#include #include typedef struct { double x, y; } Position, Velocity; // Forward declare components so they can be accessed from systems ECS_COMPONENT_DECLARE(Position); ECS_COMPONENT_DECLARE(Velocity); void SetVelocity(ecs_iter_t *it) { for (int i = 0; i < it->count; i ++) { ecs_set(it->world, it->entities[i], Velocity, {1, 2}); } } void Move(ecs_iter_t *it) { Position *p = ecs_field(it, Position, 0); const Velocity *v = ecs_field(it, const Velocity, 1); for (int i = 0; i < it->count; i ++) { p[i].x += v[i].x; p[i].y += v[i].y; } } void PrintPosition(ecs_iter_t *it) { const Position *p = ecs_field(it, const Position, 0); for (int i = 0; i < it->count; i ++) { printf("%s: {%f, %f}\n", ecs_get_name(it->world, it->entities[i]), p[i].x, p[i].y); } } int main(int argc, char *argv[]) { ecs_world_t *ecs = ecs_init_w_args(argc, argv); ECS_COMPONENT_DEFINE(ecs, Position); ECS_COMPONENT_DEFINE(ecs, Velocity); // System that sets velocity using set for entities with Position. // While systems are progressing, operations like set are deferred until // it is safe to merge. By default this merge happens at the end of the // frame, but we can annotate systems to give the scheduler more information // about what it's doing, which allows it to insert sync points earlier. // // The parentheses after Velocity indicate that the component is not used // to match entities, while [out] indicates that it is written. A subsequent // system that accesses the component will cause the scheduler to insert a // sync point. // // Note that sync points are never necessary/inserted for systems that write // components provided by their signature, as these writes directly happen // in the ECS storage and are never deferred. // // The [none] annotation for Position tells the scheduler that while we // want to match entities with Position, we're not interested in reading or // writing the component value. ECS_SYSTEM(ecs, SetVelocity, EcsOnUpdate, [none] Position, [out] Velocity()); // This system reads Velocity, which causes the insertion of a sync point. ECS_SYSTEM(ecs, Move, EcsOnUpdate, Position, [in] Velocity); // Print resulting Position ECS_SYSTEM(ecs, PrintPosition, EcsPostUpdate, [in] Position); // Create a few test entities for a Position, Velocity query ecs_entity_t e1 = ecs_entity(ecs, { .name = "e1" }); ecs_set(ecs, e1, Position, {10, 20}); ecs_set(ecs, e1, Velocity, {1, 2}); ecs_entity_t e2 = ecs_entity(ecs, { .name = "e2" }); ecs_set(ecs, e2, Position, {10, 20}); ecs_set(ecs, e2, Velocity, {3, 4}); // Run systems. Debug logging enables us to see the generated schedule ecs_log_set_level(1); ecs_progress(ecs, 0.0f); ecs_log_set_level(-1); // Restore so we don't get world cleanup logs // Output: // info: pipeline rebuild: // info: | schedule: threading: 0, staging: 1: // info: | | system SetVelocity // info: | | merge // info: | schedule: threading: 0, staging: 1: // info: | | system Move // info: | | system PrintPosition // info: | | merge // e1: {11.000000, 22.000000} // e2: {11.000000, 22.000000} // // The "merge" lines indicate sync points. // // Removing the '[out] Velocity()' annotation from the system will remove // the first sync point from the schedule. // // To create the same system with ecs_system_init, do: // ecs_system_init(ecs, &(ecs_system_desc_t){ // .query.terms = { // { // .id = ecs_id(Position), // .inout = EcsInOutNone // }, // { // .id = ecs_id(Velocity), // .inout = EcsOut, // .src.id = EcsIsEntity // } // }, // .entity = { // .name = "SetVelocity", // .add = ecs_ids(ecs_dependson(EcsOnUpdate)) // }, // .callback = SetVelocity // }); return ecs_fini(ecs); }