[ci] Shift to clang-format-10 to resolve bug in clang-format-{7-9}

[kunaltyagi]

clang-format-8 has a bug which causes some incorrect formatting. Shifting to v10 would resolve that.

Also modified is the cpp standard in the .clang-format. Should that be Cpp14 instead of Latest?

[kunaltyagi]

This PR requires a sync between this and all other open PR which are dealing with:

• increasing code-coverage of formatter: [tracking] format module by clang-format #3814, Add compatibility note for clang-format and fix a small issue in .clang-format #3877, [apps] Apply clang-format to 3d_rec_framework #3880, [apps] Apply clang-format to modeler #3884, [apps] Apply clang-format to ungrouped apps #3885
• ml or octree modules

[diivm]

Any reason why we are shifting to v10 as opposed to v11?

[kunaltyagi]

Latest version available by default in Ubuntu 20.04 LTS. The diff is similar and migration to 11 later (when 20.04 or 20.10 or smthg adds it) will be easy.

[taketwo]

I'm going to push a new image to DockerHub and then trigger re-run here.

[SunBlack]

@divmadan v11 is not published yet, so it would be not the best idea to switch to an unstable version ;-)

Another question: Why Standard: Latest and not Standard: c++14, as we don't support C++17 or later?

[taketwo]

Another question: Why Standard: Latest and not Standard: c++14, as we don't support C++17 or later?

Fair point.

Sorry for hastily merging this without enough time for a proper discussion. But I needed to have this in master to unblock other PRs since the fmt Docker image was already updated and formatting check started to fail without this PR.

[diivm]

Just to add a point, Latest and Cpp11 gave similar diffs for our whitelisted directories.

[kunaltyagi]

Should we have a PR for Cpp14 as the standard? Or is sticking to "Latest" for formatting purpose ok?

[taketwo]

Probably? Our code is C++14 code, not C++20, so why confusing the formatter?