-
-
Notifications
You must be signed in to change notification settings - Fork 4.2k
Refactor non-core Curve methods into extension traits #16930
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Merged
Merged
+187
−161
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
BenjaminBrienen
approved these changes
Dec 22, 2024
IQuick143
approved these changes
Dec 29, 2024
ecoskey
pushed a commit
to ecoskey/bevy
that referenced
this pull request
Jan 6, 2025
# Objective The way `Curve` presently achieves dyn-compatibility involves shoving `Self: Sized` bounds on a bunch of methods to forbid them from appearing in vtables. (This is called *explicit non-dispatchability*.) The `Curve` trait probably also just has way too many methods on its own. In the past, using extension traits instead to achieve similar functionality has been discussed. The upshot is that this would allow the "core" of the curve trait, on which all the automatic methods rely, to live in a very simple dyn-compatible trait, while other functionality is implemented by extensions. For instance, `dyn Curve<T>` cannot use the `Sized` methods, but `Box<dyn Curve<T>>` is `Sized`, hence would automatically implement the extension trait, containing the methods which are currently non-dispatchable. Other motivations for this include modularity and code organization: the `Curve` trait itself has grown quite large with the addition of numerous adaptors, and refactoring it to demonstrate the separation of functionality that is already present makes a lot of sense. Furthermore, resampling behavior in particular is dependent on special traits that may be mimicked or analogized in user-space, and creating extension traits to achieve similar behavior in user-space is something we ought to encourage by example. ## Solution `Curve` now contains only `domain` and the `sample` methods. `CurveExt` has been created, and it contains all adaptors, along with the other sampling convenience methods (`samples`, `sample_iter`, etc.). It is implemented for all `C` where `C: Curve<T> + Sized`. `CurveResampleExt` has been created, and it contains all resampling methods. It is implemented for all `C` where `C: Curve<T> + ?Sized`. ## Testing It compiles and `cargo doc` succeeds. --- ## Future work - Consider writing extension traits for resampling curves in related domains (e.g. resampling for `Curve<T>` where `T: Animatable` into an `AnimatableKeyframeCurve`). - `CurveExt` might be further broken down to separate the adaptor and sampling methods. --- ## Migration Guide `Curve` has been refactored so that much of its functionality is now in extension traits. Adaptors such as `map`, `reparametrize`, `reverse`, and so on now require importing `CurveExt`, while the resampling methods `resample_*` require importing `CurveResampleExt`. Both of these new traits are exported through `bevy::math::curve` and through `bevy::math::prelude`.
mrchantey
pushed a commit
to mrchantey/bevy
that referenced
this pull request
Feb 4, 2025
# Objective The way `Curve` presently achieves dyn-compatibility involves shoving `Self: Sized` bounds on a bunch of methods to forbid them from appearing in vtables. (This is called *explicit non-dispatchability*.) The `Curve` trait probably also just has way too many methods on its own. In the past, using extension traits instead to achieve similar functionality has been discussed. The upshot is that this would allow the "core" of the curve trait, on which all the automatic methods rely, to live in a very simple dyn-compatible trait, while other functionality is implemented by extensions. For instance, `dyn Curve<T>` cannot use the `Sized` methods, but `Box<dyn Curve<T>>` is `Sized`, hence would automatically implement the extension trait, containing the methods which are currently non-dispatchable. Other motivations for this include modularity and code organization: the `Curve` trait itself has grown quite large with the addition of numerous adaptors, and refactoring it to demonstrate the separation of functionality that is already present makes a lot of sense. Furthermore, resampling behavior in particular is dependent on special traits that may be mimicked or analogized in user-space, and creating extension traits to achieve similar behavior in user-space is something we ought to encourage by example. ## Solution `Curve` now contains only `domain` and the `sample` methods. `CurveExt` has been created, and it contains all adaptors, along with the other sampling convenience methods (`samples`, `sample_iter`, etc.). It is implemented for all `C` where `C: Curve<T> + Sized`. `CurveResampleExt` has been created, and it contains all resampling methods. It is implemented for all `C` where `C: Curve<T> + ?Sized`. ## Testing It compiles and `cargo doc` succeeds. --- ## Future work - Consider writing extension traits for resampling curves in related domains (e.g. resampling for `Curve<T>` where `T: Animatable` into an `AnimatableKeyframeCurve`). - `CurveExt` might be further broken down to separate the adaptor and sampling methods. --- ## Migration Guide `Curve` has been refactored so that much of its functionality is now in extension traits. Adaptors such as `map`, `reparametrize`, `reverse`, and so on now require importing `CurveExt`, while the resampling methods `resample_*` require importing `CurveResampleExt`. Both of these new traits are exported through `bevy::math::curve` and through `bevy::math::prelude`.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Labels
A-Math
Fundamental domain-agnostic mathematical operations
C-Code-Quality
A section of code that is hard to understand or change
C-Usability
A targeted quality-of-life change that makes Bevy easier to use
D-Straightforward
Simple bug fixes and API improvements, docs, test and examples
M-Migration-Guide
A breaking change to Bevy's public API that needs to be noted in a migration guide
S-Ready-For-Final-Review
This PR has been approved by the community. It's ready for a maintainer to consider merging it
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Objective
The way
Curvepresently achieves dyn-compatibility involves shovingSelf: Sizedbounds on a bunch of methods to forbid them from appearing in vtables. (This is called explicit non-dispatchability.) TheCurvetrait probably also just has way too many methods on its own.In the past, using extension traits instead to achieve similar functionality has been discussed. The upshot is that this would allow the "core" of the curve trait, on which all the automatic methods rely, to live in a very simple dyn-compatible trait, while other functionality is implemented by extensions. For instance,
dyn Curve<T>cannot use theSizedmethods, butBox<dyn Curve<T>>isSized, hence would automatically implement the extension trait, containing the methods which are currently non-dispatchable.Other motivations for this include modularity and code organization: the
Curvetrait itself has grown quite large with the addition of numerous adaptors, and refactoring it to demonstrate the separation of functionality that is already present makes a lot of sense. Furthermore, resampling behavior in particular is dependent on special traits that may be mimicked or analogized in user-space, and creating extension traits to achieve similar behavior in user-space is something we ought to encourage by example.Solution
Curvenow contains onlydomainand thesamplemethods.CurveExthas been created, and it contains all adaptors, along with the other sampling convenience methods (samples,sample_iter, etc.). It is implemented for allCwhereC: Curve<T> + Sized.CurveResampleExthas been created, and it contains all resampling methods. It is implemented for allCwhereC: Curve<T> + ?Sized.Testing
It compiles and
cargo docsucceeds.Future work
Curve<T>whereT: Animatableinto anAnimatableKeyframeCurve).CurveExtmight be further broken down to separate the adaptor and sampling methods.Migration Guide
Curvehas been refactored so that much of its functionality is now in extension traits. Adaptors such asmap,reparametrize,reverse, and so on now require importingCurveExt, while the resampling methodsresample_*require importingCurveResampleExt. Both of these new traits are exported throughbevy::math::curveand throughbevy::math::prelude.