[cppyy] fix iadd pythonization for np/array buffers

[aaronj0]

Backport of wlav/CPyCppyy#51
Fixes #18768

Currently, the iadd implementation for std::vector in Pythonize.cxx works perfectly with Python lists and iterables for which we have an ItemGetter. But the code path for numpy/array buffers do not. The difference between x+=y and x.__iadd__(y) leads to unintuitive behaviour: if the intent is to obtain the resulting iterator when calling insert, the user can do d.insert(d.end(), a) which makes more sense than expecting it from d+=a. When the user does call d+=a, since __iadd__ is overriden, operation does not end with the call to VectorIAdd which should just return the iterator, and reflect the change by insertion on d. Instead, it ends with d = d.__iadd__(a) which reassigns d to the returned iterator, losing the vector in the process. This PR fixes this by returning self in our pythonization.

[github-actions]

Test Results

    20 files      20 suites   3d 16h 35m 20s ⏱️
 3 013 tests  3 011 ✅ 0 💤 2 ❌
58 672 runs  58 670 ✅ 0 💤 2 ❌

For more details on these failures, see this check.

Results for commit d81e2c3.

♻️ This comment has been updated with latest results.

[dpiparo]

Thanks for these changes. Can we add a (some) test(s)?