Struct MemoryConsolidator 

pub struct MemoryConsolidator {
    pub config: ConsolidationConfig,
}

The memory consolidation engine — responsible for keeping a fighter’s memory store lean and battle-ready.

Like a seasoned corner crew between rounds, the consolidator trims the fat, reinforces core muscle memory, and ensures the fighter enters the next bout at peak mental sharpness.

Fields

config: ConsolidationConfig

The consolidation training regimen.

Implementations

impl MemoryConsolidator

pub fn new(config: ConsolidationConfig) -> Self

Create a new consolidator with the given configuration.

pub fn with_defaults() -> Self

Create a consolidator with sensible default settings — a balanced training regimen suitable for most fighters.

pub async fn consolidate( &self, memory: &MemorySubstrate, fighter_id: &FighterId, ) -> PunchResult<ConsolidationResult>

Run full memory consolidation for a fighter — the complete recovery session between bouts.

Steps:

1. Apply confidence decay based on memory age
2. Prune memories below the minimum confidence threshold
3. Merge similar memories (by key similarity)
4. If still over max capacity, prune lowest-confidence oldest memories

pub fn apply_decay(&self, confidence: f64, age_days: f64) -> f64

Compute decayed confidence based on age.

Muscle memory fades without practice — confidence erodes over time at the configured decay rate: confidence * (1.0 - decay_rate) ^ age_days.

pub fn should_consolidate(&self, memory_count: usize) -> bool

Check whether a fighter’s memory store needs consolidation — is the fighter carrying too much weight for their class?

pub fn keys_are_similar(a: &str, b: &str) -> bool

Check if two memory keys are similar enough to merge.

Uses a normalized edit distance (Levenshtein-like). Two keys are considered similar if their normalized similarity score exceeds the configured threshold — like recognizing two punches as variations of the same combo.

pub fn merge_values(values: &[(&str, f64)]) -> (String, f64)

Merge multiple memory values into a single consolidated entry.

Like combining footage from multiple training sessions: the best performance (highest confidence value) is kept, and the overall confidence is averaged across all observations.

Trait Implementations

impl Clone for MemoryConsolidator

fn clone(&self) -> MemoryConsolidator

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MemoryConsolidator

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.