feat: add Network Flow Analyzer (Edmonds-Karp max flow)

Implements the Edmonds-Karp algorithm (BFS-based Ford-Fulkerson) for
computing maximum flow between any two vertices in the graph.

Features:
- compute(source, sink): returns max flow value using shortest
  augmenting paths (O(V*E^2) time complexity)
- Treats undirected edges as bidirectional arcs with capacity
  equal to edge weight (defaults to 1.0 for unweighted edges)
- getEdgeFlows(): per-edge flow assignments with net direction
- getFlowOnEdge(v1, v2): flow on specific edge
- getMinCut(): edges forming the minimum cut (max-flow min-cut
  theorem), via BFS reachability in residual graph
- getSourceSide(): vertices reachable from source in residual
- getBottleneckEdges(): fully saturated edges (flow == capacity)
- getTotalCapacity(): sum of all edge capacities
- getUtilisation(): percentage of source capacity used
- decomposeFlowPaths(): decompose max flow into individual
  source-to-sink paths with flow values
- getAugmentingPathCount(): number of flow paths found
- FlowResult: immutable result snapshot with all analytics
- getSummary(): formatted multi-line report with flow paths
  and edge utilisation (marks saturated edges)

Applications: transportation capacity, communication bandwidth,
bipartite matching, supply chain optimization.

55 new tests covering: constructor validation, input validation,
simple/diamond/triangle/parallel/star/chain/complete graphs,
disconnected vertices, default weights, edge flows, min cut,
source side, bottleneck detection, capacity, utilisation,
flow decomposition, recomputation, flow conservation.

Author: sauravbhattacharya001