fixed the checkJointTypes function for the case of TORQUE control

[AlexAntn]

This PR fixes issue #718 .

Since we have an array of PidInfo / TrqPidInfo, we can't rely on class inheritance to query the parameters of each PID (an array of Derived is not a kind-of array of Base). As such, we need to check if the PID type is torque and handle it differently.

[pattacini]

I can see a verbatim copy'n'paste of lots of code lines we could thus refactor and made much more compact.

[pattacini]

@AlexAntn we can get away w/ less by refactoring the main routine and make it call 2 sub-routines that implement each the body of the 2 for loops.

In this case, indeed, each subroutine can stick to the generic case of the base class as we don't deal w/ pointers arithmetic inside the subroutines' body.

[AlexAntn]

Ìn the last commit to this issue I have refactored the loops in order to reduce duplicated code. I have also changed how we obtain the firstjoint in order to have a single loop.

These changes have been tested directly on the robot and are working (yarprobotinterface started with no issues)

[pattacini]

LGTM, thanks a lot @AlexAntn 🚀

[pattacini]

I'll wait for the CI to complete and then squash and merge the PR.

@Nicogene in the meantime, you may be willing to review it as well.

[Nicogene]

LGTM

[Nicogene]

I think we can squash all the commit in just one

[pattacini]

I think we can squash all the commit in just one

Don't do that please, otherwise CI will start over from scratch!

I can simply use the corresponding button from the GH GUI, instead, to squash the commits.