Fix #1143, SB_UT corrections and clear event count history after setup

[thnkslprpt]

Checklist

• I reviewed the Contributing Guide.
• I signed and emailed the appropriate Contributor License Agreement to [email protected] and copied [email protected].

Describe the contribution

• Fixes Unit test should clear event count history after setup #1143
  • Clears event history after setup calls and updates EVENTCOUNT asserts to the lower number of events only expected from the test under question (rather than including the total events from the setup calls as well)

While looking at this, I realised that lots of tests were checking events related the 'setup', not the actual function that is the focus of the test.

For example, there are dozens of asserts for CFE_SB_PIPE_ADDED_EID and CFE_SB_SUBSCRIPTION_RCVD_EID - even though CFE_SB_CreatePipe() and CFE_SB_Subscribe() have their own tests...

These asserts are largely redundant and just add clutter to the tests - I've removed the obvious cases from the tests I was already updating with UT_ClearEventHistory().

Test_SB_Cmds_MapInfoDef was checking for CFE_SB_PIPE_ADDED_EID and CFE_SB_SUBSCRIPTION_RCVD_EID which are part of the setup, not the function under test. I changed this to check for the command counter being incremented instead.

Test_Unsubscribe_Basic, Test_Unsubscribe_AppId and Test_Unsubscribe_Local were checking for CFE_SB_SUBSCRIPTION_RCVD_EID. They should really be checking for CFE_SB_SUBSCRIPTION_REMOVED_EID given that CFE_SB_Unsubscribe() is the focus of these tests.

• In the case of Test_Unsubscribe_Local, it was actually sending a CFE_SB_UNSUB_NO_SUBS_EID event. The reason this wasn't noticed may be because CFE_SB_UnsubscribeLocal() returns CFE_SUCCESS (and sends the CFE_SB_UNSUB_NO_SUBS_EID event) if the pipe is not subscribed to the MsgId.
  • I've amended the test to subscribe using CFE_SB_SubscribeLocal(), instead of CFE_SB_Subscribe(), and now the call to CFE_SB_UnsubscribeLocal() in the test actually works and returns CFE_SUCCESS while sending the CFE_SB_SUBSCRIPTION_REMOVED_EID event.

Even though these are coverage tests, they should still be clear and at least check for what they are testing...

Testing performed
GitHub CI actions all passing successfully (incl. Build + Run, Unit/Functional Tests etc.).
Local testing confirms net coverage unchanged.

Expected behavior changes
No change to code behavior. Test behavior updated as described above.

System(s) tested on
Debian GNU/Linux 11 (bullseye)
Current main branch of cFS bundle.

Contributor Info
Avi Weiss @thnkslprpt

[skliper]

Even though these are coverage tests, they should still be clear and at least check for what they are testing...

Love it, great catch on the UT bug in Test_Unsubscribe_Local!

[thnkslprpt]

Love it, great catch on the UT bug in Test_Unsubscribe_Local!

Cheers Jake.

Yeah SubscribeLocal()/UnsubscribeLocal() are fully tested with functional tests, but still best to keep the coverage tests correct and up-to-date anyway of course.