Struct PathRanker 

pub struct PathRanker;

Stateless path-finding and scoring engine

Implementations

impl PathRanker

pub fn find_paths( graph: &KnowledgeGraph, start: &str, end: &str, max_hops: usize, ) -> Vec<KgPath>

Find all simple paths from start to end with at most max_hops edges.

Uses iterative DFS with an explicit stack to avoid stack overflow on large graphs.

pub fn score_path( graph: &KnowledgeGraph, path: &KgPath, config: &PathRankingConfig, ) -> f64

Score a single path according to config.

score = (Σ edge_weights) * weight_factor * hop_penalty^hop_count * (Π edge_confidences) * confidence_factor + (Σ node_importance) * importance_bonus

pub fn rank_paths( graph: &KnowledgeGraph, paths: Vec<KgPath>, config: &PathRankingConfig, ) -> Vec<(KgPath, f64)>

Sort paths by descending score and return (path, score) pairs.

pub fn shortest_path_dijkstra( graph: &KnowledgeGraph, start: &str, end: &str, ) -> Option<KgPath>

Find the shortest path from start to end using Dijkstra’s algorithm.

Distance metric: d = 1/weight (higher weight = shorter distance). Edges with weight ≤ 0 are treated as distance f64::INFINITY.

pub fn most_relevant_paths( graph: &KnowledgeGraph, start: &str, end: &str, max_hops: usize, top_k: usize, config: &PathRankingConfig, ) -> Vec<(KgPath, f64)>

Find all paths up to max_hops, rank them, and return the top-k.