output is checked if it is identical to a desired output. Errors are printed additionally to an statement whether the test failed or not.
An equally to the test file named result file (xxx.res) is used for validation of output. More specific: the output of a test process is compared to the content of this result file. If they are identical the test is successful, else it is failed.
Each test can be located somewhere within the given directory. If in the test files directory a similar named xxx.conf is found, this configuration is used. If no configuration file was found, the testManager will look in a directory one level higher. IMPORTANT: as this new directory can contain multiple .conf-files only an "base.conf" file is used to avoid ambiguity.
So this is allowed and gets used:
- sampletests/base.conf
- sampletests/mytest/mytestA.conf
- sampletests/mytest/mytestA.bmx <- uses mytest.conf
- sampletests/mytest/mytestB.bmx <- uses ../base.conf
- sampletests/mytest2/base.conf
- sampletests/mytest2/mytestA.bmx <- uses base.conf
- ... and so on
Each configuration file inherits the configuration of a parent folder's "base.conf" (which also inherits from its parent).