Crate bevy_reflect

Expand description

Reflection in Rust.

Reflection is a powerful tool provided within many programming languages that allows for meta-programming: using information about the program to affect the program. In other words, reflection allows us to inspect the program itself, its syntax, and its type information at runtime.

This crate adds this missing reflection functionality to Rust. Though it was made with the Bevy game engine in mind, it’s a general-purpose solution that can be used in any Rust project.

At a very high level, this crate allows you to:

Dynamically interact with Rust values
Access type metadata at runtime
Serialize and deserialize (i.e. save and load) data

It’s important to note that because of missing features in Rust, there are some limitations with this crate.

§The `Reflect` Trait

At the core of bevy_reflect is the Reflect trait.

One of its primary purposes is to allow all implementors to be passed around as a dyn Reflect trait object. This allows any such type to be operated upon completely dynamically (at a small runtime cost).

Implementing the trait is easily done using the provided derive macro:

#[derive(Reflect)]
struct MyStruct {
  foo: i32
}

This will automatically generate the implementation of Reflect for any struct or enum.

It will also generate other very important trait implementations used for reflection:

GetTypeRegistration
Typed
Struct, TupleStruct, or Enum depending on the type

§Requirements

We can implement Reflect on any type that satisfies both of the following conditions:

The type implements Any. This is true if and only if the type itself has a 'static lifetime.
All fields and sub-elements themselves implement Reflect (see the derive macro documentation for details on how to ignore certain fields when deriving).

Additionally, using the derive macro on enums requires a third condition to be met:

All fields and sub-elements must implement FromReflect— another important reflection trait discussed in a later section.

§The `Reflect` Subtraits

Since Reflect is meant to cover any and every type, this crate also comes with a few more traits to accompany Reflect and provide more specific interactions. We refer to these traits as the reflection subtraits since they all have Reflect as a supertrait. The current list of reflection subtraits include:

As mentioned previously, the last three are automatically implemented by the derive macro.

Each of these traits come with their own methods specific to their respective category. For example, we can access our struct’s fields by name using the Struct::field method.

let my_struct: Box<dyn Struct> = Box::new(MyStruct {
  foo: 123
});
let foo: &dyn Reflect = my_struct.field("foo").unwrap();
assert_eq!(Some(&123), foo.downcast_ref::<i32>());

Since most data is passed around as dyn Reflect, the Reflect trait has methods for going to and from these subtraits.

Reflect::reflect_kind, Reflect::reflect_ref, Reflect::reflect_mut, and Reflect::reflect_owned all return an enum that respectively contains zero-sized, immutable, mutable, and owned access to the type as a subtrait object.

For example, we can get out a dyn Tuple from our reflected tuple type using one of these methods.

let my_tuple: Box<dyn Reflect> = Box::new((1, 2, 3));
let ReflectRef::Tuple(my_tuple) = my_tuple.reflect_ref() else { unreachable!() };
assert_eq!(3, my_tuple.field_len());

And to go back to a general-purpose dyn Reflect, we can just use the matching Reflect::as_reflect, Reflect::as_reflect_mut, or Reflect::into_reflect methods.

§Value Types

Types that do not fall under one of the above subtraits, such as for primitives (e.g. bool, usize, etc.) and simple types (e.g. String, Duration), are referred to as value types since methods like Reflect::reflect_ref return a ReflectRef::Value variant. While most other types contain their own dyn Reflect fields and data, these types generally cannot be broken down any further.

§Dynamic Types

Each subtrait comes with a corresponding dynamic type.

The available dynamic types are:

These dynamic types may contain any arbitrary reflected data.

let mut data = DynamicStruct::default();
data.insert("foo", 123_i32);
assert_eq!(Some(&123), data.field("foo").unwrap().downcast_ref::<i32>())

They are most commonly used as “proxies” for other types, where they contain the same data as— and therefore, represent— a concrete type. The Reflect::clone_value method will return a dynamic type for all non-value types, allowing all types to essentially be “cloned”. And since dynamic types themselves implement Reflect, we may pass them around just like any other reflected type.

let original: Box<dyn Reflect> = Box::new(MyStruct {
  foo: 123
});

// `cloned` will be a `DynamicStruct` representing a `MyStruct`
let cloned: Box<dyn Reflect> = original.clone_value();
assert!(cloned.represents::<MyStruct>());
assert!(cloned.is::<DynamicStruct>());

§Patching

These dynamic types come in handy when needing to apply multiple changes to another type. This is known as “patching” and is done using the Reflect::apply and Reflect::try_apply methods.

let mut value = Some(123_i32);
let patch = DynamicEnum::new("None", ());
value.apply(&patch);
assert_eq!(None, value);

§`FromReflect`

It’s important to remember that dynamic types are not the concrete type they may be representing. A common mistake is to treat them like such when trying to cast back to the original type or when trying to make use of a reflected trait which expects the actual type.

let original: Box<dyn Reflect> = Box::new(MyStruct {
  foo: 123
});

let cloned: Box<dyn Reflect> = original.clone_value();
let value = cloned.take::<MyStruct>().unwrap(); // PANIC!

To resolve this issue, we’ll need to convert the dynamic type to the concrete one. This is where FromReflect comes in.

FromReflect is a trait that allows an instance of a type to be generated from a dynamic representation— even partial ones. And since the FromReflect::from_reflect method takes the data by reference, this can be used to effectively clone data (to an extent).

It is automatically implemented when deriving Reflect on a type unless opted out of using #[reflect(from_reflect = false)] on the item.

#[derive(Reflect)]
struct MyStruct {
  foo: i32
}
let original: Box<dyn Reflect> = Box::new(MyStruct {
  foo: 123
});

let cloned: Box<dyn Reflect> = original.clone_value();
let value = <MyStruct as FromReflect>::from_reflect(&*cloned).unwrap(); // OK!

When deriving, all active fields and sub-elements must also implement FromReflect.

Fields can be given default values for when a field is missing in the passed value or even ignored. Ignored fields must either implement Default or have a default function specified using #[reflect(default = "path::to::function")].

See the derive macro documentation for details.

All primitives and simple types implement FromReflect by relying on their Default implementation.

The GetPath trait allows accessing arbitrary nested fields of a Reflect type.

Using GetPath, it is possible to use a path string to access a specific field of a reflected type.

#[derive(Reflect)]
struct MyStruct {
  value: Vec<Option<u32>>
}

let my_struct = MyStruct {
  value: vec![None, None, Some(123)],
};
assert_eq!(
  my_struct.path::<u32>(".value[2].0").unwrap(),
  &123,
);

§Type Registration

This crate also comes with a TypeRegistry that can be used to store and retrieve additional type metadata at runtime, such as helper types and trait implementations.

The derive macro for Reflect also generates an implementation of the GetTypeRegistration trait, which is used by the registry to generate a TypeRegistration struct for that type. We can then register additional type data we want associated with that type.

For example, we can register ReflectDefault on our type so that its Default implementation may be used dynamically.

#[derive(Reflect, Default)]
struct MyStruct {
  foo: i32
}
let mut registry = TypeRegistry::empty();
registry.register::<MyStruct>();
registry.register_type_data::<MyStruct, ReflectDefault>();

let registration = registry.get(std::any::TypeId::of::<MyStruct>()).unwrap();
let reflect_default = registration.data::<ReflectDefault>().unwrap();

let new_value: Box<dyn Reflect> = reflect_default.default();
assert!(new_value.is::<MyStruct>());

Because this operation is so common, the derive macro actually has a shorthand for it. By using the #[reflect(Trait)] attribute, the derive macro will automatically register a matching, in-scope ReflectTrait type within the GetTypeRegistration implementation.

use bevy_reflect::prelude::{Reflect, ReflectDefault};

#[derive(Reflect, Default)]
#[reflect(Default)]
struct MyStruct {
  foo: i32
}

§Reflecting Traits

Type data doesn’t have to be tied to a trait, but it’s often extremely useful to create trait type data. These allow traits to be used directly on a dyn Reflect while utilizing the underlying type’s implementation.

For any object-safe trait, we can easily generate a corresponding ReflectTrait type for our trait using the #[reflect_trait] macro.

#[reflect_trait] // Generates a `ReflectMyTrait` type
pub trait MyTrait {}
impl<T: Reflect> MyTrait for T {}

let mut registry = TypeRegistry::new();
registry.register_type_data::<i32, ReflectMyTrait>();

The generated type data can be used to convert a valid dyn Reflect into a dyn MyTrait. See the trait reflection example for more information and usage details.

§Serialization

By using reflection, we are also able to get serialization capabilities for free. In fact, using bevy_reflect can result in faster compile times and reduced code generation over directly deriving the serde traits.

The way it works is by moving the serialization logic into common serializers and deserializers:

All of these structs require a reference to the registry so that type information can be retrieved, as well as registered type data, such as ReflectSerialize and ReflectDeserialize.

The general entry point are the “untyped” versions of these structs. These will automatically extract the type information and pass them into their respective “typed” version.

The output of the ReflectSerializer will be a map, where the key is the type path and the value is the serialized data. The TypedReflectSerializer will simply output the serialized data.

The ReflectDeserializer can be used to deserialize this map and return a Box<dyn Reflect>, where the underlying type will be a dynamic type representing some concrete type (except for value types).

Again, it’s important to remember that dynamic types may need to be converted to their concrete counterparts in order to be used in certain cases. This can be achieved using FromReflect.

#[derive(Reflect, PartialEq, Debug)]
struct MyStruct {
  foo: i32
}

let original_value = MyStruct {
  foo: 123
};

// Register
let mut registry = TypeRegistry::new();
registry.register::<MyStruct>();

// Serialize
let reflect_serializer = ReflectSerializer::new(&original_value, &registry);
let serialized_value: String = ron::to_string(&reflect_serializer).unwrap();

// Deserialize
let reflect_deserializer = ReflectDeserializer::new(&registry);
let deserialized_value: Box<dyn Reflect> = reflect_deserializer.deserialize(
  &mut ron::Deserializer::from_str(&serialized_value).unwrap()
).unwrap();

// Convert
let converted_value = <MyStruct as FromReflect>::from_reflect(&*deserialized_value).unwrap();

assert_eq!(original_value, converted_value);

§Limitations

While this crate offers a lot in terms of adding reflection to Rust, it does come with some limitations that don’t make it as featureful as reflection in other programming languages.

§Non-Static Lifetimes

One of the most obvious limitations is the 'static requirement. Rust requires fields to define a lifetime for referenced data, but Reflect requires all types to have a 'static lifetime. This makes it impossible to reflect any type with non-static borrowed data.

§Function Reflection

Another limitation is the inability to fully reflect functions and methods. Most languages offer some way of calling methods dynamically, but Rust makes this very difficult to do. For non-generic methods, this can be done by registering custom type data that contains function pointers. For generic methods, the same can be done but will typically require manual monomorphization (i.e. manually specifying the types the generic method can take).

§Manual Registration

Since Rust doesn’t provide built-in support for running initialization code before main, there is no way for bevy_reflect to automatically register types into the type registry. This means types must manually be registered, including their desired monomorphized representations if generic.

§Features

§`bevy`

Default	Dependencies
❌	`bevy_math`, `glam`, `smallvec`

This feature makes it so that the appropriate reflection traits are implemented on all the types necessary for the Bevy game engine. enables the optional dependencies: bevy_math, glam, and smallvec. These dependencies are used by the Bevy game engine and must define their reflection implementations within this crate due to Rust’s orphan rule.

§`documentation`

Default	Dependencies
❌	`bevy_reflect_derive/documentation`

This feature enables capturing doc comments as strings for items that derive Reflect. Documentation information can then be accessed at runtime on the TypeInfo of that item.

This can be useful for generating documentation for scripting language interop or for displaying tooltips in an editor.

Re-exports§

pub use erased_serde;

Modules§

access: Representation for individual element accesses within a path.
attributes
prelude
serde
std_traits
utility: Helpers for working with Bevy reflection.

Macros§

hash_error
impl_from_reflect_value: A macro used to generate a FromReflect trait implementation for the given type.
impl_reflect: A replacement for #[derive(Reflect)] to be used with foreign types which the definitions of cannot be altered.
impl_reflect_value: A macro used to generate reflection trait implementations for the given type.
impl_type_path: A replacement for deriving TypePath for use on foreign types.

Structs§

AccessError: An error originating from an Access of an element within a type.
ArrayInfo: A container for compile-time array info.
ArrayIter: An iterator over an Array.
DynamicArray: A fixed-size list of reflected values.
DynamicEnum: A dynamic representation of an enum.
DynamicList: A list of reflected values.
DynamicMap: An ordered mapping between reflected values.
DynamicStruct: A struct type which allows fields to be added at runtime.
DynamicTuple: A tuple which allows fields to be added at runtime.
DynamicTupleStruct: A tuple struct which allows fields to be added at runtime.
EnumInfo: A container for compile-time enum info, used by TypeInfo.
FieldIter: An iterator over the field values of a struct.
ListInfo: A container for compile-time list info.
ListIter: An iterator over an List.
MapInfo: A container for compile-time map info.
MapIter: An iterator over the key-value pairs of a Map.
NamedField: The named field of a reflected struct.
OffsetAccess: An Access combined with an offset for more helpful error reporting.
ParseError: An error that occurs when parsing reflect path strings.
ParsedPath: A pre-parsed path to an element within a type.
ReflectDeserialize: A struct used to deserialize reflected instances of a type.
ReflectFromPtr: Reflect values are commonly used in situations where the actual types of values are not known at runtime. In such situations you might have access to a *const () pointer that you know implements Reflect, but have no way of turning it into a &dyn Reflect.
ReflectFromReflect: Type data that represents the FromReflect trait and allows it to be used dynamically.
ReflectSerialize: A struct used to serialize reflected instances of a type.
StructInfo: A container for compile-time named struct info.
StructVariantInfo: Type info for struct variants.
TupleFieldIter: An iterator over the field values of a tuple.
TupleInfo: A container for compile-time tuple info.
TupleStructFieldIter: An iterator over the field values of a tuple struct.
TupleStructInfo: A container for compile-time tuple struct info.
TupleVariantInfo: Type info for tuple variants.
TypePathTable: Provides dynamic access to all methods on TypePath.
TypeRegistration: Runtime storage for type metadata, registered into the TypeRegistry.
TypeRegistry: A registry of reflected types.
TypeRegistryArc: A synchronized wrapper around a TypeRegistry.
UnitVariantInfo: Type info for unit variants.
UnnamedField: The unnamed field of a reflected tuple or tuple struct.
ValueInfo: A container for compile-time info related to general value types, including primitives.
VariantFieldIter: An iterator over the fields in the current enum variant.

Enums§

Access: A singular element access within a path. Multiple accesses can be combined into a ParsedPath.
AccessErrorKind: The kind of AccessError, along with some kind-specific information.
ApplyError: A enumeration of all error outcomes that might happen when running try_apply.
DynamicVariant: A dynamic representation of an enum variant.
ReflectKind: A zero-sized enumuration of the “kinds” of a reflected type.
ReflectMut: A mutable enumeration of “kinds” of a reflected type.
ReflectOwned: An owned enumeration of “kinds” of a reflected type.
ReflectPathError: An error returned from a failed path string query.
ReflectRef: An immutable enumeration of “kinds” of a reflected type.
TypeInfo: Compile-time type information for various reflected types.
VariantField
VariantInfo: A container for compile-time enum variant info.
VariantType: Describes the form of an enum variant.

Traits§

Array: A trait used to power array-like operations via reflection.
DynamicTypePath: Dynamic dispatch for TypePath.
Enum: A trait used to power enum-like operations via reflection.
FromReflect: A trait that enables types to be dynamically constructed from reflected data.
FromType: Trait used to generate TypeData for trait reflection.
GetField: A convenience trait which combines fetching and downcasting of struct fields.
GetPath: A trait which allows nested Reflect values to be retrieved with path strings.
GetTupleField: A convenience trait which combines fetching and downcasting of tuple fields.
GetTupleStructField: A convenience trait which combines fetching and downcasting of tuple struct fields.
GetTypeRegistration: A trait which allows a type to generate its TypeRegistration for registration into the TypeRegistry.
List: A trait used to power list-like operations via reflection.
Map: A trait used to power map-like operations via reflection.
Reflect: The core trait of bevy_reflect, used for accessing and modifying data dynamically.
ReflectPath: Something that can be interpreted as a reflection path in GetPath.
Struct: A trait used to power struct-like operations via reflection.
Tuple: A trait used to power tuple-like operations via reflection.
TupleStruct: A trait used to power tuple struct-like operations via reflection.
TypeData: A trait used to type-erase type metadata.
TypePath: A static accessor to type paths and names.
Typed: A static accessor to compile-time type information.

Functions§

array_apply: Applies the reflected array data to the given array.
array_debug: The default debug formatter for Array types.
array_hash: Returns the u64 hash of the given array.
array_partial_eq: Compares two arrays (one concrete and one reflected) to see if they are equal.
array_try_apply: Tries to apply the reflected array data to the given array and returns a Result.
enum_debug: The default debug formatter for Enum types.
enum_hash: Returns the u64 hash of the given enum.
enum_partial_eq: Compares an Enum with a Reflect value.
list_apply: Applies the elements of b to the corresponding elements of a.
list_debug: The default debug formatter for List types.
list_hash: Returns the u64 hash of the given list.
list_partial_eq: Compares a List with a Reflect value.
list_try_apply: Tries to apply the elements of b to the corresponding elements of a and returns a Result.
map_apply: Applies the elements of reflected map b to the corresponding elements of map a.
map_debug: The default debug formatter for Map types.
map_partial_eq: Compares a Map with a Reflect value.
map_try_apply: Tries to apply the elements of reflected map b to the corresponding elements of map a and returns a Result.
struct_debug: The default debug formatter for Struct types.
struct_partial_eq: Compares a Struct with a Reflect value.
tuple_apply: Applies the elements of b to the corresponding elements of a.
tuple_debug: The default debug formatter for Tuple types.
tuple_partial_eq: Compares a Tuple with a Reflect value.
tuple_struct_debug: The default debug formatter for TupleStruct types.
tuple_struct_partial_eq: Compares a TupleStruct with a Reflect value.
tuple_try_apply: Tries to apply the elements of b to the corresponding elements of a and returns a Result.

Attribute Macros§

reflect_trait: A macro that automatically generates type data for traits, which their implementors can then register.

Derive Macros§

FromReflect: Derives the FromReflect trait.
Reflect: The main derive macro used by bevy_reflect for deriving its Reflect trait.
TypePath: Derives the TypePath trait, providing a stable alternative to [std::any::type_name].