Double unwrap for `program_base` in Cairo runner `initialize_state`

[Thomas Coratger]

Issue Description

In the initialize_state function of the CairoRunner structure, there is redundant code in my opinion, that can be improved for clarity and efficiency.

At the beginning of the function, there is a conditional check (if let Some(prog_base) = self.program_base) to ensure that self.program_base is not None.

cairo-vm/vm/src/vm/runners/cairo_runner.rs

Lines 414 to 431 in 4ea224a

 if let Some(prog_base) = self.program_base { 
     let initial_pc = Relocatable { 
         segment_index: prog_base.segment_index, 
         offset: prog_base.offset + entrypoint, 
     }; 
     self.initial_pc = Some(initial_pc); 
     vm.segments 
         .load_data(prog_base, &self.program.shared_program_data.data) 
         .map_err(RunnerError::MemoryInitializationError)?; 
  
     // Mark all addresses from the program segment as accessed 
     let base = self 
         .program_base 
         .unwrap_or_else(|| Relocatable::from((0, 0))); 
     for i in 0..self.program.shared_program_data.data.len() { 
         vm.segments.memory.mark_as_accessed((base + i)?); 
     } 
 } 

However, a few lines later, the unwrap method is used again on self.program_base to obtain base, and in case of None, it sets base to Relocatable::from((0, 0)).

cairo-vm/vm/src/vm/runners/cairo_runner.rs

Lines 425 to 427 in 4ea224a

 let base = self 
     .program_base 
     .unwrap_or_else(|| Relocatable::from((0, 0))); 

This redundant check and assignment can be eliminated because the None case will never happen inside the if statement, making the base declaration unnecessary. Instead, we can directly use the prog_base variable in the loop to mark memory segments as accessed.