Decomposition of the execute_best_transactions method

[therustmonk]

📝 Summary

I've prepared a fairly significant structural revision: I turned execute_best_transactions into a BestTransactionsExecutor struct, and instead of a scoped method, we now have access to all contextual data between method calls of this struct.

💡 Motivation and Context

I'm pursuing several important goals:

• Reduce the execution method size (split it into parts) to make it more readable

• Turning the build process into an explicit finite state machine (FSM)

  • easier to debug
  • allows changing or combining steps depending on configuration
  • potential for increased performance by skipping groups of calls under certain conditions

• Additionally, BestTransactionsExecutor can be turned into an Iterator 🔥

• Since the decomposition is based on method calls, it's:

  • faster than an FSM built on dynamic states and traits
  • easier to maintain than combinators

• This change will enable deeper decomposition of the build process and simplify runtime control, or conversely, the emission of useful data at any stage of the build, for example: we could create an iterator that yields events we can respond to - sending various data or metrics, interrupting the build, etc.

---

✅ I have completed the following steps:

• Run make lint
• Run make test
• Added tests (if applicable)