Summary:
This is a very crucial change, everything else in this stack depends on it.
Here is the set of constraints that we have for ShadowNode-based class hierarchy:
* `ShadowNode` is a base class that defines basic operations on the nodes. It does not have virtual methods by design (virtual dispatch hurts performance; we want to limit and centralize that in `ComponentDescriptor`s).
* `ConcreteShadowNode<>` template *statically* wires particular `ShadowNode` and particular `Props` establishing a type-safe relationship between them ensured on compile time.
* Not all `ShadowNode`s are "layoutable", not all "layoutable" `ShadowNode`s use Yoga to do layout.
* Layotability (and YogaLayotability) feature is implemented as two classes `LayoutableShadowNode` and `YogaLayoutableShadowNode`. These classes essentially need to know something about the ShadowNode nature of the object (get the list of children or replace some children).
Before the change, `LayoutableShadowNode` and `YogaLayoutableShadowNode`classes did not inherit `ShadowNode` because we don't want to use *virtual inheritance* for `ShadowNode`: `ConcreteShadowNode<>` already inherits `ShadowNode`, so if we make `LayoutableShadowNode` inherits `ShadowNode`, we will have to somehow flatten this inheritance hierarchy (and the virtual inheritance is an answer to that). (Yes, C++ supports multiple inheritance and virtual inheritance.)
Before this change, we solved this dilemma this way: We have a subclass-template `ConcreteViewShadowNode<>` that inherits `ConcreteShadowNode<>` and `YogaLayoutableShadowNode`. Then, we had a bunch of methods that implement the rerouting of some functionality from `YogaLayoutableShadowNode` to `ShadowNode` and vise-versa. (See the diagram "Before".)
That worked fine, except the caveats:
* That wiring is nasty, complex, hard to reason about and overall limiting.
* There is no way to statically cast `LayoutableShadowNode` to `ShadowNode` (because there is no common base class). That forces us to use dynamic_cast on some perf critical paths (including layout, diffing and so on).
* Adding features that rely on interop between `LayoutableShadowNode` and `ShadowNode` is a nightmare.
It should be a better way to deal with this dilemma, and this diff implements a different approach: We can have a base class of `ConcreteShadowNode<>` as a template parameter. With this approach, we can make `LayoutableShadowNode` inherit `ShadowNode` and when we need to instantiate a `ConcreteShadowNode<>` that needs to be layoutable, we can just specify `YogaLayoutableShadowNode` as a base class. (See the diagram "After".)
This simple change will allow us to simplify a lot of things. The rest of the stack is about getting rid of unnecessary moving parts. Which will finally allow us to build "Inline Views" feature.
```
╭──────────────────────╮
│ ◎ ○ ○ ░░░░░░░░░░░░░░░│
├──────────────────────┤
│ │
│ │
│ Before │
│ │ ┌────────────────────────────┐ ┌────────────────────────────┐
│ │ │ │ │ │
│ │ │ ShadowNode │ │ LayoutableShadowNode │
└──────────────────────┘ │ │ │ │
└────────────────────────────┘ └────────────────────────────┘
▲ ▲
│ │
╔════════════════════════════╗ ┌────────────────────────────┐
║ ║ │ │
┌────────�║ ConcreteShadowNode<> ║ │ YogaLayoutableShadowNode │
│ ║ ║ │ │
│ ╚════════════════════════════╝ └────────────────────────────┘
│ ▲ ▲
│ │ │
│ │ │
│ │ │
│ │ ╔════════════════════════════╗ │
│ │ ║ ║ │
│ └─────║ ConcreteViewShadowNode<> ║─┘
│ ║ ║
│ ╚════════════════════════════╝
│ ▲
│ │
│ ┌──────────────┴───────────────┐
│ │ │
│ │ │
│ │ │
┌────────────────────────────┐ ┌────────────────────────────┐ ┌────────────────────────────┐
│ │ │ │ │ │
│ TextShadowNode │ │ ViewShadowNode │ │ ParagraphShadowNode │
│ │ │ │ │ │
└────────────────────────────┘ └────────────────────────────┘ └────────────────────────────┘
╭──────────────────────╮
│ ◎ ○ ○ ░░░░░░░░░░░░░░░│
├──────────────────────┤
│ │ ┌────────────────────────────┐
│ │ │ │
│ After │ │ ShadowNode │
│ │ │ │
│ │ └────────────────────────────┘
│ │ ▲
└──────────────────────┘ ┌────────────────────────────────────┤
│ │
│ ╔════════════════════════════╗
┌────────────────────────────┐ ║ ConcreteShadowNode ║
│ │ ║ <Base: ShadowNode> ║
│ LayoutableShadowNode │ ║ ║
│ │ ╚════════════════════════════╝
└────────────────────────────┘ ▲
▲ │
│ ┌────────────────────────────┐
┌────────────────────────────┐ │ │
│ │ │ TextShadowNode │
│ YogaLayoutableShadowNode │ │ │
│ │ └────────────────────────────┘
└────────────────────────────┘
▲
│
╔════════════════════════════╗
║ ConcreteShadowNode ║
║ <Base: ║
║ YogaLayoutableShadowNode> ║
╚════════════════════════════╝
▲
│
╔════════════════════════════╗
║ ║
║ ConcreteViewShadowNode<> ║
║ ║
╚════════════════════════════╝
▲
├─────────────────────────────────────┐
│ │
┌────────────────────────────┐ ┌────────────────────────────┐
│ │ │ │
│ ParagraphShadowNode │ │ ViewShadowNode │
│ │ │ │
└────────────────────────────┘ └────────────────────────────┘
```
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D19963353
fbshipit-source-id: b65c8a5064bdb54ab64f08a8e546aa9e2b5a486b
Summary:
Having the overridden function that returns a different type is apparently not a good idea (and might cause bugs and unexpected behavior), so it was renamed. The function also got a new return type (`const &` instead of `std::shared_ptr`) for simplicity, better performance, and smaller code size.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D19837694
fbshipit-source-id: b7a96424bd040409371724907b3fb3931cd8a2e8
Summary:
This pull request fixes a few `#include`s in the Fabric source:
* Changes `<Glog/logging.h>` to `<glog/logging.h>` in two files, which was an issue for case-sensitive file systems
* In `ParagraphProps.cpp`, changes the include of `attributedstring/primitives.h` from a relative style to a more absolute style.
## Changelog
[Internal] [Fixed] - Fabric: Fix various includes
Pull Request resolved: https://github.com/facebook/react-native/pull/27947
Test Plan: Fabric compiles perfectly, and the Fabric test suite passes.
Reviewed By: shergin
Differential Revision: D19788148
Pulled By: sammy-SC
fbshipit-source-id: 30a297e8aea50c7c7fd5f48745edf04811a6bdbd
Summary:
The reason for this change is that it is the primary root that we want people to be using and the naming should reflect that.
#nocancel
build-break
overriding_review_checks_triggers_an_audit_and_retroactive_review
Changelog: [Internal]
Oncall Short Name: fbobjc_sheriff
Differential Revision: D19431128
fbshipit-source-id: c7208e20ed0f5f5eb6c2849428c09a6d4af9b6f3
Summary:
Replace with `:container_evicting_cache_map`
Changelog: [Internal]
Rename `//xplat/folly:evicting_cache_map` to `//xplat/folly:container_eviciting_cache_map` to follow the new target name conventions
Reviewed By: JoshuaGross, nlutsenko
Differential Revision: D19374914
fbshipit-source-id: fadcaeac20fe61af79e1eecb1be5642f2ef96285
Summary:
This diff changes the signature of ComponentDescriptor constructor to make it simpler and easier to support: now all arguments are passed via struct that contains all these arguments as fields.
Now the ComponentDescriptor constructor accepts three arguments one of which is optional. This causes some confusion and the possibility of bugs in all subclasses that needs to implement a custom constructor. Mostly because in every case we need to ensure that the constructor:
* Accepts and pass down all parameters/arguments;
* Accepts the right types of those parameters (shared vs weak pointers, references vs values).
* Accepts all thee arguments and pass them (including flavor!). We failed this point several times.
Overal that makes the code simpler and allows changing the set of parameters relatively easy. (There is no plan for it!)
Look at the LOC balance: less code!
Changelog: [INTERNAL]
Reviewed By: sammy-SC
Differential Revision: D18548173
fbshipit-source-id: 5d038b135e004f6c054026b3235ed57db99c086d
Summary:
Having automatic defaults/an optional arg for `convertRawProp` has caused way more problems than it is worth. Remove the default argument.
Changelog: [Internal]
Reviewed By: shergin
Differential Revision: D19151594
fbshipit-source-id: 839ec8d138b2c3c083f221a2871582454004648c
Summary:
Motivation: in Marketplace, there's a TextInput in JS that sends { lineHeight: null } to C++. In Paper this was correctly handled as a default value, which just causes that property to not be set. In C++ in Fabric, it was being parsed incorrectly as 0 (because of incorrect defaults being passed into `convertRawProp`), causing the text to not be visible at all. The solution is to make sure that its default value is NaN, which is handled correctly, and causes the Text to render again.
Changelog: [Internal]
Reviewed By: shergin
Differential Revision: D19128049
fbshipit-source-id: fe985428f3ed8b90d56cfa387fbc2d1476d19d36
Summary:
Here the deal with the clamp thing:
A layout engine asks nodes to measure their content providing layout constraints (min and max size) and that would be kind from nodes to return a result that in space of those constraints. Sometimes, satisfying that is an additional separate step of that operation, e.g. if we know that the intrinsic size of some node is `{100, 100}`, we need to clamp that to satisfy constraint `[min: {200, 200}, max: {inf, inf}]`.
In case when we need to cache measure information, it becomes tricky enough when we need to make the cache work most efficiently. For the case of measuring Text, we have an insight: the available maximum width is important, the rest three metrics (maximum height and minimum size) are not.
That means that any changes in those three metrics don't affect the actual measurement result but will affect the final value because the actual value will be clamped by constraints that didn't influence measuring in the first place.
That's why moving clamping outside of a cache is the first step into making it more efficient.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D18848584
fbshipit-source-id: 8feeb3f1a9d0e9691abac8be43639487a626fec3
Summary:
The diff implements a new class called `AttributedStringBox` that represents an object storing a shared `AttributedString` *or* a shared pointer to some opaque platform-specific object that can be used as an attributed string. The class serves two main purposes:
- Represent type-erased attributed string entity (which can be platform-specific or platform-independent);
- Represent a container that can be copied with constant complexity.
Why? Several reasons:
- Sometimes it makes sense to keep an attributed string as a shared resource. This way we don't need to pay for expensive copying and we also implement a copy-on-write semantics on top of that if needed.
- We need to extend a TextLayoutMeasure API to support measuring some platform-specific attributed string implementation to remove the necessity of converting a string back and forth between representations. That's especially important for TextInput because we will need to measure that very efficiently (and the source of measuring, in this case, is a platform attributed string).
In other words, we need something to store inside TextInputState to measure and update very efficiently. The source of this data might be a native TextInput control or a data from React, to represent that kinda object we need this data structure (and interfaces that deal with it).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D18670793
fbshipit-source-id: bc0164f801f28642f7c6da340af12acf33b85d24
Summary:
iOS and other platforms have direct access to C++ StateT structs, whereas Java only has access to a Java map equivalent - state updates from Java can't update complex types, or must incur significant cost to reconstruct large objects from their folly::dynamic representation (not to mention the complexity of implementing the Java-to-C++ struct converters). Thus it's hard for Java to update StateT's with complex types on the C++ side.
This diff makes a minor change to Android's updateState which uses both the folly::dynamic data from Java as well as the previous State, so each StateT can have fields that are read-only from the Java perspective.
Motivation: For AndroidTextInput we need to set params from Java, without being able to send all of the State params from Java.
In this diff, we introduce a new State constructor that takes the previous State value and a folly::dynamic. It is up to each State implementation how the additional parameter will be used.
We migrate every existing component except for AndroidTextInput in this diff.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D18672365
fbshipit-source-id: 4469e0a3c7658c204089c6fed39394979883f124
Summary:
Use a similar setup as Paragraph, and support in Fabric:
- Correct measuring of AndroidTextInput
- Correct display of AndroidTextInput attributed strings
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D18669957
fbshipit-source-id: 84e0ad8021c9edf8219e0c673c781276ca29787d
Summary:
Before this change `AttributedString::Fragment` had two ShadowView objects (`shadowView` and `parentShadowView`). This diff unifies those two things into one. That allows us to save some CPU and memory and makes things a bit simpler. Besides that, now the length of NSAttributedString and AttributedString is now always the same (it's one Unicode character for an attachment for both).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D18607658
fbshipit-source-id: 502ae244e98a52694adc0d646650f8ea0d7922ae
Summary:
We need to call all custom base constructors as part of a custom constuctor of the superclass otherwise C++ will use the default one (and that's not what we want here).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D18607654
fbshipit-source-id: 6e6154ce4b546c0717a8a67e3c497dfcd696ec41
Summary:
We will have several consumers for the measure infra soon (TextInput will use the same TextLayoutManager). It makes sense to move the cache there.
In the future, iOS and Android implementations will probably use a bit different (platform-specific) cache implementations because we will implement the ability to measure "opaque"/platform-specific text containers alongside with normal AttributeStrings.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: mdvacca
Differential Revision: D18445855
fbshipit-source-id: 7b7a65152ac13c74525da695612ae034904e82bf
Summary:
As soon as we have traits infra, it's pretty straight-forward to implement this kinda flag. Without the traits, it's challenging to build that in a performant and elegant way. The challenges of the dynamic_cast-involved approach are:
* How to check for that feature in YogaLayoutableShadowNode?
* Even if we use `dynamic_cast`, to which class to cast?
* We also need to do that in constructor... and dynamic_cast-like checks for `this` don't work in constructors (C++ design limitation).
* How to scale that if we have more classes like Paragraph (and we will have it for sure)?
* Performance.
* Relying on RTTI.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D18525347
fbshipit-source-id: b1915f43ff3fe4364ab6345fb9d1becc591b5a35
Summary:
This diff exposes textBreakStrategy as part of ParagraphAttributes. This is necessary to support the textBreakStrategy prop into Text for Android
Changelog: [Internal]
Reviewed By: JoshuaGross
Differential Revision: D18101404
fbshipit-source-id: e7b665cefe48cf8c764f73a1c51eede16245d1ec
Summary:
This diff extends ParagraphState to expose not only the AttributedString associated to Text components, but also ParagraphAttributes that describes the visual high level props of the Paragraph
Changelog: [Internal]
Reviewed By: JoshuaGross
Differential Revision: D18101407
fbshipit-source-id: 5f5d5ca35cc03e4bf983fd24654be9506d1901a1
Summary:
This is another step on the journey of removing `enable_shared_from_this`.
It's unclear why we used that before but it's clear now that using shared_ptr here is not necessary because all computation around happens inside the single callstack, so by definition we don't have object life-time concerns here.
Changelog: [Internal] Small Fabric-specific optimization.
Reviewed By: sammy-SC
Differential Revision: D17973957
fbshipit-source-id: 09a65c78e22083ed21b041240307f4858379cc60
Summary:
`xplat` targets add different deps based on what platform the target is being built for.
for anything using `fb_xplat`, we can put all ios supermodules in `fbobjc_labels` and all android sms in `fbandroid_labels`
There's some weirdness with python targets like `thrift_gen` in `/xplat/mobileconfig/tools/generator/gen-py/BUCK` that don't have platform-specific labels because the except_for list for `fbandroid` doesn't need the `fbsource//` prefix (see changes in `/ios/isolation/infra.mobileconfig.sm`)
Changelog: [Internal]
Reviewed By: shergin, joshleibsly
Differential Revision: D17884952
fbshipit-source-id: e245364cf515b75682990094d24f789d53b1f3f5
Summary:
Currently, the same ComponentDescriptor class cannot be registered as a responder for components with different names. However, we have marginal cases where we really need it. The examples are `UnimplementedView` or possible universal interop with the classic RN or any other UI framework.
This change adds a special optional argument to ComponentDescript constructor that allows implementing this functionality.
Reviewed By: fkgozali
Differential Revision: D17211915
fbshipit-source-id: 18f59e09fe06b875a8e8975b7b2ab423489238bb
Summary: Support existing, backwards-compatible AndroidTextInput component for minimal support of TextInput on Android.
Reviewed By: shergin, mdvacca
Differential Revision: D17086758
fbshipit-source-id: 25726f22229e0d5dfe96eb36b386a5317601283d
Summary: Storing a strong shared pointer to `ComponentDescriptor` can cause a memory leak. Therefore we enforce all call sides and params to be weak pointers. The only Scheduler preserves a retaining pointer to it (to prevent preliminary deallocation).
Reviewed By: sammy-SC
Differential Revision: D17115540
fbshipit-source-id: fdea7d19f742ff04d5ba5470dd9748a5b226aa7c
Summary:
Remove check whether `measureCache_` is nullptr. It was part of the experiment
which is no longer running.
Reviewed By: shergin
Differential Revision: D16360332
fbshipit-source-id: 2fc8baa93a87754da6255bf1c134901447349f7a
Summary:
All props to Eric Lewis! cc ericlewis
This revert of revert of land of changes originally published in #24873 (with some slight fixes). The change removes usage of LocalData from the `<Text>` component.
After this change we only have ---one (maybe two)--- three components left using LocalData.
Reviewed By: mdvacca
Differential Revision: D15962376
fbshipit-source-id: 19f41109ce9d71ce30d618a45eb2b547a11f29a2
Summary: This fixes the parsing of textAlign prop in BaseTextProps class. The name is textAlign and not alignment.
Reviewed By: JoshuaGross
Differential Revision: D16057477
fbshipit-source-id: a00a472af4c7df0a6cd3d6efb17b30cd9134907b
Summary:
First of all, seems it's the right thing to do. Fabric C++ code is cross-platfrom and should run on *all* platforms including Windows, Linux, and Mac.
While we don't have a real *production* use cases where we need compilation for desktops, having CXX target is really handy for two reasons:
* It simplifies local test running process. Instead of going to `/fbandroid/` and executing something like `buck test fbsource//xplat/js/react-native-github/ReactCommon/fabric/core:coreAndroid` (note the suffix). We can just do `buck test fbsource//xplat/js/react-native-github/ReactCommon/fabric/core:core` everywhere and it works now out of the box. Running tests with "Apple" flavor never worked for me.
* It allows creating synthetic benchmark tests (using Google Benchmark) that can be used as a rough approximation of code micro-optimizations.
Reviewed By: JoshuaGross
Differential Revision: D15608678
fbshipit-source-id: d2449035685dbca6ab983480f5334ec4ac11cd35
Summary:
Updates the paragraph component to use State instead of Local Data, part of the path to a Fabric TextInput 💯
## Changelog
[General] [Changed] - Fabric: Use State instead of Local Data for Paragraph
Pull Request resolved: https://github.com/facebook/react-native/pull/24873
Differential Revision: D15410979
Pulled By: shergin
fbshipit-source-id: 3c9517d2495a64c4dbd213b6efb5ff55287900e3
Summary: QE expired a while ago. Remove the experiment code.
Reviewed By: fkgozali
Differential Revision: D15133830
fbshipit-source-id: 562c3998cc7860497eefa9899dfb6bfcee4fe210
Summary:
With recent changes, simple cloning a component does not mean that underlying Yoga node will be dirtied. Autodirtying is only performed if the child nodes were dirtied and/or if the YGStyles were changed.
That's not the case for Text component which does not have direct layotable children, so we have to call `dirtyLayout` explicitly.
Reviewed By: JoshuaGross
Differential Revision: D14508277
fbshipit-source-id: 2c52d7d40da963a976c7d28a13781cc1755ef591
Summary: Surprisingly, we have some significant amount of text measuring requests where the string is empty. So, there is no need to go to platform specific layer to find that the size of those strings is zero.
Reviewed By: mdvacca
Differential Revision: D14297315
fbshipit-source-id: bf84cf27d5c0893262e8b27da8ff42fc77bcd6c5
Summary: `ParagraphMeasurementCache` was replaced with templated version of itself.
Reviewed By: mdvacca
Differential Revision: D14296515
fbshipit-source-id: 29e370f07baf14b25430f85a06f603907aed5563
Summary:
In React Native there are several use cases where React State is not the only input that affects the component tree. E.g., in case of a <Modal> component, the screen size directly affects the layout of components inside modal; in the case of uncontrolled <TextInput> component, the text inside the input affects the layout of the surrounding components. `State` is a special (legit!) workaround for all those similar cases. Native part of React Native maintains a special shared object between all nodes of the same family and use that during cloning and commits.
See coming commits to know how to use that.
In the near future State will fully replace LocalData concept but for simplicity they both exist for now.
Reviewed By: mdvacca
Differential Revision: D14217184
fbshipit-source-id: 6e018c5b68208d662462013bce0f4e2733d2f673
Summary: All our C++ Fabric tests are cross-platform, so it makes sense to run them for all platforms (especially because platform may behaive differently).
Reviewed By: JoshuaGross, mdvacca
Differential Revision: D13984574
fbshipit-source-id: e384c03c7f9839be38a1910e04ba2f7725abc378
Summary: Nothing really changed; the change is only to better express an original intent.
Reviewed By: JoshuaGross
Differential Revision: D13962464
fbshipit-source-id: f385db8ba8662f2150181e47fc6a2a981f809e96
Summary: The goal is to be able to use MobileConfig params inside of core React Native C++ code. This works on Catalyst iOS now and Wilde; need to add support for FB4A and Catalyst Android.
Reviewed By: fkgozali
Differential Revision: D13883007
fbshipit-source-id: 115fe6cc884d2a0b9ca26dadf867a5f0ae99f262