Fix clang-tidy checks, including some use-after-move

[esseivaju]

Fix a few clang-tidy checks. There were a few use-after-move in LocalTransporter.cc, Solid.cc and ValueGridBuilder.cc that could lead to bugs. In addition, ignore some checks making a lot of noise. The remaining warnings are mostly false alerts and could use a // NOLINT comment explaining why the code is fine.

[sethrj]

Why is this override? It's not like you can declare a destructor "final"...

[esseivaju]

To document/verify that the base class has a virtual dtor. This is following the discussion isocpp/CppCoreGuidelines#1448 virtual instead of override is also fine.

[sethrj]

That's weird. The parent class doesn't have a ~GeantLoggerAdapter method. Maybe next c++ version will name the destructor __del__ so you can actually tell it's overriding... I think I prefer virtual here since it says "polymorphic and safe to destruct" rather than "I'm replacing what the parent classdoes".

[esseivaju]

The parent G4coutDestination still has a virtual destructor. Even though technically destructor override their parent's I agree that semantically virtual makes sense since they are chained and don't replace the parent's behavior. However, the advantage of override is that it will result in an error if the parent dtor is no longer virtual. Using virtual would silently redeclare it virtual in the derived which would be confusing.

Anyway, in that specific case, it's not really important since this is a TU-local class but in general I prefer the destructor be declared override if it needs to be defined again in the derived class for the reason above.

Maybe for that case we could declare the class final? It's local anyway and would silence the warning

[sethrj]

Declaring as final sounds good to me, and now that I know about this behavior I think override for virtual destructors is sensible. (It's partly weird because the daughter classes will automatically define virtual override destructors...)

[sethrj]

Thanks @esseivaju ! Looks like there's only 20 warnings left, after those are gone we can make "warning" fatal 😄

[esseivaju]

Even though some checks are not showing up on the CI, they'll show up in my environment. If we don't want these checks I suppose we can keep them disabled for future-proofing when we switch to ubuntu-24.04 runners.