Merge pull request #299 from pangenome/id-increment-fix

Id increment fix

src/gfa_to_handle.cpp

@@ -26,6 +26,7 @@ std::map<char, uint64_t> gfa_line_counts(const char* filename) {
 
 void gfa_to_handle(const string& gfa_filename,
                    handlegraph::MutablePathMutableHandleGraph* graph,
+                   bool compact_ids,
                    uint64_t n_threads,
                    bool progress) {
 
@@ -52,12 +53,10 @@ void gfa_to_handle(const string& gfa_filename,
         line_counts = gfa_line_counts(filename);
         x.join();
     }
-    uint64_t id_increment = min_id - 1;
+    uint64_t id_increment = (compact_ids ? min_id - 1 : 0);
     uint64_t node_count = line_counts['S'];
     uint64_t edge_count = line_counts['L'];
     uint64_t path_count = line_counts['P'];
-    // set the min id as an offset
-    graph->set_id_increment(id_increment);
     // build the nodes
     {
         std::unique_ptr<algorithms::progress_meter::ProgressMeter> progress_meter;
@@ -87,8 +86,8 @@ void gfa_to_handle(const string& gfa_filename,
             filename,
             [&](gfak::edge_elem e) {
                 if (e.source_name.empty()) return;
-                handlegraph::handle_t a = graph->get_handle(stol(e.source_name), !e.source_orientation_forward);
-                handlegraph::handle_t b = graph->get_handle(stol(e.sink_name), !e.sink_orientation_forward);
+                handlegraph::handle_t a = graph->get_handle(stol(e.source_name) - id_increment, !e.source_orientation_forward);
+                handlegraph::handle_t b = graph->get_handle(stol(e.sink_name) - id_increment, !e.sink_orientation_forward);
                 graph->create_edge(a, b);
                 if (progress) progress_meter->increment(1);
             });
@@ -115,7 +114,7 @@ void gfa_to_handle(const string& gfa_filename,
                         uint64_t i = 0;
                         for (auto& s : p->gfak.segment_names) {
                             graph->append_step(p->path,
-                                               graph->get_handle(std::stoi(s),
+                                               graph->get_handle(std::stoi(s) - id_increment,
                                                                  // in gfak, true == +
                                                                  !p->gfak.orientations[i++]));
                         }
@@ -153,5 +152,10 @@ void gfa_to_handle(const string& gfa_filename,
             progress_meter->finish();
         }
     }
+
+    if (compact_ids) {
+        graph->optimize();
+    }
+
 }
 }

src/gfa_to_handle.hpp

@@ -33,7 +33,8 @@ std::map<char, uint64_t> gfa_line_counts(const char* filename);
 /// Handle graph must be empty when passed into function.
 void gfa_to_handle(const string& gfa_filename,
                    handlegraph::MutablePathMutableHandleGraph* graph,
-                   uint64_t n_threads = 1,
-                   bool show_progress = false);
+                   bool compact_ids,
+                   uint64_t n_threads,
+                   bool show_progress);
 
 }

src/subcommand/build_main.cpp

@@ -29,6 +29,7 @@ int main_build(int argc, char** argv) {
     args::Group graph_files_io(parser, "[ Graph Files IO ]");
     args::Flag to_gfa(graph_files_io, "to_gfa", "Write the graph to stdout in GFAv1 format.", {'G', "to-gfa"});
     args::Group graph_sorting(parser, "[ Graph Sorting ]");
+    args::Flag optimize(graph_sorting, "optimize", "Compact the graph id space into a dense integer range.", {'O', "optimize"});
     args::Flag toposort(graph_sorting, "sort", "Apply a general topological sort to the graph and order the node ids"
                                         "  accordingly. A bidirected adaptation of Kahn’s topological sort (1962)"
                                         "  is used, which can handle components with no heads or tails. Here, both heads and tails are taken into account.", {'s', "sort"});
@@ -74,7 +75,7 @@ int main_build(int argc, char** argv) {
     	return 1;
     }
     if (gfa_filename.size()) {
-        gfa_to_handle(gfa_filename, &graph, args::get(nthreads), args::get(progress));
+        gfa_to_handle(gfa_filename, &graph, args::get(optimize), args::get(nthreads), args::get(progress));
     }
 
     const uint64_t num_threads = args::get(nthreads) ? args::get(nthreads) : 1;

src/utils.cpp

@@ -72,7 +72,7 @@ namespace utils {
 																				   "To save time in the future, please use odgi build -i=[FILE], --idx=[FILE] -o=[FILE], --out=[FILE] "
 																				   "to generate a graph in ODGI format. Such a graph can be supplied to all ODGI subcommands. Building graph in ODGI format from given GFA." << std::endl;
 			}
-			gfa_to_handle(infile, &graph, num_threads, progress);
+			gfa_to_handle(infile, &graph, false, num_threads, progress);
 			graph.set_number_of_threads(num_threads);
 		} else {
 			ifstream f(infile.c_str());