Add begin/end member functions for iDynTree vectors

[GiulioRomualdi]

This PR:

• Implement begin()/end() and cbegin()/cend() methods for VectorFixSize
• Implement begin()/end() and cbegin()/cend() methods for VectorDynSize
• Update the CHANGELOG.md

The implementation of these members will allow the user to use range-based for loop.

Related issue #645

[traversaro]

Thanks for the contribution @GiulioRomualdi ! Technically this is a (small) new feature, so I would prefer to merge it in devel and eventually release it as iDynTree 1.1 , instead of merging it in master and release it as part of iDynTree 1.0.2, to comply with SemVer (see https://semver.org/).

However, as it is the "saturday that is made for the men, not the men for the saturday", i.e. we follow guidelines only if we gain advantage from them, not because they are valuable on its own, if you have any reason why you want this in master let me know, thanks!

[GiulioRomualdi]

Hi @traversaro. Devel is fine!

[traversaro]

Just to clarify, now that I cleaned up a bit the release process of iDynTree, I think if we need a iDynTree 1.1 we can easily release it when we need it, without waiting years. : )

[traversaro]

Hi @traversaro. Devel is fine!

Perfect, then feel free to change the target branch.

[GiulioRomualdi]

Rebased on top of devel

[GiulioRomualdi]

The spurious commits are there because master is not merged in devel

[DanielePucci]

As for future reference,

"saturday that is made for the men, not the men for the saturday"

it is Gospel of Mark, 2:27. I had to google for it, I believed it was actually Gospel of Mathew.

CC
@traversaro @GiulioRomualdi

[traversaro]

The spurious commits are there because master is not merged in devel

Sorry, I fixed it now.

[GiulioRomualdi]

The spurious commits are there because master is not merged in devel

Sorry, I fixed it now.

Rebased on devel. You can merge

[traversaro]

Thanks!

[S-Dafarra]

Sorry for the late comment. Was there a reason not to use an std::iterator? I am not a great fan of using a pointer to double as iterator 🤔

[traversaro]

Sorry for the late comment. Was there a reason not to use an std::iterator? I am not a great fan of using a pointer to double as iterator 🤔

What do you think @GiulioRomualdi ? Until we actually release iDynTree 1.1, we can change the interface any time we want, so feel free to propose a change if you see fit.

[GiulioRomualdi]

std::iterator has been deprecated in C++17, that's the reason why I didn't use it. I don't know which is the alternative :sad:

[S-Dafarra]

To be honest I did not know. Apparently the alternative is simply to define your own iterator without inheriting from std::iterator, but defining all the aliases (https://www.fluentcpp.com/2018/05/08/std-iterator-deprecated/).