/*

   igraph library.

   Copyright (C) 2024  The igraph development team

   This program is free software; you can redistribute it and/or modify

   it under the terms of the GNU General Public License as published by

   the Free Software Foundation; either version 2 of the License, or

   (at your option) any later version.

   This program is distributed in the hope that it will be useful,

   but WITHOUT ANY WARRANTY; without even the implied warranty of

   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License

   along with this program; if not, write to the Free Software

   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA

   02110-1301 USA

*/

#include <igraph.h>

#include <cstdlib>

extern "C" int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size) {

    igraph_t graph;

    igraph_vector_int_t edges;

    igraph_vector_t weights;

    igraph_set_warning_handler(igraph_warning_handler_ignore);

    if (Size % 3 == 0 || Size % 3 == 2 || Size > (3 * 256) + 1 || Size < 1) {

        return 0;

    }

    igraph_vector_int_init(&edges, ((Size-1) / 3) * 2);

    igraph_vector_init(&weights, (Size-1) / 3);

    for (size_t i=0; i < ((Size-1) / 3); ++i) {

        VECTOR(edges)[i * 2] = Data[i * 3 + 1];

        VECTOR(edges)[i * 2 + 1] = Data[i * 3 + 2];

        // We keep the weights strictly positive, as this is required by some algorithms.

        VECTOR(weights)[i] = ((double) Data[i * 3 + 3] + 1.0) / 105.0;

    }

    // Turn on attribute handling.

    igraph_set_attribute_table(&igraph_cattribute_table);

    igraph_rng_seed(igraph_rng_default(), 42);

    if (igraph_create(&graph, &edges, Data[0], IGRAPH_DIRECTED) == IGRAPH_SUCCESS) {

        igraph_matrix_int_t merges, im;

        igraph_matrix_t mat;

        igraph_vector_int_t membership, membership2, iv, iv2;

        igraph_vector_t mv, v;

        igraph_real_t m, r;

        igraph_int_t i;

        igraph_bool_t b;

        /* Limit graph size for the sake of performance. */

        if (igraph_vcount(&graph) <= 64) {

            igraph_attribute_combination_t comb;

            igraph_matrix_int_init(&merges, 0, 0);

            igraph_matrix_int_init(&im, 0, 0);

            igraph_matrix_init(&mat, 0, 0);

            igraph_vector_int_init(&membership, 0);

            igraph_vector_int_init(&membership2, 0);

            igraph_vector_int_init(&iv, 0);

            igraph_vector_int_init(&iv2, 0);

            igraph_vector_init(&mv, 0);

            igraph_vector_init(&v, 0);

            SETEANV(&graph, "weight", &weights);

            // Currently "strength" is used only used as a vertex attribute to exercise the

            // attribute handling code durig vertex contraction, which is convenient to do

            // using the output from community detection. It is not used as input to any

            // community detection methods.

            igraph_strength(&graph, &v, igraph_vss_all(), IGRAPH_OUT, IGRAPH_LOOPS, &weights);

            SETVANV(&graph, "strength", &v);

            igraph_attribute_combination(&comb,

                                         "weight", IGRAPH_ATTRIBUTE_COMBINE_SUM,

                                         "strength", IGRAPH_ATTRIBUTE_COMBINE_MAX,

                                         IGRAPH_NO_MORE_ATTRIBUTES);

            // Ignore edge directions in label propagation for now, as on some weighted graphs the

            // algorithm seems to never complete. See https://github.com/igraph/igraph/issues/2561

            igraph_community_label_propagation(&graph, &membership, IGRAPH_ALL, &weights, NULL, NULL, IGRAPH_LPA_FAST);

            igraph_community_label_propagation(&graph, &membership, IGRAPH_ALL, &weights, NULL, NULL, IGRAPH_LPA_RETENTION);

            igraph_community_label_propagation(&graph, &membership, IGRAPH_ALL, &weights, NULL, NULL, IGRAPH_LPA_DOMINANCE);

            igraph_community_walktrap(&graph, &weights, 3, &merges, &mv, &membership);

            igraph_community_edge_betweenness(&graph, &iv, &v, &merges, &iv2, &mv, &membership2, IGRAPH_DIRECTED, &weights, NULL);

            igraph_community_leiden(&graph, &weights, NULL, NULL, 1.5, 0.01, false, 2, &membership, &i, &r);

            // Take the opportunity to run functions that can use the output of community detection,

            // potentially with weights.

            {

                igraph_community_comparison_t method[] = {

                    IGRAPH_COMMCMP_VI,

                    IGRAPH_COMMCMP_NMI,

                    IGRAPH_COMMCMP_SPLIT_JOIN,

                    IGRAPH_COMMCMP_RAND,

                    IGRAPH_COMMCMP_ADJUSTED_RAND

                };

                for (size_t i=0; i < sizeof(method) / sizeof(method[0]); i++) {

                    if ((method[i] == IGRAPH_COMMCMP_RAND ||

                         method[i] == IGRAPH_COMMCMP_ADJUSTED_RAND) &&

                        igraph_vcount(&graph) < 2) {

                        continue;

                    }

                    igraph_compare_communities(&membership, &membership2, &r, method[i]);

                }

            }

            {

                igraph_t graph2;

                igraph_copy(&graph2, &graph);

                igraph_contract_vertices(&graph2, &membership, &comb);

                igraph_destroy(&graph2);

            }

            igraph_modularity(&graph, &membership, &weights, 1.5, IGRAPH_UNDIRECTED, &m);

            igraph_modularity_matrix(&graph, &weights, 0.75, &mat, IGRAPH_DIRECTED);

            // NOTE: Currently Infomap dominates this fuzz target due to being

            // slower and more complex than the other algorithms.

            // TODO: Currently disabled because of extremely bad performance.

            // igraph_community_infomap(&graph, &weights, NULL, 1, &membership, &r);

            igraph_simplify(&graph, true, true, &comb);

            EANV(&graph, "weight", &weights);

            // Compute 'length' vector for community_voronoi()

            igraph_vector_update(&v, &weights);

            for (igraph_int_t i=0; i < igraph_vector_size(&v); i++) {

                VECTOR(v)[i] = 1 / (1 + VECTOR(v)[i]);

            }

            igraph_community_voronoi(&graph, &membership, &iv, &m, &v, &weights, IGRAPH_OUT, 1.0);

            igraph_to_undirected(&graph, IGRAPH_TO_UNDIRECTED_COLLAPSE, &comb);

            EANV(&graph, "weight", &weights);

            igraph_community_fastgreedy(&graph, &weights, &merges, &mv, &membership);

            igraph_community_leiden(&graph, &weights, NULL, NULL, 1.5, 0.01, false, 2, &membership, &i, &r);

            igraph_community_multilevel(&graph, &weights, 0.8, &membership, &im, &mv);

            // community_spinglass() only works on connected graphs

            igraph_is_connected(&graph, &b, IGRAPH_WEAK);

            if (b) {

                igraph_community_spinglass(&graph, &weights, &r, NULL, &membership, NULL, 10, false, 1.0, 0.01, 0.99, IGRAPH_SPINCOMM_UPDATE_CONFIG, 1.0, IGRAPH_SPINCOMM_IMP_ORIG, 1.0);

            }

            DELALL(&graph);

            igraph_attribute_combination_destroy(&comb);

            igraph_vector_destroy(&v);

            igraph_vector_destroy(&mv);

            igraph_vector_int_destroy(&iv2);

            igraph_vector_int_destroy(&iv);

            igraph_vector_int_destroy(&membership2);

            igraph_vector_int_destroy(&membership);

            igraph_matrix_destroy(&mat);

            igraph_matrix_int_destroy(&im);

            igraph_matrix_int_destroy(&merges);

        }

        igraph_destroy(&graph);

    }

    igraph_vector_int_destroy(&edges);

    igraph_vector_destroy(&weights);

    IGRAPH_ASSERT(IGRAPH_FINALLY_STACK_EMPTY);

    return 0;  // Non-zero return values are reserved for future use.

}