Add flag to control asset versioning when getting write access

[roby2014]

Description

Added a flag to Assets::write in order to control whether the version of assets should be incremented when getting write access.

Checklist

• Self-review changes.
• Evaluate impact on the documentation.
• Write new samples.

[github-actions]

PR Preview Action v1.4.4
🚀 Deployed preview to https://GameDevTecnico.github.io/cubos/pr-preview/pr-727/
on branch gh-pages at 2023-10-17 01:26 UTC

[codecov]

Codecov Report

All modified lines are covered by tests ✅

Comparison is base (174a787) 38.15% compared to head (69ead5e) 38.15%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #727   +/-   ##
=======================================
  Coverage   38.15%   38.15%           
=======================================
  Files          99       99           
  Lines        6576     6576           
=======================================
  Hits         2509     2509           
  Misses       4067     4067           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[RiscadoA]

This PR should also add a method used to bump the version manually, but I'm not sure how we would do this. Maybe we should add a third locking function, instead of adding an argument to the existing one.

E.g.:

auto assetWrite = assets->potentialWrite(asset);
/* ... potentially do changes, just like before */
if (modified)
{
    // consumes the assetWrite lock, unlocking the asset, but before that, increments the asset's version.
    assets->confirmWrite(std::move(assetWrite));
}

The thing is, currently AssetWrite is just an alias for core::memory::WriteGuard<T, std::unique_lock<std::shared_mutex>>, and for this to work, it would also need to store the UUID of the asset being accessed. Otherwise, we would need to add an asset handle argument to the confirmWrite method, which is duplicating information and thus error prone.

Maybe instead of making it an alias, we could make it a child class of the WriteGuard and then just store the extra UUID field, which could then be used in the confirmWrite implementation.

[RiscadoA]

Forgot to request changes!