(Re)introduce a list of macros that can avoid testing - #2188

@aphecetche yes, this logics looks fine for me, but I don't understand failing checkcode, it mentions Detectors/gconfig/commonConfig.C as excluded (while it is not and its test is requested) then produces a Fatal on it.

Concerning currently excluded macros:

Detectors/FIT/T0/macros/run_reco_t0.C
Detectors/FIT/T0/macros/readHitsDigits.C
Detectors/ITSMFT/ITS/macros/test/CheckLookUp.C

are operational macros (I used them occasionally) and passing the test, why they were excluded?

@shahor02 the o2checkcode fails because some macros require simulation parts that are not built. Means I have to change a bit the logic : always call o2_add_test_root_macro but add a parameter to signal whether or not it should actually perform the test, depending on whether or not BUILD_SIMULATION is on.

For the 3 missing macros, no particular reason as far as I can tell, will put them back.

@aphecetche I refer to this failure in checkode:

-- [ ] [ ] [E] Detectors/gconfig/commonConfig.C
CMake Error at cmake/O2ReportNonTestedMacros.cmake:79 (message):
  Macro Detectors/gconfig/commonConfig.C should be tested
Call Stack (most recent call first):
  CMakeLists.txt:112 (o2_report_non_tested_macros)

macro declared as excluded and then reported to be in neither of the list (while it is requested to be tested)

@shahor02 o2_add_test_root_macro for commonConfig.C is called in a sub_directory which is not included when BUILD_SIMULATION=OFF (which is the case when some pythia or geant stuff is not found, e.g. on o2checkcode CI slaves)

Contrary to my comment above though, can't really add a parameter to o2_add_test_root_macro as in the case for commonConfig.C exemplifies, sometimes a full CMakeLists.txt is skipped.

Guess the exclusion list have to include a couple of if(BUILD_SIMULATION) and if(pythia6_FOUND) ...

I see, thanks!

on closer look :

Detectors/FIT/T0/macros/run_reco_t0.C
Detectors/FIT/T0/macros/readHitsDigits.C

were excluded because they are the same as the ones in Detectors/FIT/macros so in the end I've simply removed those duplicates.

For the rest, macros that are specific to simulation are now put on the exclusion list when required. That should fix the o2checkcode build.

@shahor02 @davidrohr Don't exactly understand the issue with clang-format ...

Doing git-clang-format locally the line reported by travis I get :

> git-clang-format --commit 347200c0716c5d2e0467af7b7946b9f5b58790a2 --diff
no modified files to format

..... but the others are green.

@aphecetche It complains about the macro which should be anyway removed since duplicated (as you found). I'll merge this PR then remove the macro and update the lists