[SofaKernel] FIX operator>> in Mat.h and add corresponding test.

[damienmarchal]

The operator>> was not functionning properly as it was returning
with a failbit set because of the in.peek() on empty string. This error
was generating a warning in UniformMass.

The PR:

• fix this error
• adds a test that validate that the code is correct and does not generate error anymore.
• makes some sout/serr cleaning in uniformmass.

---

This PR:

• builds with SUCCESS for all platforms on the CI.
• does not generate new warnings.
• does not generate new unit test failures.
• does not generate new scene test failures.
• does not break API compatibility.
• is more than 1 week old (or has fast-merge label).

Reviewers will merge only if all these checks are true.

[guparan]

Thanks @damienmarchal 👍
Could you explain how the new test checkRigidAttribute ends up using the operator>> you fixed?

[damienmarchal]

Hi Guillaume,

The problem of matrix reading was visible because of the UniformMass was generating a warning at run-time. The test was designed to fail if the uniformmass generates such a warning or cannot read the given values. Finally the error was not in UniformMass but in Mat.h, so I fixed the problem and keept the written test (not failing now).