import Graph, { GraphData } from "../dataStructures/graph"
import edmondsKarp from "../maximumFlowAlgorithms/edmondsKarp"
import bellmanFord from "../shortestPathAlgorithms/bellmanFord"
import { getResidualCapacity, sendFlow, getResidualGraph, getOptimalGraph } from "../utils"

/**
 * The Cycle-Canceling algorithm solves the minimum-cost flow problem
 * starting from a feasible graph obtained with a maximum flow algorithm
 * and, as long as negative cost cycles exist, augment the flow along
 * this cycles. This implementation uses the Edmonds-Karp algorithm to
 * solve the maximum flow problem in O(n * m^2), and the Bellman-Fort
 * algorithm to find the negative cycles in O(m * n).
 * Time complexity: O (n * m^2 * C * U).
 * @param {Graph | GraphData} The graph to visit.
 * @returns {[Graph, number, number]} The optimal graph, the maximum flow and the minimum cost.
 */
export default function cycleCanceling(graph: Graph | GraphData): [Graph, number, number] {
    if (!(graph instanceof Graph)) {
        graph = new Graph(graph)
    }

    const [optimalGraph, maximumFlow, , sinkNodeId] = edmondsKarp(graph)
    const residualGraph = getResidualGraph(optimalGraph)

    let negativeCycle = bellmanFord(residualGraph, sinkNodeId)

    while (negativeCycle && Array.isArray(negativeCycle)) {
        negativeCycle.push(negativeCycle[0])

        const residualCapacity = getResidualCapacity(residualGraph, negativeCycle)

        sendFlow(residualGraph, negativeCycle, residualCapacity)

        negativeCycle = bellmanFord(residualGraph, sinkNodeId)
    }

    // Calculates the minimum cost.
    let minimumCost = 0
    for (const node of residualGraph.getNodes()) {
        for (const arc of node.getArcs()) {
            if (arc.cost < 0) {
                minimumCost -= arc.cost
            }
        }
    }

    return [getOptimalGraph(residualGraph), maximumFlow, minimumCost]
}