You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The idea of using sub-FSMs with a common interface for single tests is nice. It would be even cooler to extend that thought to parametrize these sub-FSMs with input and expected output values.
What I had in mind is:
Extend the TestHandler interface to include a method Bool success() which indicates if the test was a success
Provide a default module ``mkTestCheckExpected(#parameter t input, #parameter t2 expectedOut) (TestHandler) which extends the currently implemented functionality of mkTestsMainTest by a check if the provided output of the dut matches the parameter expectedOut
Create a module mkDashBoardTestbench which uses the mkTestCheckExpected module and keeps track of successful and failed tests and displays the total statistics of totally passed tests.
Extend the documentation to showcase an example of the new functionality.
The main idea is to provide a basic functionality similar to assertEquals in programming so it's easy to generate testcases and see if the latest commits to a project actually improved the code (i.e. more tests passing).
The text was updated successfully, but these errors were encountered:
Such a feature was always on "my list" but I never got around to it. My plan was to use the ModuleContext features of Bluespec to annotate "Testing" modules that are automatically enumerated and run when compiling with e.g. make test and use some kind of assert functions to check for failures. Using BlueTCL one could even get the annotated modules programmatically to enable e.g. parallel simulation. In the end it would look a bit like tests in rust https://doc.rust-lang.org/book/ch11-01-writing-tests.html
And each module of type BSVToolsTest (name is just an example) would collect test information via ModuleCollect?
The bracket notation is what I understand under ModuleContext, or do you think of something else?
The idea of using sub-FSMs with a common interface for single tests is nice. It would be even cooler to extend that thought to parametrize these sub-FSMs with input and expected output values.
What I had in mind is:
TestHandler
interface to include amethod Bool success()
which indicates if the test was a successThe main idea is to provide a basic functionality similar to assertEquals in programming so it's easy to generate testcases and see if the latest commits to a project actually improved the code (i.e. more tests passing).
The text was updated successfully, but these errors were encountered: