bevy_reflect/serde/ser/
processor.rs

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
use serde::Serializer;

use crate::{PartialReflect, TypeRegistry};

/// Allows overriding the default serialization behavior of
/// [`ReflectSerializer`] and [`TypedReflectSerializer`] for specific values.
///
/// When serializing a reflected value, you may want to override the default
/// behavior and use your own logic for serialization. This logic may also be
/// context-dependent, and only apply for a single use of your
/// [`ReflectSerializer`]. To achieve this, you can create a processor and pass
/// it into your serializer.
///
/// Whenever the serializer attempts to serialize a value, it will first call
/// [`try_serialize`] on your processor, which may take ownership of the
/// serializer and write into the serializer (successfully or not), or return
/// ownership of the serializer back, and continue with the default logic.
///
/// The deserialization equivalent of this is [`ReflectDeserializerProcessor`].
///
/// # Compared to [`SerializeWithRegistry`]
///
/// [`SerializeWithRegistry`] allows you to define how your type will be
/// serialized by a [`TypedReflectSerializer`], given the extra context of the
/// [`TypeRegistry`]. If your type can be serialized entirely using that, then
/// you should prefer implementing that trait instead of using a processor.
///
/// However, you may need more context-dependent data which is only present in
/// the scope where you create the [`TypedReflectSerializer`]. For example, if
/// you need to use a reference to a value while serializing, then there is no
/// way to do this with [`SerializeWithRegistry`] as you can't pass that
/// reference into anywhere. This is where a processor is useful, as the
/// processor can capture local variables.
///
/// A [`ReflectSerializerProcessor`] always takes priority over a
/// [`SerializeWithRegistry`] implementation, so this is also useful for
/// overriding serialization behavior if you need to do something custom.
///
/// # Examples
///
/// Serializing a reflected value when saving an asset to disk, and replacing
/// asset handles with the handle path (if it has one):
///
/// ```
/// # use core::any::Any;
/// # use serde::Serialize;
/// # use bevy_reflect::{PartialReflect, Reflect, TypeData, TypeRegistry};
/// # use bevy_reflect::serde::{ReflectSerializer, ReflectSerializerProcessor};
/// #
/// # #[derive(Debug, Clone, Reflect)]
/// # struct Handle<T>(T);
/// # #[derive(Debug, Clone, Reflect)]
/// # struct Mesh;
/// #
/// # struct ReflectHandle;
/// # impl TypeData for ReflectHandle {
/// #     fn clone_type_data(&self) -> Box<dyn TypeData> {
/// #         unimplemented!()
/// #     }
/// # }
/// # impl ReflectHandle {
/// #     fn downcast_handle_untyped(&self, handle: &(dyn Any + 'static)) -> Option<UntypedHandle> {
/// #         unimplemented!()
/// #     }
/// # }
/// #
/// # #[derive(Debug, Clone)]
/// # struct UntypedHandle;
/// # impl UntypedHandle {
/// #     fn path(&self) -> Option<&str> {
/// #         unimplemented!()
/// #     }
/// # }
/// # type AssetError = Box<dyn core::error::Error>;
/// #
/// #[derive(Debug, Clone, Reflect)]
/// struct MyAsset {
///     name: String,
///     mesh: Handle<Mesh>,
/// }
///
/// struct HandleProcessor;
///
/// impl ReflectSerializerProcessor for HandleProcessor {
///     fn try_serialize<S>(
///         &self,
///         value: &dyn PartialReflect,
///         registry: &TypeRegistry,
///         serializer: S,
///     ) -> Result<Result<S::Ok, S>, S::Error>
///     where
///         S: serde::Serializer,
///     {
///         let Some(value) = value.try_as_reflect() else {
///             // we don't have any info on this type; do the default serialization logic
///             return Ok(Err(serializer));
///         };
///         let type_id = value.reflect_type_info().type_id();
///         let Some(reflect_handle) = registry.get_type_data::<ReflectHandle>(type_id) else {
///             // this isn't a `Handle<T>`
///             return Ok(Err(serializer));
///         };
///
///         let untyped_handle = reflect_handle
///             .downcast_handle_untyped(value.as_any())
///             .unwrap();
///         if let Some(path) = untyped_handle.path() {
///             Ok(Ok(serializer.serialize_str(path)?))
///         } else {
///             Ok(Ok(serializer.serialize_unit()?))
///         }
///     }
/// }
///
/// fn save(type_registry: &TypeRegistry, asset: &MyAsset) -> Result<Vec<u8>, AssetError> {
///     let mut asset_bytes = Vec::new();
///
///     let processor = HandleProcessor;
///     let serializer = ReflectSerializer::with_processor(asset, type_registry, &processor);
///     let mut ron_serializer = ron::Serializer::new(&mut asset_bytes, None)?;
///
///     serializer.serialize(&mut ron_serializer)?;
///     Ok(asset_bytes)
/// }
/// ```
///
/// [`ReflectSerializer`]: crate::serde::ReflectSerializer
/// [`TypedReflectSerializer`]: crate::serde::TypedReflectSerializer
/// [`try_serialize`]: Self::try_serialize
/// [`SerializeWithRegistry`]: crate::serde::SerializeWithRegistry
/// [`ReflectDeserializerProcessor`]: crate::serde::ReflectDeserializerProcessor
pub trait ReflectSerializerProcessor {
    /// Attempts to serialize the value which a [`TypedReflectSerializer`] is
    /// currently looking at.
    ///
    /// If you want to override the default serialization, return
    /// `Ok(Ok(value))` with an `Ok` output from the serializer.
    ///
    /// If you don't want to override the serialization, return ownership of
    /// the serializer back via `Ok(Err(serializer))`.
    ///
    /// You can use the type registry to read info about the type you're
    /// serializing, or just try to downcast the value directly:
    ///
    /// ```
    /// # use bevy_reflect::{TypeRegistration, TypeRegistry, PartialReflect};
    /// # use bevy_reflect::serde::ReflectSerializerProcessor;
    /// # use core::any::TypeId;
    /// struct I32AsStringProcessor;
    ///
    /// impl ReflectSerializerProcessor for I32AsStringProcessor {
    ///     fn try_serialize<S>(
    ///         &self,
    ///         value: &dyn PartialReflect,
    ///         registry: &TypeRegistry,
    ///         serializer: S,
    ///     ) -> Result<Result<S::Ok, S>, S::Error>
    ///     where
    ///         S: serde::Serializer
    ///     {
    ///         if let Some(value) = value.try_downcast_ref::<i32>() {
    ///             let value_as_string = format!("{value:?}");
    ///             Ok(Ok(serializer.serialize_str(&value_as_string)?))
    ///         } else {
    ///             // Not an `i32`, just do the default serialization
    ///             Ok(Err(serializer))
    ///         }
    ///     }
    /// }
    /// ```
    ///
    /// [`TypedReflectSerializer`]: crate::serde::TypedReflectSerializer
    /// [`Reflect`]: crate::Reflect
    fn try_serialize<S>(
        &self,
        value: &dyn PartialReflect,
        registry: &TypeRegistry,
        serializer: S,
    ) -> Result<Result<S::Ok, S>, S::Error>
    where
        S: Serializer;
}

impl ReflectSerializerProcessor for () {
    fn try_serialize<S>(
        &self,
        _value: &dyn PartialReflect,
        _registry: &TypeRegistry,
        serializer: S,
    ) -> Result<Result<S::Ok, S>, S::Error>
    where
        S: Serializer,
    {
        Ok(Err(serializer))
    }
}