queries/group_by

• C

  View on GitHub

  #include <group_by.h>
  #include <stdio.h>
   
  // Group by is a feature of cached queries that allows applications to assign a
  // group id to each matched table. Tables that are assigned the same group id
  // are stored together in "groups". This ensures that when a query is iterated,
  // tables that share a group are iterated together.
  //
  // Groups in the cache are ordered by group id, which ensures that tables with
  // lower ids are iterated before table with higher ids. This is the same 
  // mechanism that is used by the cascade feature, which groups tables by depth
  // in a relationship hierarchy.
  //
  // This makes groups a more efficient, though less granular mechanism for 
  // ordering entities. Order is maintained at the group level, which means that
  // once a group is created, tables can get added and removed to the group
  // with is an O(1) operation.
  //
  // Groups can also be used as an efficient filtering mechanism. See the 
  // group_iter example for more details.
   
  typedef struct {
      double x, y;
  } Position;
   
  int main(int argc, char *argv[]) {
      ecs_world_t *ecs = ecs_init_w_args(argc, argv);
   
      ECS_COMPONENT(ecs, Position);
      ECS_TAG(ecs, Tag); // Dummy tag to put entities in different tables
   
      // Create a relationship to use for the group_by function. Tables will
      // be assigned the relationship target as group id
      ECS_TAG(ecs, Group);
   
      // Targets for the relationship, which will be used as group ids.
      ECS_TAG(ecs, First);
      ECS_TAG(ecs, Second);
      ECS_TAG(ecs, Third);
   
      // Grouped query
      ecs_query_t *q = ecs_query(ecs, {
          .terms = {
              { .id = ecs_id(Position) }
          },
          // Group tables by the relationship target of the "Group" relationship.
          // A custom group_by function can be provided to derive a group id from
          // a table, see the group_by_custom example for more details.
          .group_by = Group
      });
   
      // Create entities in 6 different tables with 3 group ids
      ecs_entity_t e1 = ecs_new_w_pair(ecs, Group, Third);
      ecs_set(ecs, e1, Position, {1, 1});
   
      ecs_entity_t e2 = ecs_new_w_pair(ecs, Group, Second);
      ecs_set(ecs, e2, Position, {2, 2});
   
      ecs_entity_t e3 = ecs_new_w_pair(ecs, Group, First);
      ecs_set(ecs, e3, Position, {3, 3});
   
      ecs_entity_t e4 = ecs_new_w_pair(ecs, Group, Third);
      ecs_set(ecs, e4, Position, {4, 4});
      ecs_add(ecs, e4, Tag);
   
      ecs_entity_t e5 = ecs_new_w_pair(ecs, Group, Second);
      ecs_set(ecs, e5, Position, {5, 5});
      ecs_add(ecs, e5, Tag);
   
      ecs_entity_t e6 = ecs_new_w_pair(ecs, Group, First);
      ecs_set(ecs, e6, Position, {6, 6});
      ecs_add(ecs, e6, Tag);
   
      // The query cache now looks like this:
      //  - group First:
      //     - table [Position, (Group, First)]
      //     - table [Position, Tag, (Group, First)]
      //
      //  - group Second:
      //     - table [Position, (Group, Second)]
      //     - table [Position, Tag, (Group, Second)]
      //  
      //  - group Third:
      //     - table [Position, (Group, Third)]
      //     - table [Position, Tag, (Group, Third)]
      //
   
      // Iterate query, print position & table components
      ecs_iter_t it = ecs_query_iter(ecs, q);
      while (ecs_query_next(&it)) {
          Position *p = ecs_field(&it, Position, 0);
          char *table_str = ecs_table_str(ecs, it.table);
          char *group_str = ecs_get_path(ecs, ecs_iter_get_group(&it));
   
          printf(" - group %s: table [%s]\n", group_str, table_str);
   
          for (int i = 0; i < it.count; i ++) {
              printf("     {%f, %f}\n", p[i].x, p[i].y);
          }
   
          ecs_os_free(group_str);
          ecs_os_free(table_str);
          printf("\n");
      }
   
      // Output:
      //  - group First: table [Position, (Group,First)]
      //      {3.000000, 3.000000}
      // 
      //  - group First: table [Position, Tag, (Group,First)]
      //      {6.000000, 6.000000}
      // 
      //  - group Second: table [Position, (Group,Second)]
      //      {2.000000, 2.000000}
      // 
      //  - group Second: table [Position, Tag, (Group,Second)]
      //      {5.000000, 5.000000}
      // 
      //  - group Third: table [Position, (Group,Third)]
      //      {1.000000, 1.000000}
      // 
      //  - group Third: table [Position, Tag, (Group,Third)]
      //      {4.000000, 4.000000}
   
      return ecs_fini(ecs);
  }

• C++

  View on GitHub

  #include <group_by.h>
  #include <iostream>
   
  // Group by is a feature of cached queries that allows applications to assign a
  // group id to each matched table. Tables that are assigned the same group id
  // are stored together in "groups". This ensures that when a query is iterated,
  // tables that share a group are iterated together.
  //
  // Groups in the cache are ordered by group id, which ensures that tables with
  // lower ids are iterated before table with higher ids. This is the same 
  // mechanism that is used by the cascade feature, which groups tables by depth
  // in a relationship hierarchy.
  //
  // This makes groups a more efficient, though less granular mechanism for 
  // ordering entities. Order is maintained at the group level, which means that
  // once a group is created, tables can get added and removed to the group
  // with is an O(1) operation.
  //
  // Groups can also be used as an efficient filtering mechanism. See the 
  // set_group example for more details.
   
  struct Position {
      double x, y;
  };
   
  // Dummy tag to put entities in different tables
  struct Tag { };
   
  // Create a relationship to use for the group_by function. Tables will
  // be assigned the relationship target as group id
  struct Group { };
   
  // Targets for the relationship, which will be used as group ids.
  struct First { };
  struct Second { };
  struct Third { };
   
  int main() {
      flecs::world ecs;
   
      // Register components in order so that id for First is lower than Third
      ecs.component<First>();
      ecs.component<Second>();
      ecs.component<Third>();
   
      // Grouped query
      flecs::query<Position> q = ecs.query_builder<Position>()
          .group_by<Group>()
          .build();
   
      // Create entities in 6 different tables with 3 group ids
      ecs.entity().add<Group, Third>()
          .set<Position>({1, 1});
      ecs.entity().add<Group, Second>()
          .set<Position>({2, 2});
      ecs.entity().add<Group, First>()
          .set<Position>({3, 3});
   
      ecs.entity().add<Group, Third>()
          .set<Position>({4, 4})
          .add<Tag>();
      ecs.entity().add<Group, Second>()
          .set<Position>({5, 5})
          .add<Tag>();
      ecs.entity().add<Group, First>()
          .set<Position>({6, 6})
          .add<Tag>();
   
      // The query cache now looks like this:
      //  - group First:
      //     - table [Position, (Group, First)]
      //     - table [Position, Tag, (Group, First)]
      //
      //  - group Second:
      //     - table [Position, (Group, Second)]
      //     - table [Position, Tag, (Group, Second)]
      //  
      //  - group Third:
      //     - table [Position, (Group, Third)]
      //     - table [Position, Tag, (Group, Third)]
      //
   
      q.run([&](flecs::iter& it) {
          while (it.next()) {
              auto p = it.field<Position>(0);
   
              flecs::entity group = ecs.entity(it.group_id());
              std::cout << " - group " << group.path() << ": "
                        << "table [" << it.table().str() << "]\n";
   
              for (auto i : it) {
                  std::cout << "     {" << p[i].x << ", " << p[i].y << "}\n";
              }
   
              std::cout << "\n";
          }
      });
   
      // Output:
      //  - group ::First: table [Position, (Group,First)]
      //      {3, 3}
      // 
      //  - group ::First: table [Position, Tag, (Group,First)]
      //      {6, 6}
      // 
      //  - group ::Second: table [Position, (Group,Second)]
      //      {2, 2}
      // 
      //  - group ::Second: table [Position, Tag, (Group,Second)]
      //      {5, 5}
      // 
      //  - group ::Third: table [Position, (Group,Third)]
      //      {1, 1}
      // 
      //  - group ::Third: table [Position, Tag, (Group,Third)]
      //      {4, 4}
  }