I have to get list of all meshes (windows/loops/elementary circuits, shortest cycles that together cover all edges of graph and no one contains other cycles) in unweighted graph representing electrical circuit for mesh analysis of that circuit (I can assume it's a planar graph). The graph (represented as tuples (A,B,R) meaning two vertices and resistance of edge) is loaded from file.
I'm using Python with NetworkX library, but it's cycle_basis function doesn't return meshes (they are apparently not the same as cycle basis). The graph is undirected, so I can't use simple_cycles function. I tried to modify BFS for this task, but I couldn't guarantee that no cycle is contained in other cycle.
I probably need something like this: Algorithm for finding minimal cycles in a graph, but this question doesn't have any proof for algorithm in last comment, and I hope the answer to my problem is simpler than "implement Horton's algorithm".