HandMk5CouplingHandler: refactor it in order to make it configurable

[Nicogene]

It fixes #661

This is a first draft, I thoght that the configuration file will become:

[COUPLING]
icub_hand_mk5 (0 1 2 3 4 5 6 7 8 9 10 11) (l_thumb_add l_thumb_oc l_index_add l_index_oc l_middle_oc l_ring_pinky_oc reserved reserved reserved reserved reserved reserved) (0.0 0.0 0.0 0.0 0.0 0.0) (90.0 82.1 15.0 90.0 90.0 90.0)

[COUPLING_PARAMS]
L0x     -0.00555 -0.00555 -0.00555 -0.00555 -0.00555
L0y      0.00285  0.00285  0.00285  0.00285  0.00285
q2bias     180.0    180.0    180.0    180.0    180.0
q1off       4.29     4.29     4.29     4.29     4.29
k         0.0171   0.0171   0.0171   0.0171   0.0171
d        0.02006  0.02006  0.02006  0.02006  0.02006
l         0.0085   0.0085   0.0085   0.0085   0.0085
b        0.00624  0.00624  0.00624  0.00624  0.00624

Right now I am asking to specify all the coefficient for all the fingers, but @xEnVrE said that only thumb index and pinky is required, maybe I can ask only those?

cc @pattacini @traversaro

[xEnVrE]

Right now I am asking to specify all the coefficient for all the fingers, but @xEnVrE said that only thumb index and pinky is required, maybe I can ask only those?

In my opinion, I would leave them all just for clarity also considering that in the code they are easily handled within a for cycle and just once at configuration time.

I will review the PR within today and tomorrow.

[xEnVrE]

Thanks @Nicogene. I have added some comments for requested changes.

[Nicogene]

As suggested by @xEnVrE, with the usage of a map, we got rid of some functions that are now useless

[xEnVrE]

Thanks @Nicogene, I added some comments for requested changes. It seems that at the moment the code was not using the values for the middle and the ring and still using those of the index for them.

[Nicogene]

@xEnVrE for the pinkie/pinky I am a little confused what is the correct name, in the documentation there is a pinkie

In any case it is ok for me if we prefer pinky

[xEnVrE]

for the pinkie/pinky I am a little confused what is the correct name, in the documentation there is a pinkie

Then there is a bit of a mix. In this part of the documentation:

pinky is used instead.

@Nicogene

[xEnVrE]

LGTM! Thanks @Nicogene

Important
We need to wait for icub-tech-iit/ergocub-software#169 to be ready for merging I guess.

[Nicogene]

@traversaro I think we can ignore the failures, they seems strange conda/windows failures that seems unrelated

 LINK : fatal error LNK1181: cannot open input file 'C:\Program.obj' [D:\a\gazebo-yarp-plugins\gazebo-yarp-plugins\build\plugins\modelposepublisher\gazebo_yarp_modelposepublisher.vcxproj]
  Building Custom Rule D:/a/gazebo-yarp-plugins/gazebo-yarp-plugins/thrift/worldinterface/CMakeLists.txt

I saw it also in other repositories (icub-basic-demos)

[traversaro]

For the future, please avoid functional and whitespace changes in the same commit, thanks!

[traversaro]

@Nicogene can you add a note in the CHANGELOG, also considering that (if I am not wrong) this change breaks old mk5 simulations, so there is a strict dependency between future gyp releases and ergocub-software releases.

[traversaro]

@traversaro I think we can ignore the failures, they seems strange conda/windows failures that seems unrelated

 LINK : fatal error LNK1181: cannot open input file 'C:\Program.obj' [D:\a\gazebo-yarp-plugins\gazebo-yarp-plugins\build\plugins\modelposepublisher\gazebo_yarp_modelposepublisher.vcxproj]
  Building Custom Rule D:/a/gazebo-yarp-plugins/gazebo-yarp-plugins/thrift/worldinterface/CMakeLists.txt

I saw it also in other repositories (icub-basic-demos)

See #664 . If you see this kind of environmental regression, please report them in new issues. Unfortunately in GitHub actions only the last person that modified the ci files is getting notification of CI failures (see #664), so it is not possible for me or any other mantainer to get all possible notifications.