Relationships

The relationships feature makes it possible to describe entity graphs natively in ECS. Graphs are created by adding and removing relationships from one entity to another entity. See this blog for an introduction to entity relationships.

Adding/removing relationships is similar to adding/removing regular components, with as difference that instead of a single component id, a relationship adds a pair of two things to an entity. In this pair, the first element represents the relationship (e.g. "Eats"), and the second element represents the relationship target (e.g. "Apples").

Relationships can be used to describe many things, from hierarchies to inventory systems to trade relationships between players in a game. The following sections go over how to use relationships, and what features they support.

Definitions

Name	Description
Id	An id that can be added and removed
Component	Id with a single element (same as an entity id)
Pair	Id with two elements
Tag	Component or pair not associated with data
Relationship	Used to refer to first element of pair
Target	Used to refer to second element of pair
Source	Entity to which an id is added

Examples

Make sure to check out the code examples in the repository:

• relationships (C)
• relationships (C++)

Introduction

The following code is a simple example that uses relationships:

ecs_entity_t Likes = ecs_new_id(world);
ecs_entity_t Bob = ecs_new_id(world);
ecs_entity_t Alice = ecs_new_id(world);
 
// Bob Likes Alice
ecs_add_pair(world, Bob, Likes, Alice);
 
// Bob Likes Alice no more
ecs_remove_pair(world, Bob, Likes, Alice);

auto Likes = world.entity();
auto Bob = world.entity();
auto Alice = world.entity();
 
// Bob Likes Alice
Bob.add(Likes, Alice);
 
// Bob Likes Alice no more
Bob.remove(Likes, Alice);

In this example, we refer to Bob as the "source", Likes as the "relationship" and Alice as the "target". A relationship when combined with an target is called a "relationship pair".

The same relationship can be added multiple times to an entity, as long as its target is different:

ecs_entity_t Bob = ecs_new_id(world);
ecs_entity_t Eats = ecs_new_id(world);
ecs_entity_t Apples = ecs_new_id(world);
ecs_entity_t Pears = ecs_new_id(world);
 
ecs_add_pair(world, Bob, Eats, Apples);
ecs_add_pair(world, Bob, Eats, Pears);
 
ecs_has_pair(world, Bob, Eats, Apples); // true
ecs_has_pair(world, Bob, Eats, Pears); // true

auto Bob = world.entity();
auto Eats = world.entity();
auto Apples = world.entity();
auto Pears = world.entity();
 
Bob.add(Eats, Apples);
Bob.add(Eats, Pears);
 
Bob.has(Eats, Apples); // true
Bob.has(Eats, Pears); // true

An application can query for relationships with the (Relationship, Target) notation:

// Find all entities that eat apples
ecs_query_t *q = ecs_query_new(world, "(Eats, Apples)");
 
// Find all entities that eat anything
ecs_query_t *q = ecs_query_new(world, "(Eats, *)");
 
// Or with the ecs_query_init function:
ecs_query_t *q = ecs_query_init(world, &(ecs_query_desc_t){
    .filter.terms = {{ecs_pair(Eats, Apples)}}
});

// Find all entities that eat apples
auto q = world.query("(Eats, Apples)");
 
// Find all entities that eat anything
auto q = world.query("(Eats, *)");
 
// With the query builder API:
auto q = world.query_builder<>()
  .term(Eats, Apples)
  .build();
 
// Or when using pair types, when both relationship & target are compile time types:
auto q = world.query<flecs::pair<Eats, Apples>>();

This example just shows a simple relationship query. Relationship queries are much more powerful than this as they provide the ability to match against entity graphs of arbitrary size. For more information on relationship queries see the query manual.

Relationship queries

There are a number of ways an application can query for relationships. The following kinds of queries are available for all (unidirectional) relationships, and are all constant time:

Test if entity has a relationship pair

ecs_has_pair(world, Bob, Eats, Apples);

Bob.has(Eats, Apples);

Test if entity has a relationship wildcard

ecs_has_pair(world, Bob, Eats, EcsWildcard);

Bob.has(Eats, flecs::Wildcard);

Get parent for entity

ecs_entity_t parent = ecs_get_parent(world, Bob);

flecs::entity parent = Bob.parent();

Find first target of a relationship for entity

ecs_entity_t food = ecs_get_target(world, Bob, Eats, 0);

flecs::entity food = Bob.target(Eats);

Find all targets of a relationship for entity

int32_t index = 0;
while ((food = ecs_get_target(world, Bob, Eats, index ++))) {
  // ...
}

int32_t index = 0;
while ((food = Bob.target(Eats, index ++))) {
  // ...
}

Find target of a relationship with component

ecs_entity_t parent = ecs_get_target_for(world, Bob, EcsChildOf, Position);

flecs::entity parent = Bob.target_for<Position>(flecs::ChildOf);

Iterate all pairs for entity

const ecs_type_t *type = ecs_get_type(world, Bob);
for (int i = 0; i < type->count; i ++) {
  ecs_id_t id = type->array[i];
  if (ECS_IS_PAIR(id)) {
    ecs_entity_t first = ecs_pair_first(world, id);
    ecs_entity_t second = ecs_pair_second(world, id);
  }
}

Bob.each([](flecs::id id) {
  if (id.is_pair()) {
    flecs::entity first = id.pair().first();
    flecs::entity second = id.pair().second();
  }
});

Find all entities with a pair

ecs_filter_t *f = ecs_filter(world, {
  .terms[0] = ecs_pair(Eats, Apples)
});
 
ecs_iter_t it = ecs_filter_iter(world, f);
while (ecs_filter_next(&it)) {
  for (int i = 0; i < it.count; i ++) {
    // Iterate as usual
  }
}
 
ecs_filter_fini(f);

world.filter_builder()
  .term(Eats, Apples)
  .build()
  .each([](flecs::entity e) {
    // Iterate as usual
  });

Find all entities with a pair wildcard

ecs_filter_t *f = ecs_filter(world, {
  .terms[0] = ecs_pair(Eats, EcsWildcard)
});
 
ecs_iter_t it = ecs_filter_iter(world, f);
while (ecs_filter_next(&it)) {
  ecs_id_t pair_id = ecs_field_id(&it, 1);
  ecs_entity_t food = ecs_pair_second(world, pair_id); // Apples, ...
  for (int i = 0; i < it.count; i ++) {
    // Iterate as usual
  }
}
ecs_filter_fini(f);

world.filter_builder()
  .term(Eats, flecs::Wildcard)
  .build()
  .each([](flecs::iter& it, size_t i) {
    flecs::entity food = it.pair().second(); // Apples, ...
    flecs::entity e = it.entity(i);
    // Iterate as usual
  });

Iterate all children for a parent

ecs_iter_t it = ecs_children(world, parent);
while (ecs_children_next(&it)) {
  for (int i = 0; i < it.count; i ++) {
    ecs_entity_t child = it.entities[i];
    // ...
  }
}

parent.children([](flecs::entity child) {
   // ...
});

More advanced queries are possible with filters, queries and rules. See the Queries manual for more details.

Relationship components

Relationship pairs, just like regular component, can be associated with data. To associate data with a relationship pair, at least one of its elements needs to be a component. A pair can be associated with at most one type. To determine which type is associated with a relationship pair, the following rules are followed in order:

• If neither the first nor second elements are a type, the pair is a tag
• If the first element has the tag property, the pair is a tag
• If the first element is a type, the pair type is the first element
• If the second element is a type, the pair type is the second element

The following examples show how these rules can be used:

typedef struct {
  float x, y;
} Position;
 
typedef struct {
  float amount;
} Eats;
 
// Components
ECS_COMPONENT(world, Position);
ECS_COMPONENT(world, Eats);
 
// Tags
ecs_entity_t Likes = ecs_new_id(world);
ecs_entity_t Begin = ecs_new_id(world);
ecs_entity_t End = ecs_new_id(world);
ecs_entity_t Apples = ecs_new_id(world);
 
ecs_entity_t e = ecs_new_id(world);
 
// Both Likes and Apples are tags, so (Likes, Apples) is a tag
ecs_add_pair(world, e, Likes, Apples);
 
// Eats is a type and Apples is a tag, so (Eats, Apples) has type Eats
ecs_set_pair(world, e, Eats, Apples, { .amount = 1 });
 
// Begin is a tags and Position is a type, so (Begin, Position) has type Position
ecs_set_pair_second(world, e, Begin, Position, {0, 0});
ecs_set_pair_second(world, e, End, Position, {10, 20});  // same for End
 
// ChildOf has the Tag property, so even though Position is a type, the pair
// does not assume the Position type
ecs_add_pair(world, e, EcsChildOf, Position);

struct Position {
  float x, y;
};
 
struct Eats {
  float amount;
};
 
// Empty types (types without members) are automatically interpreted as tags
struct Begin { };
struct End { };
 
// Tags
flecs::entity Likes = world.entity();
flecs::entity Apples = world.entity();
 
flecs::entity e = world.entity();
 
// Both Likes and Apples are tags, so (Likes, Apples) is a tag
e.add(Likes, Apples);
 
// Eats is a type and Apples is a tag, so (Eats, Apples) has type Eats
e.set<Eats>(Apples, { 1 });
 
// Begin is a tags and Position is a type, so (Begin, Position) has type Position
e.set<Begin, Position>({0, 0});
e.set<End, Position>({10, 20}); // Same for End
 
// ChildOf has the Tag property, so even though Position is a type, the pair
// does not assume the Position type
e.add(flecs::ChildOf, world.id<Position>());

Using relationships to add components multiple times

A limitation of components is that they can only be added once to an entity. Relationships make it possible to get around this limitation, as a component can be added multiple times, as long as the pair is unique. Pairs can be constructed on the fly from new entity identifiers, which means this is possible:

typedef struct {
  float x;
  float y;
} Position;
 
ecs_entity_t e = ecs_new_id(world);
 
ecs_entity_t first = ecs_new_id(world);
ecs_entity_t second = ecs_new_id(world);
ecs_entity_t third = ecs_new_id(world);
 
// Add component position 3 times, for 3 different objects
ecs_add_pair(world, e, Position, first, {1, 2});
ecs_add_pair(world, e, Position, second, {3, 4});
ecs_add_pair(world, e, Position, third, {5, 6});

struct Position {
  float x;
  float y;
}
 
auto e = world.entity();
 
auto first = world.entity();
auto second = world.entity();
auto third = world.entity();
 
// Add component position 3 times, for 3 different objects
e.set<Position>(first, {1, 2});
e.set<Position>(second, {3, 4});
e.set<Position>(third, {5, 6});

Relationship wildcards

When querying for relationship pairs, it is often useful to be able to find all instances for a given relationship or target. To accomplish this, an application can use wildcard expressions. Consider the following example, that queries for all entities with a Likes relationship:

ecs_query_t *q = ecs_query_init(world, &(ecs_query_desc_t){
  .filter.terms = {
    {ecs_pair(Likes, EcsWildcard)}
  }
});
 
ecs_iter_t it = ecs_query_iter(world, q);
 
while (ecs_query_next(&it)) {
  ecs_id_t id = ecs_field_id(&it, 1); // Obtain pair id
 
  // Get relationship & target
  ecs_entity_t rel = ecs_pair_first(world, id);
  ecs_entity_t obj = ecs_pair_second(world, id);
 
  for (int i = 0; i < it.count; it++) {
    printf("entity %d has relationship %s, %s\n",
      it.entities[i],
      ecs_get_name(world, rel),
      ecs_get_name(world, obj));
  }
}

auto q = world.query_builder()
  .term(Likes, flecs::Wildcard)
  .build();
 
q.iter([](flecs::iter& it) {
  auto id = it.pair(1);
 
  for (auto i : it) {
    cout << "entity " << it.entity(i) << " has relationship "
      << id.first().name() << ", "
      << id.second().name() << endl;
  }
});

Wildcards may appear in query expressions, using the * character:

ecs_query_t *q = ecs_query_init(world, &(ecs_query_desc_t){
  .filter.expr = "(Likes, *)"
});

auto q = world.query("(Likes, *)");

Wildcards may used for the relationship or target part of a pair, or both:

"(Likes, *)" // Matches all Likes relationships
"(*, Alice)" // Matches all relationships with Alice as target
"(*, *)"     // Matches all relationships

Inspecting relationships

An application can use pair wildcard expressions to find all instances of a relationship for an entity. The following example shows how to find all Eats relationships for an entity:

// Bob eats apples and pears
ecs_entity_t Eats = ecs_new_entity(world, "Eats");
ecs_entity_t Apples = ecs_new_entity(world, "Apples");
ecs_entity_t Pears = ecs_new_entity(world, "Pears");
 
ecs_entity_t Bob = ecs_new_id(world);
ecs_add_pair(world, Bob, Eats, Apples);
ecs_add_pair(world, Bob, Eats, Pears);
 
// Find all (Eats, *) relationships in Bob's type
ecs_table_t *bob_table = ecs_get_table(world, Bob);
ecs_type_t bob_type = ecs_get_type(world, Bob);
ecs_id_t wildcard = ecs_pair(Eats, EcsWildcard);
ecs_id_t *ids = ecs_vector_first(bob_type, ecs_id_t);
int32_t cur = -1;
 
while (-1 != (cur = ecs_search_offset(world, bob_table, cur + 1, wildcard, 0))){
  ecs_entity_t obj = ecs_pair_second(world, ids[cur]);
  printf("Bob eats %s\n", ecs_get_name(world, obj));
}

// Bob eats apples and pears
auto Bob = world.entity();
auto Eats = world.entity();
auto Apples = world.entity();
auto Pears = world.entity();
 
Bob.add(Eats, Apples);
Bob.add(Eats, Pears);
 
// Find all (Eats, *) relationships in Bob's type
bob.match(world.pair(Eats, flecs::Wildcard), [](flecs::id id) {
  cout << "Bob eats " << id.second().name() << endl;
});
 
// For target wildcard pairs, each() can be used:
bob.each(Eats, [](flecs::entity obj) {
  cout << "Bob eats " << obj.name() << endl;
})

Builtin relationships

Flecs comes with a few builtin relationships that have special meaning within the framework. While they are implemented as regular relationships and therefore obey the same rules as any custom relationship, they are used to enhance the features of different parts of the framework. The following two sections describe the builtin relationships of Flecs.

The IsA relationship

The IsA relationship is a builtin relationship that allows applications to express that one entity is equivalent to another. This relationship is at the core of component sharing and plays a large role in queries. The IsA relationship can be used like any other relationship, as is shown here:

ecs_entity_t Apple = ecs_new_id(world);
ecs_entity_t Fruit = ecs_new_id(world);
ecs_add_pair(world, Apple, EcsIsA, Fruit);

auto Apple = world.entity();
auto Fruit = world.entity();
Apple.add(flecs::IsA, Fruit);

In C++, adding an IsA relationship has a shortcut:

Apple.is_a(Fruit);

This indicates to Flecs that an Apple is equivalent to a Fruit and should be treated as such. This equivalence is one-way, as a Fruit is not equivalent to an Apple. Another way to think about this is that IsA allows an application to express subsets and supersets. An Apple is a subset of Fruit. Fruit is a superset of Apple.

We can also add IsA relationships to Apple:

ecs_entity_t GrannySmith = ecs_new_id(world);
ecs_add_pair(world, GrannySmith, EcsIsA, Apple);

auto GrannySmith = world.entity();
GrannySmith.add(flecs::IsA, Apple);

This specifies that GrannySmith is a subset of Apple. A key thing to note here is that because Apple is a subset of Fruit, GrannySmith is a subset of Fruit as well. This means that if an application were to query for (IsA, Fruit) it would both match Apple and GrannySmith. This property of the IsA relationship is called "transitivity" and it is a feature that can be applied to any relationship. See the section on Transitivity for more details.

Component sharing

An entity with an IsA relationship to another entity is equivalent to the other entity. So far the examples showed how querying for an IsA relationship will find the subsets of the thing that was queried for. In order for entities to be treated as true equivalents though, everything the superset contains (its components, tags, relationships) must also be found on the subsets. Consider:

ecs_entity_t Spaceship = ecs_new_id(world);
ecs_set(world, Spaceship, MaxSpeed, {100});
ecs_set(world, SpaceShip, Defense, {50});
 
ecs_entity_t Frigate = ecs_new_id(world);
ecs_add(world, Frigate, EcsIsA, Spaceship);
ecs_set(world, Frigate, Defense, {100});

auto Spaceship = world.entity()
  .set<MaxSpeed>({100})
  .set<Defense>({50});
 
auto Frigate = world.entity()
  .is_a(SpaceShip) // shorthand for .add(flecs::IsA, Spaceship)
  .set<Defense>({75});

Here, the Frigate "inherits" the contents of SpaceShip. Even though MaxSpeed was never added directly to Frigate, an application can do this:

// Obtain the inherited component from Spaceship
const MaxSpeed *v = ecs_get(world, Frigate, MaxSpeed);
v->value == 100; // true

// Obtain the inherited component from Spaceship
const MaxSpeed *v = Frigate.get<MaxSpeed>();
v->value == 100; // true

While the Frigate entity also inherited the Defense component, it overrode this with its own value, so that the following example works:

// Obtain the overridden component from Frigate
const Defense *v = ecs_get(world, Frigate, Defense);
v->value == 75; // true

// Obtain the overridden component from Frigate
const Defense *v = Frigate.get<Defense>();
v->value == 75; // true

The ability to share components is also applied transitively, so Frigate could be specialized further into a FastFrigate:

ecs_entity_t FastFrigate = ecs_new_id(world);
ecs_add(world, FastFrigate, EcsIsA, Frigate);
ecs_set(world, FastFrigate, MaxSpeed, {200});
 
// Obtain the overridden component from FastFrigate
const MaxSpeed *s = ecs_get(world, Frigate, MaxSpeed);
s->value == 200; // true
 
// Obtain the inherited component from Frigate
const Defense *d = Frigate.get<Defense>();
d->value == 75; // true

auto FastFrigate = world.entity()
  .is_a(Frigate)
  .set<MaxSpeed>({200});
 
// Obtain the overridden component from FastFrigate
const MaxSpeed *s = Frigate.get<MaxSpeed>();
s->value == 200; // true
 
// Obtain the inherited component from Frigate
const Defense *d = Frigate.get<Defense>();
d->value == 75; // true

This ability to inherit and override components is one of the key enabling features of Flecs prefabs, and is further explained in the Inheritance section of the manual.

The ChildOf relationship

The ChildOf relationship is the builtin relationship that allows for the creation of entity hierarchies. The following example shows how hierarchies can be created with ChildOf:

ecs_entity_t Spaceship = ecs_new_id(world);
ecs_entity_t Cockpit = ecs_new_id(world);
 
ecs_add_pair(world, Cockpit, EcsChildOf, Spaceship);

auto Spaceship = world.entity();
auto Cockpit = world.entity();
 
Cockpit.add(flecs::ChildOf, Spaceship);

In C++, adding a ChildOf relationship has a shortcut:

Cockpit.child_of(Spaceship);

The ChildOf relationship is defined so that when a parent is deleted, its children are also deleted. For more information on specifying cleanup behavior for relationships, see the Relationship cleanup properties section.

The ChildOf relationship is defined as a regular relationship in Flecs. There are however a number of features that interact with ChildOf. The following sections describe these features.

Namespacing

Entities in flecs can have names, and name lookups can be relative to a parent. Relative name lookups can be used as a namespacing mechanism to prevent clashes between entity names. This example shows a few examples of name lookups in combination with hierarchies:

// Create two entities with a parent/child name
ecs_entity_t parent = ecs_entity(world, {
  .name = "Parent"
});
 
ecs_entity_t child = ecs_entity(world, {
  .name = "Child"
});
 
// Create the hierarchy
ecs_add_pair(world, child, EcsChildOf, parent);
 
child = ecs_lookup_fullpath(world, "Parent::Child"); // true
child = ecs_lookup_path(world, parent, "Child"); // true

auto parent = world.entity("Parent");
auto child = world.entity("Child")
  .child_of(parent);
 
child == world.lookup("Parent::Child"); // true
child == parent.lookup("Child"); // true

Scoping

In some scenarios a number of entities all need to be created with the same parent. Rather than adding the relationship to each entity, it is possible to configure the parent as a scope, which ensures that all entities created afterwards are created in the scope. The following example shows how:

ecs_entity_t parent = ecs_new_id(world);
ecs_entity_t prev = ecs_set_scope(world, parent);
 
// Note that we're not using the ecs_new_id function for the children. This
// function only generates a new id, and does not add the scope to the entity.
ecs_entity_t child_a = ecs_new(world, 0);
ecs_entity_t child_b = ecs_new(world, 0);
 
// Restore the previous scope
ecs_set_scope(world, prev);
 
ecs_has_pair(world, child_a, EcsChildOf, parent); // true
ecs_has_pair(world, child_b, EcsChildOf, parent); // true

auto parent = world.entity();
auto prev = world.set_scope(parent);
 
auto child_a = world.entity();
auto child_b = world.entity();
 
// Restore the previous scope
world.set_scope(prev);
 
child_a.has(flecs::ChildOf, parent); // true
child_b.has(flecs::ChildOf, parent); // true

Scopes in C++ can also be used with the scope function on an entity, which accepts a (typically lambda) function:

auto parent = world.entity().scope([&]{
  auto child_a = world.entity();
  auto child_b = world.entity();
 
  child_a.has(flecs::ChildOf, parent); // true
  child_b.has(flecs::ChildOf, parent); // true
});

Scopes are the mechanism that ensure contents of a module are created as children of the module, without having to explicitly add the module as a parent.

Cleanup properties

When entities that are used as tags, components, relationships or relationship targets are deleted, cleanup policies ensure that the store does not contain any dangling references. Any cleanup policy provides this guarantee, so while they are configurable, applications cannot configure policies that allows for dangling references.

Note: this only applies to entities (like tags, components, relationships) that are added to other entities. It does not apply to components that store an entity value, so:

struct MyComponent {
  entity e; // not covered by cleanup policies
}

e.add(ChildOf, parent); // covered by cleanup policies

The default policy is that any references to the entity will be removed. For example, when the tag Archer is deleted, it will be removed from all entities that have it, which is similar to invoking the remove_all operation:

ecs_remove_all(world, Archer);

world.remove_all(Archer);

Since entities can be used in relationship pairs, just calling remove_all on just the entity itself does not guarantee that no dangling references are left. A more comprehensive description of what happens is:

ecs_remove_all(world, Archer);
ecs_remove_all(world, ecs_pair(Archer, EcsWildcard));
ecs_remove_all(world, ecs_pair(EcsWildcard, Archer));

world.remove_all(Archer);
world.remove_all(Archer, flecs::Wildcard);
world.remove_all(flecs::Wildcard, Archer);

This succeeds in removing all possible references to Archer. Sometimes this behavior is not what we want however. Consider a parent-child hierarchy, where we want to delete the child entities when the parent is deleted. Instead of removing (ChildOf, parent) from all children, we need to delete the children.

We also want to specify this per relationship. If an entity has (Likes, parent) we may not want to delete that entity, meaning the cleanup we want to perform for Likes and ChildOf may not be the same.

This is what cleanup policies are for: to specify which action needs to be executed under which condition. They are applied to entities that have a reference to the entity being deleted: if I delete the Archer tag I remove the tag from all entities that have it.

To configure a cleanup policy for an entity, a (Condition, Action) pair can be added to it. If no policy is specified, the default cleanup action (Remove) is performed.

There are three cleanup actions:

• Remove: as if doing remove_all(entity) (default)
• Delete: as if doing delete_with(entity)
• Panic: throw a fatal error (default for components)

There are two cleanup conditions:

• OnDelete: the component, tag or relationship is deleted
• OnDeleteTarget: a target used with the relationship is deleted

Policies apply to both regular and pair instances, so to all entities with T as well as (T, *).

Examples

The following examples show how to use cleanup policies

**(OnDelete, Remove)**

// Remove Archer from entities when Archer is deleted
ECS_TAG(world, Archer);
ecs_add_pair(world, EcsOnDelete, EcsRemove);
 
ecs_entity_t e = ecs_new_w_id(world, Archer);
 
// This will remove Archer from e
ecs_delete(world, Archer);

// Delete entities with Archer when Archer is deleted
world.component<Archer>()
  .add(flecs::OnDelete, flecs::Remove);
 
auto e = world.entity().add<Archer>();
 
// This will remove Archer from e
world.component<Archer>().destruct();

**(OnDelete, Delete)**

// Delete entities with Archer when Archer is deleted
ECS_TAG(world, Archer);
ecs_add_pair(world, EcsOnDelete, EcsDelete);
 
ecs_entity_t e = ecs_new_w_id(world, Archer);
 
// This will delete e
ecs_delete(world, Archer);

// Delete entities with Archer when Archer is deleted
world.component<Archer>()
  .add(flecs::OnDelete, flecs::Delete);
 
auto e = world.entity().add<Archer>();
 
// This will delete e
world.component<Archer>().destruct();

**(OnDeleteTarget, Delete)**

// Delete children when deleting parent
ECS_TAG(world, ChildOf);
ecs_add_pair(world, EcsOnDeleteTarget, EcsDelete);
 
ecs_entity_t p = ecs_new_id(world);
ecs_entity_t e = ecs_new_w_pair(world, ChildOf, p);
 
// This will delete both p and e
ecs_delete(world, p);

// Delete children when deleting parent
world.component<ChildOf>()
  .add(flecs::OnDeleteTarget, flecs::Delete);
 
auto p = world.entity();
auto e = world.entity().add<ChildOf>(p);
 
// This will delete both p and e
p.destruct();

Cleanup order

While cleanup actions allow for specifying what needs to happen when a particular entity is deleted, or when an entity used with a particular relationship is deleted, they do not enforce a strict cleanup order. The reason for this is that there can be many orderings that satisfy the cleanup policies.

This is important to consider especially when writing OnRemove triggers or hooks, as the order in which they are invoked highly depends on the order in which entities are cleaned up.

Take an example with a parent and a child that both have the Node tag:

world.observer<Node>()
  .event(flecs::OnRemove)
  .each([](flecs::entity e) { });
 
flecs::entity p = world.entity().add<Node>();
flecs::entity c = world.entity().add<Node>().child_of(p);

In this example, when calling p.destruct() the observer is first invoked for the child, and then for the parent, which is to be expected as the child is deleted before the parent. Cleanup policies do not however guarantee that this is always the case.

An application could also call world.component<Node>().destruct() which would delete the Node component and all of its instances. In this scenario the cleanup policies for the ChildOf relationship are not considered, and therefore the ordering is undefined. Another typical scenario in which ordering is undefined is when an application has cyclical relationships with a Delete cleanup action.

Cleanup order during world teardown

Cleanup issues often show up during world teardown as the ordering in which entities are deleted is controlled by the application. While world teardown respects cleanup policies, there can be many entity delete orderings that are valid according to the cleanup policies, but not all of them are equally useful. There are ways to organize entities that helps world cleanup to do the right thing. These are:

Organize components, triggers, observers and systems in modules. Storing these entities in modules ensures that they stay alive for as long as possible. This leads to more predictable cleanup ordering as components will be deleted as their entities are, vs. when the component is deleted. It also ensures that triggers and observers are not deleted while matching events are still being generated.

Avoid organizing components, triggers, observers and systems under entities that are not modules. If a non-module entity with children is stored in the root, it will get cleaned up along with other regular entities. If you have entities such as these organized in a non-module scope, consider adding the EcsModule/flecs::Module tag to the root of that scope.

The next section goes into more detail on why this improves cleanup behavior and what happens during world teardown.

World teardown sequence

To understand why some ways to organize entities work better than others, having an overview of what happens during world teardown is useful. Here is a list of the steps that happen when a world is deleted:

1. Find all root entities World teardown starts by finding all root entities, which are entities that do not have the builtin ChildOf relationship.
2. Filter out modules, components, observers and systems This ensures that components are not cleaned up before the entities that use them, and triggers, observers and systems are not cleaned up while there are still conditions under which they could be invoked.
3. Filter out entities that have no children If entities have no children they cannot cause complex cleanup logic. This also decreases the likelihood of initiating cleanup actions that could impact other entities.
4. Delete root entities The root entities that were not filtered out will be deleted.
5. Delete everything else The last step will delete all remaining entities. At this point cleanup policies are no longer considered and cleanup order is undefined.

Relationship properties

Relationship properties are tags that can be added to relationships to modify their behavior.

Tag property

A relationship can be marked as a tag in which case it will never contain data. By default the data associated with a pair is determined by whether either the relationship or target are components. For some relationships however, even if the target is a component, no data should be added to the relationship. Consider the following example:

typedef struct {
  float x;
  float y;
} Position;
 
ECS_TAG(world, Serializable);
ECS_COMPONENT(world, Position);
 
ecs_entity_t e = ecs_new_id(world);
ecs_set(world, e, Position, {10, 20});
ecs_add_pair(world, e, Serializable, ecs_id(Position));
 
// Gets value from Position component
const Position *p = ecs_get(world, e, Position);
 
// Gets (unintended) value from (Serializable, Position) pair
const Position *p = ecs_get_pair_object(world, e, Serializable, Position);

struct Serializable { }; // Tag, contains no data
 
struct Position {
  float x, y;
};
 
auto e = ecs.entity()
  .set<Position>({10, 20})
  .add<Serializable, Position>(); // Because Serializable is a tag, the pair
                                  // has a value of type Position
 
// Gets value from Position component
const Position *p = e.get<Position>();
 
// Gets (unintended) value from (Serializable, Position) pair
const Position *p = e.get<Serializable, Position>();

To prevent data from being associated with pairs that can apply to components, the Tag property can be added to relationships:

// Ensure that Serializable never contains data
ecs_add_id(world, Serializable, EcsTag);
 
// Because Serializable is marked as a Tag, no data is added for the pair
// even though Position is a component
ecs_add_pair(world, e, Serializable, ecs_id(Position));
 
// This is still OK
const Position *p = ecs_get(world, e, Position);
 
// This no longer works, the pair has no data
const Position *p = ecs_get_pair_object(world, e, Serializable, Position);

// Ensure that Serializable never contains data
ecs.component<Serializable>()
  .add<flecs::Tag>();
 
auto e = ecs.entity()
  .set<Position>({10, 20})
  .add<Serializable, Position>(); // Because Serializable marked as a Tag, no
                                  // data is added for the pair even though
                                  // Position is a component
 
// Gets value from Position component
const Position *p = e.get<Position>();
 
// This no longer works, the pair has no data
const Position *p = e.get<Serializable, Position>();

The Tag property is only interpreted when it is added to the relationship part of a pair.

Final property

Entities can be annotated with the Final property, which prevents using them with IsA relationship. This is similar to the concept of a final class as something that cannot be extended. The following example shows how use Final:

ecs_entity_t e = ecs_new_id(world);
ecs_add_id(world, e, EcsFinal);
 
ecs_entity_t i = ecs_new_id(world);
ecs_add_pair(world, e, i, EcsIsA, e); // not allowed

auto e = ecs.entity()
  .add(flecs::Final);
 
auto i = ecs.entity()
  .is_a(e); // not allowed

Queries may use the final property to optimize, as they do not have to explore subsets of a final entity. For more information on how queries interpret final, see the Query manual. By default, all components are created as final.

DontInherit property

The DontInherit property prevents inheriting a component from a base entity (IsA target). Consider the following example:

ecs_entity_t TagA = ecs_new_id(world);
ecs_entity_t TagB = ecs_new_id(world);
ecs_add_id(world, TagB, EcsDontInherit);
 
ecs_entity_t base = ecs_new_id(world);
ecs_add_id(world, base, TagA);
ecs_add_id(world, base, TagB);
 
ecs_entity_t inst = ecs_new_w_pair(world, base);
ecs_has_id(world, inst, TagA); // true
ecs_has_id(world, inst, TagB); // false

struct TagA = { };
struct TagB = { };
 
world.component<TagB>().add(flecs::DontInherit);
 
auto base = world.entity()
  .add<TagA>()
  .add<TagB>();
 
auto inst = world.entity().is_a(base);
inst.has<TagA>(); // true
inst.has<TagB>(); // false

The builtin Prefab, Disabled, Identifier and ChildOf tags/relationships are marked as DontInherit.

AlwaysOverride property

The AlwaysOverride property ensures that a component is always automatically overridden when an inheritance (IsA) relationship is added. The behavior of this property is as if OVERRIDE | Component is always added together with Component.

ECS_COMPONENT(world, Position);
ECS_COMPONENT(world, Velocity);
 
ecs_add_id(world, ecs_id(Position), EcsAlwaysOverride);
 
ecs_entity_t base = ecs_new_id(world);
ecs_set(world, base, Position, {10, 20});
ecs_set(world, base, Velocity, {1, 2});
 
ecs_entity_t inst = ecs_new_w_pair(world, base);
ecs_has(world, inst, Position); // true
ecs_has(world, inst, Velocity); // true
ecs_owns(world, inst, Position); // true
ecs_owns(world, inst, Velocity); // false

world.component<Position>().add(flecs::AlwaysOverride);
 
auto base = world.entity()
  .set<Position>({10, 20})
  .add<Velocity>({1, 2})
 
auto inst = world.entity().is_a(base);
inst.has<Position>(); // true
inst.has<Velocity>(); // true
inst.owns<Position>(); // true
inst.owns<Velocity>(); // false

Transitive property

Relationships can be marked as transitive. A formal-ish definition if transitivity in the context of relationships is:

If Relationship(EntityA, EntityB) And Relationship(EntityB, EntityC) Then Relationship(EntityA, EntityC)

What this means becomes more obvious when translated to a real-life example:

If Manhattan is located in New York, and New York is located in the USA, then Manhattan is located in the USA.

In this example, LocatedIn is the relationship and Manhattan, New York and USA are entities A, B and C. Another common example of transitivity is found in OOP inheritance:

If a Square is a Rectangle and a Rectangle is a Shape, then a Square is a Shape.

In this example IsA is the relationship and Square, Rectangle and Shape are the entities.

When relationships in Flecs are marked as transitive, queries can follow the transitive relationship to see if an entity matches. Consider this example dataset:

ecs_entity_t LocatedIn = ecs_new_id(world);
ecs_entity_t Manhattan = ecs_new_id(world);
ecs_entity_t NewYork = ecs_new_id(world);
ecs_entity_t USA = ecs_new_id(world);
 
ecs_add_pair(world, Manhattan, LocatedIn, NewYork);
ecs_add_pair(world, NewYork, LocatedIn, USA);

auto LocatedIn = world.entity();
auto Manhattan = world.entity();
auto NewYork = world.entity();
auto USA = world.entity();
 
Manhattan.add(LocatedIn, NewYork);
NewYork.add(LocatedIn, USA);

If we were now to query for (LocatedIn, USA) we would only match NewYork, because we never added (LocatedIn, USA) to Manhattan. To make sure queries Manhattan as well we have to make the LocatedIn relationship transitive. We can simply do this by adding the transitive property to the relationship entity:

ecs_add_id(world, LocatedIn, Transitive);

LocatedIn.add(flecs::Transitive);

When now querying for (LocatedIn, USA), the query will follow the LocatedIn relationship and return both NewYork and Manhattan. For more details on how queries use transitivity, see the Transitive Relationships section in the query manual.

Reflexive property

A relationship can be marked reflexive which means that a query like Relationship(Entity, Entity) should evaluate to true. The utility of Reflexive becomes more obvious with an example:

Given this dataset:

IsA(Oak, Tree)

we can ask whether an oak is a tree:

IsA(Oak, Tree)
- Yes, an Oak is a tree (Oak has (IsA, Tree))

We can also ask whether a tree is a tree, which it obviously is:

IsA(Tree, Tree)
- Yes, even though Tree does not have (IsA, Tree)

However, this does not apply to all relationships. Consider a dataset with a LocatedIn relationship:

LocatedIn(SanFrancisco, UnitedStates)

we can now ask whether SanFrancisco is located in SanFrancisco, which it is not:

LocatedIn(SanFrancisco, SanFrancisco)
- No

In these examples, IsA is a reflexive relationship, whereas LocatedIn is not.

Acyclic property

A relationship can be marked with the Acyclic property to indicate that it cannot contain cycles. Both the builtin ChildOf and IsA relationships are marked acyclic. Knowing whether a relationship is acyclic allows the storage to detect and throw errors when a cyclic relationship is introduced by accident.

Note that because cycle detection requires expensive algorithms, adding Acyclic to a relationship does not guarantee that an error will be thrown when a cycle is accidentally introduced. While detection may improve over time, an application that runs without errors is no guarantee that it does not contain acyclic relationships with cycles.

Traversable property

Traversable relationships are allowed to be traversed automatically by queries, for example using the up bitflag (upwards traversal, see query traversal flags). Traversable relationships are also marked as Acyclic, which ensures a query won't accidentally attempt to traverse a relationship that contains cycles.

Events are propagated along the edges of traversable relationships. A typical example of this is when a component value is changed on a prefab. The event of this change will be propagated by traversing the IsA relationship downwards, for all instances of the prefab. Event propagation does not happen for relationships that are not marked with Traversable.

Exclusive property

The Exclusive property enforces that an entity can have only a single instance of a relationship. When a second instance is added, it replaces the first instance. An example of a relationship with the Exclusive property is the builtin ChildOf relationship:

ecs_add_pair(world, child, EcsChildOf, parent_a);
ecs_add_pair(world, child, EcsChildOf, parent_b); // replaces (ChildOf, parent_a)

e.child_of(parent_a);
e.child_of(parent_b); // replaces (ChildOf, parent_a)

To create a custom exclusive relationship, add the Exclusive property:

ecs_entity_t MarriedTo = ecs_new_id(world);
ecs_add_id(world, MarriedTo, EcsExclusive);

flecs::entity MarriedTo = world.entity()
  .add(flecs::Exclusive);

Union property

The Union is similar to Exclusive in that it enforces that an entity can have only a single instance of a relationship. The difference between Exclusive and Union is that Union combines different relationship targets in a single table. This reduces table fragmentation, and as a result speeds up add/remove operations. This increase in add/remove speed does come at a cost: iterating a query with union terms is more expensive than iterating a regular relationship.

The API for using the Union property is similar to regular relationships, as this example shows:

ecs_entity_t Movement = ecs_new_id(world);
ecs_add_id(world, Movement, EcsUnion);
 
ecs_entity_t Walking = ecs_new_id(world);
ecs_entity_t Running = ecs_new_id(world);
 
ecs_entity_t e = ecs_new_id(world);
ecs_add_pair(world, e, Movement, Running);
ecs_add_pair(world, e, Movement, Walking); // replaces (Movement, Running)

flecs::entity Movement = world.entity().add(flecs::Union);
flecs::entity Walking = world.entity();
flecs::entity Running = world.entity();
 
flecs::entity e = world.entity().add(Movement, Running);
e.add(Movement, Walking); // replaces (Movement, Running)

When compared to regular relationships, union relationships have some differences and limitations:

• Relationship cleanup does not work yet for union relationships
• Removing a union relationship removes any target, even if the specified target is different
• Filters and rules do not support union relationships
• Union relationships cannot have data
• Union relationship query terms can use only the And operator
• Queries with a (R, *) term will return (R, *) as term id for each entity

Symmetric property

The Symmetric property enforces that when a relationship (R, Y) is added to entity X, the relationship (R, X) will be added to entity Y. The reverse is also true, if relationship (R, Y) is removed from X, relationship (R, X) will be removed from Y.

The symmetric property is useful for relationships that do not make sense unless they are bidirectional. Examples of such relationships are AlliesWith, MarriedTo, TradingWith and so on. An example:

ecs_entity_t MarriedTo = ecs_new_w_id(world, EcsSymmetric);
ecs_entity_t Bob = ecs_new_id(world);
ecs_entity_t Alice = ecs_new_id(world);
ecs_add_pair(world, Bob, MarriedTo, Alice); // Also adds (MarriedTo, Bob) to Alice

auto MarriedTo = world.entity().add(flecs::Symmetric);
auto Bob = ecs.entity();
auto Alice = ecs.entity();
Bob.add(MarriedTo, Alice); // Also adds (MarriedTo, Bob) to Alice

With property

The With relationship can be added to components to indicate that it must always come together with another component. The following example shows how With can be used with regular components/tags:

ecs_entity_t Responsibility = ecs_new_id(world);
ecs_entity_t Power = ecs_new_w_pair(world, EcsWith, Responsibility);
 
// Create new entity that has both Power and Responsibility
ecs_entity_t e = ecs_new_w_id(world, Power);

auto Responsibility = world.entity();
auto Power = world.entity().add(flecs::With, Responsibility);
 
// Create new entity that has both Power and Responsibility
auto e = world.entity().add(Power);

When the With relationship is added to a relationship, the additional id added to the entity will be a relationship pair as well, with the same target as the original relationship:

ecs_entity_t Likes = ecs_new_id(world);
ecs_entity_t Loves = ecs_new_w_pair(world, EcsWith, Likes);
ecs_entity_t Pears = ecs_new_id(world);
 
// Create new entity with both (Loves, Pears) and (Likes, Pears)
ecs_entity_t e = ecs_new_w_pair(world, Loves, Pears);

auto Likes = world.entity();
auto Loves = world.entity().add(flecs::With, Likes);
auto Pears = world.entity();
 
// Create new entity with both (Loves, Pears) and (Likes, Pears)
auto e = world.entity().add(Loves, Pears);

OneOf property

The OneOf property enforces that the target of the relationship is a child of a specified entity. OneOf can be used to indicate that the target needs to be either a child of the relationship (common for enum relationships), or of another entity.

The following example shows how to constrain the relationship target to a child of the relationship:

ecs_entity_t Food = ecs_new_id(world);
 
// Enforce that target of relationship is child of Food
ecs_add_id(world, Food, EcsOneOf);
 
ecs_entity_t Apples = ecs_new_w_pair(world, EcsChildOf, Food);
ecs_entity_t Fork = ecs_new_id(world);
 
// This is ok, Apples is a child of Food
ecs_entity_t a = ecs_new_w_pair(world, Food, Apples);
 
// This is not ok, Fork is not a child of Food
ecs_entity_t b = ecs_new_w_pair(world, Food, Fork);

// Enforce that target of relationship is child of Food
auto Food = world.entity().add(flecs::OneOf);
auto Apples = world.entity().child_of(Food);
auto Fork = world.entity();
 
// This is ok, Apples is a child of Food
auto a = world.entity().add(Food, Apples);
 
// This is not ok, Fork is not a child of Food
auto b = world.entity().add(Food, Fork);

The following example shows how OneOf can be used to enforce that the relationship target is the child of an entity other than the relationship:

ecs_entity_t Food = ecs_new_id(world);
ecs_entity_t Eats = ecs_new_id(world);
 
// Enforce that target of relationship is child of Food
ecs_add_pair(world, Eats, EcsOneOf, Food);
 
ecs_entity_t Apples = ecs_new_w_pair(world, EcsChildOf, Food);
ecs_entity_t Fork = ecs_new_id(world);
 
// This is ok, Apples is a child of Food
ecs_entity_t a = ecs_new_w_pair(world, Eats, Apples);
 
// This is not ok, Fork is not a child of Food
ecs_entity_t b = ecs_new_w_pair(world, Eats, Fork);

// Enforce that target of relationship is child of Food
auto Food = world.entity();
auto Eats = world.entity().add(flecs::OneOf, Food);
auto Apples = world.entity().child_of(Food);
auto Fork = world.entity();
 
// This is ok, Apples is a child of Food
auto a = world.entity().add(Eats, Apples);
 
// This is not ok, Fork is not a child of Food
auto b = world.entity().add(Eats, Fork);

Relationship performance

This section goes over the performance implications of using relationships.

Introduction

The ECS storage needs to know two things in order to store components for entities:

• Which ids are associated with an entity
• Which types are associated with those ids

Ids represent anything that can be added to an entity. An id that is not associated with a type is called a tag. An id associated with a type is a component. For regular components, the id is a regular entity that has the builtin Component component. This component contains the information needed by the storage to associate the entity with a type. If an entity does not have the Component component, it is a tag.

Storing relationships

Relationships do not fundamentally change or extend the capabilities of the storage. Relationship pairs are two elements encoded into a single 64-bit id, which means that on the storage level they are treated the same way as regular component ids. What changes is the function that determines which type is associated with an id. For regular components this is simply a check on whether an entity has Component. To support relationships, new rules are added to determine the type of an id.

Because of this, adding/removing relationships to entities has the same performance as adding/removing regular components. This becomes more obvious when looking more closely at a function that adds a relationship pair. The following example shows how the function that adds a regular component and the function that adds a pair actually map to the same functions:

// Component
ECS_COMPONENT(world, Position);
 
// Tags
ecs_entity_t Likes = ecs_new_id(world);
ecs_entity_t Apples = ecs_new_id(world);
ecs_entity_t Npc = ecs_new_id(world);
 
ecs_entity_t e = ecs_new_id(world);
 
// The ecs_add_id function can be used to add one id to another
ecs_add_id(world, e, Npc);
 
// The ecs_add function is translated into an ecs_add_id function:
ecs_add(world, e, Position);
ecs_add_id(world, e, ecs_id(Position));
 
// The ecs_add_pair function is translated into an ecs_add_id function:
ecs_add_pair(world, e, Likes, Apples);
ecs_add_id(world, e, ecs_pair(Likes, Apples));
 
// ecs_pair() translates two ids into a single 64bit pair id.

This example also applies to C++, as the C++ API maps to the same C API functions.

While most of the storage uses the same code paths for regular components and relationships, there are a few properties of the storage that can impact performance when using relationships. These properties are not unique to relationships, but are more likely to be significant when using relationships.

Id ranges

Flecs reserves entity ids under a threshold (FLECS_HI_COMPONENT_ID, default is 256) for components. This low id range is used by the storage to more efficiently encode graph edges between tables. Graph edges for components with low ids use direct array indexing, whereas graph edges for high ids use a hashmap. Graph edges are used to find the next archetype when adding/removing component ids, and are a contributing factor to the performance overhead of add/remove operations.

Because of the way pair ids are encoded, a pair will never be in the low id range. This means that adding/removing a pair id always uses a hashmap to find the next archetype. This introduces a small overhead, which is usually 5-10% of the total cost of an operation.

Fragmentation

Fragmentation is a property of archetype-based ECS implementations where entities are spread out over more tables as the number of different component combinations increases. The overhead of fragmentation is visible in two areas:

• Table creation
• Queries (queries have to match & iterate more tables)

Applications that make extensive use of relationships might observe high levels of fragmentation, as relationships can introduce many different combinations of components. While the Flecs storage is optimized for supporting large amounts (hundreds of thousands) of tables, fragmentation is a factor to consider when using relationships.

Fragmentation can be reduced by using union relationships. There are additional storage improvements on the roadmap that will decrease the overhead of fragmentation introduced by relationships.

Relationship flattening

To reduce fragmentation, a relationship can be flattened by calling the ecs_flatten operation. This combines entities from different relationship targets in the same table, provided that they otherwise have the same components. The operation is recursive, meaning that for a specified relationship, the entire subtree is flattened.

Flattening a subtree can be done with or without preserving depth information of the original tree. Preserving depth information ensures that features like cascade still work as usual, whereas not preserving depth information further decreases fragmentation.

The flattening operation by default removes names of the flattened entities. This behavior can be overridden with an option. When flattening for the builtin ChildOf relationship, flattening with preservation of names can cause the operation to fail, if entities from different subtrees have the same name.

After the flatten operation, entities for the same target are stored in contiguous slices in the new table. The new table will have a (Target, Relationship) pair, which contains data about the original target, and the start and number of entities in the contiguous block. This information is used by queries when iterating to ensure entities for the same target can still be iterated in bulk.

Relationship flattening is an experimental feature, and some limitations apply for the current implementation. After a subtree has been flattened, the entities in that subtree can no longer be individually deleted from a target, and cannot be moved to another parent. Additionally, no components can be added/removed to entities in a flattened subtree. Relationship flattening is currently supported only for exclusive, acyclic relationships.

For additional information, see the API documentation for the ecs_flatten operation.

Table Creation

When an id added to an entity is deleted, all references to that id are deleted from the storage (see cleanup properties). For example, when the component Position is deleted it is removed from all entities, and all tables with the Position component are deleted. While not unique to relationships, it is more common for relationships to trigger cleanup actions, as relationship pairs contain regular entities.

The opposite is also true. Because relationship pairs can contain regular entities which can be created on the fly, table creation is more common than in applications that do not use relationships. While Flecs is optimized for fast table creation, creating and cleaning up tables is inherently more expensive than creating/deleting an entity. Therefore table creation is a factor to consider, especially for applications that make extensive use of relationships.

Indexing

To improve the speed of evaluating queries, Flecs has indices that store all tables for a given component id. Whenever a new table is created, it is registered with the indices for the ids the table has, including ids for relationship pairs.

While registering a table for a relationship index is not more expensive than registering a table for a regular index, a table with relationships has to also register itself with the appropriate wildcard indices for its relationships. For example, a table with relationship (Likes, Apples) registers itself with the (Likes, Apples), (Likes, *), (*, Apples) and (*, *) indices. For this reason, creating new tables with relationships has a higher overhead than a table without relationships.

Wildcard Queries

A wildcard query for a relationship pair, like (Likes, *) may return multiple results for each instance of the relationship. To find all instances of a relationship, the table index (see previous section) stores two additional pieces of information:

• The column: At which offset in the table type does the id first occur
• The count: How many occurrences of the id does the table have

If the id is not a wildcard id, the number of occurrences will always be one. When the id is a wildcard, a table type may have multiple occurrences of a relationship. For wildcard queries in the form of (Likes, *), finding all occurrences is cheap, as a query can start at the column and iterate the next count members.

For wildcard queries in the form of (*, Apples), however, the pair ids are not stored contiguously in a table type. This means that if a table has multiple instances that match (*, Apples), a query will have to perform a linear search starting from column. Once the query has found count occurrences, it can stop searching.

The following example of a table type shows how relationships are ordered, and demonstrates why (Likes, *) wildcards are easier to resolve than (*, Apples) wildcards:

Npc, (Likes, Apples), (Likes, Pears), (Likes, Bananas), (Eats, Apples), (Eats, Pears)

The index for (Likes, *) will have column=1, count=3, whereas the index for (*, Pears) will have column=2, count=2. To find all occurrences of (Likes, *) a query can start iteration at index 1 and iterate 3 elements. To find all instances of (*, Pears) a query has to start at index 2 and scan until the second instance is found.