## Overview
This PR adds the `ref` prop to `<Fragment>` in `react@canary`.
This means this API is ready for final feedback and prepared for a
semver stable release.
## What this means
Shipping Fragment refs to canary means they have gone through extensive
testing in production, we are confident in the stability of the APIs,
and we are preparing to release it in a future semver stable version.
Libraries and frameworks following the [Canary
Workflow](https://react.dev/blog/2023/05/03/react-canaries) should begin
implementing and testing these features.
## Why we follow the Canary Workflow
To prepare for semver stable, libraries should test canary features like
Fragment refs with `react@canary` to confirm compatibility and prepare
for the next semver release in a myriad of environments and
configurations used throughout the React ecosystem. This provides
libraries with ample time to catch any issues we missed before slamming
them with problems in the wider semver release.
Since these features have already gone through extensive production
testing, and we are confident they are stable, frameworks following the
[Canary Workflow](https://react.dev/blog/2023/05/03/react-canaries) can
also begin adopting canary features like Fragment refs.
This adoption is similar to how different Browsers implement new
proposed browser features before they are added to the standard. If a
frameworks adopts a canary feature, they are committing to stability for
their users by ensuring any API changes before a semver stable release
are opaque and non-breaking to their users.
Apps not using a framework are also free to adopt canary features like
Fragment refs as long as they follow the [Canary
Workflow](https://react.dev/blog/2023/05/03/react-canaries), but we
generally recommend waiting for a semver stable release unless you have
the capacity to commit to following along with the canary changes and
debugging library compatibility issues.
Waiting for semver stable means you're able to benefit from libraries
testing and confirming support, and use semver as signal for which
version of a library you can use with support of the feature.
## Docs
Check out the ["React Labs: View Transitions, Activity, and
more"](https://react.dev/blog/2025/04/23/react-labs-view-transitions-activity-and-more#fragment-refs)
blog post, and [the new docs for Fragment
refs`](https://react.dev/reference/react/Fragment#fragmentinstance) for
more info.
The View Transition docs were unclear about this but apparently the
`finished` promise never settles if the animation never started. So if
there's an error that rejects the `ready` promise, we'll never run the
clean up which can cause it to stall.
Fixes#34662.
However, ultimately that is caused by Chrome stalling our default
`onDefaultTransitionIndicator` but it should be unblocked after 10
seconds, not a minute.
In Fizz and Fiber we emit hints for suspensey images and CSS as soon as
we discover them during render. At the beginning of the stream. This
adds a similar capability when a Host Component is known to be a Host
Component during the Flight render.
The client doesn't know that these resources are in the payload until it
parses that particular component which is lazy. So they need to be
hoisted with hints. We detect when these are rendered during Flight and
add them as hints. That allows you to consume a Flight payload to
preload prefetched content without having to render it.
`<link rel="preload">` can be hoisted more or less as is.
`<link rel="stylesheet">` we preload but we don't actually insert them
anywhere until they're rendered. We do these even for non-suspensey
stylesheets since we know that when they're rendered they're going to
start loading even if they're not immediately used. They're never lazy.
`<img src>` we only preload if they follow the suspensey image pattern
since otherwise they may be more lazy e.g. by if they're in the
viewport. We also skip if they're known to be inside `<picture>`. Same
as Fizz. Ideally this would preload the other `<source>` but it's
tricky.
The downside of this is that you might conditionally render something in
only one branch given a client component. However, in that case you're
already eagerly fetching the server component's data in that branch so
it's not too much of a stretch that you want to eagerly fetch the
corresponding resources as well. If you wanted it to be lazy, you
should've done a lazy fetch of the RSC.
We don't collect hints when any of these are wrapped in a Client
Component. In those cases you might want to add your own preload to a
wrapper Shared Component.
Everything is skipped if it's known to be inside `<noscript>`.
Note that the format context is approximate (see #34601) so it's
possible for these hints to overfetch or underfetch if you try to trick
it. E.g. by rendering Server Components inside a Client Component that
renders `<noscript>`.
---------
Co-authored-by: Josh Story <josh.c.story@gmail.com>
Flight doesn't have any semantically sound notion of a parent context.
That's why we removed Server Context. Each root can really start
anywhere in the tree when you refetch subtrees. Additionally when you
dedupe elements they can end up in multiple different parent contexts.
However, we do have a DEV only version of this with debugTask being
tracked for the nearest parent element to track the context of
properties inside of it.
To apply certain DOM specific hints and optimizations when you render
host components we need some information of the context. This is usually
very local so doesn't suffer from the likelihood that you refetch in the
middle. We'll also only use this information for optimistic hints and
not hard semantics so getting it wrong isn't terrible.
```
<picture>
<img />
</picture>
<noscript>
<p>
<img />
</p>
</noscript>
```
For example, in these cases we should exclude preloading the image but
we have to know if that's the scope we're in.
We can easily get this wrong if they're split or even if they're wrapped
in client components that we don't know about like:
```
<NoScript>
<p>
<img />
</p>
</NoScript>
```
However, getting it wrong in either direction is not the end of the
world. It's about covering the common cases well.
We should favor outlining a boundary if it contains Suspensey CSS or
Suspensey Images since then we can load that content separately and not
block the main content. This also allows us to animate the reveal.
For example this should be able to animate the reveal even though the
actual HTML content isn't large in this case it's worth outlining so
that the JS runtime can choose to animate this reveal.
```js
<ViewTransition>
<Suspense>
<img src="..." />
</Suspense>
</ViewTransition>
```
For Suspensey Images, in Fizz, we currently only implement the suspensey
semantics when a View Transition is running. Therefore the outlining
only applies if it appears inside a Suspense boundary which might
animate. Otherwise there's no point in outlining. It is also only if the
Suspense boundary itself might animate its appear and not just any
ViewTransition. So the effect is very conservative.
For CSS it applies even without ViewTransition though, since it can help
unblock the main content faster.
Stacked on #34546.
Same as #34538 but for gestures.
Includes various fixes.
This shows how it ends with a Transition when you release in the
committed state. Note how the Animation of the Gesture continues until
the Transition is done so that the handoff is seamless.
<img width="853" height="134" alt="Screenshot 2025-09-20 at 7 37 29 PM"
src="https://github.com/user-attachments/assets/6192a033-4bec-43b9-884b-77e3a6f00da6"
/>
Stacked on #34522.
<img width="1025" height="200" alt="Screenshot 2025-09-19 at 6 37 28 PM"
src="https://github.com/user-attachments/assets/f25900f6-6503-48b1-876d-bd6697a29c6f"
/>
We already cover the time between "Starting Animation" and "Remaining
Effects" as "Animating". However, if the effects are forced then we can
still be animating after that. This fills in that gap.
This also fills in the gap if another render starts before the animation
finishes on the same track. It'll mark the blank space between the
previous render finishing and the next render starting as "Animating".
This should correspond roughly to the native "Animations" track.
Stacked on #34511.
We currently log all Suspended Commit as "Suspended on Images or CSS"
but it can really be other reasons too now. Like waiting on the previous
View Transition. This allows the host config configure this reason.
Now when one animation starts before another one finishes we log that as
"Waiting for the previous Animation".
<img width="592" height="257" alt="Screenshot 2025-09-17 at 11 53 45 PM"
src="https://github.com/user-attachments/assets/817af8b5-37ae-46d8-bfd1-cd3fc637f3f3"
/>
Stacked on #34510.
The "Commit" phase for a View Transition starts before the snapshot
phase (before mutation) and then stretches into the async gap of
`startViewTransition`, encompasses the mutation phase inside of its
update callback and finally the layout phase.
However, between the mutation phase and the layout phase we may suspend
the start of the view transition on fonts and/or images. In that case we
now split the Commit phase into first one before we suspend and then we
log "Waiting for Images and/or Fonts" and then another Commit phase
around the layout effects.
<img width="897" height="119" alt="Screenshot 2025-09-16 at 11 37 26 PM"
src="https://github.com/user-attachments/assets/0fe21388-bb48-4456-a594-62227d12d9b7"
/>
Stacked on #34486.
If we gave up on loading suspensey images for blocking the commit (e.g.
due to #34481), we can still block the view transition from committing
to allow an animation to include the image from the start.
At this point we have more information about the layout so we can
include only the images that are within viewport in the calculation
which may end up with a different answer.
This only applies when we attempt to run an animation (e.g. something
mutated inside a `<ViewTransition>` in a Transition). We could attempt a
`startViewTransition` if we gave up on the suspensey images just so that
we could block it even if no animation would be running.
However, this point the screen is frozen and you can no longer have sync
updates interrupt so ideally we would have already blocked the commit
from happening in the first place.
The reason to have two points where we block is that ideally we leave
the UI responsive while blocking, which blocking the commit does. In the
simple case of all images or a single image being within the viewport,
that's favorable. By combining the techniques we only end up freezing
the screen in the special case that we had a lot of images added outside
the viewport and started an animation with some image inside the
viewport (which presumably is about to finish anyway).
Stacked on #34481.
We currently track the suspended state temporarily with a global which
is safe as long as we always read it during a sync pass. However, we
sometimes read it in closures and then we have to be carefully to pass
the right one since it's possible another commit on a different root has
started at that point. This avoids this footgun.
Another reason to do this is that I want to read it in
`startViewTransition` which is in an async gap after which point it's no
longer safe. So I have to pass that through the `commitRoot` bound
function.
Stacked on #34478.
In general we don't like to deal with timeouts in suspense world. We've
had that in the past but in general it doesn't work well because if you
have a timeout and then give up you made everything wait longer for no
benefit at the end. That's why the recommendation is to remove a
Suspense boundary if you expect it to be fast and add one if you expect
it to be slow. You have to estimate as the developer.
Suspensey images suffer from this same problem. We want to apply
suspensey images to as much as possible so that it's the default to
avoid flashing because if just a few images flash it's still almost as
bad as all of them. However, we do know that it's also very common to
use images and on a slow connection or many images, it's not worth it so
we have the timeout to eventually give up.
However, this means that in cases that are always slow or connections
that are always slow, you're always punished for no reason.
Suspensey images is mainly a polish feature to make high end experiences
on high end connections better but we don't want to unnecessarily punish
all slow connections in the process or things like lots of images below
the viewport.
This PR adds an estimate for whether or not we'll likely be able to load
all the images within the timeout on a high end enough connection. If
not, we'll still do a short suspend (unless we've already exceeded the
wait time adjusted for #34478) to allow loading from cache if available.
This estimate is based on two heuristics:
1) We compute an estimated bandwidth available on the current device in
mbps. This is computed from performance entries that have loaded static
resources already on the site. E.g. this can be other images, css, or
scripts. We see how long they took. If we don't have any entries (or if
they're all cross-origin in Safari) we fallback to
`navigator.connection.downlink` in Chrome or a 5mbps default in
Firefox/Safari.
2) To estimate how many bytes we'll have to download we use the
width/height props of the img tag if available (or a 100 pixel default)
times the device pixel ratio. We assume that a good img implementation
downloads proper resolution image for the device and defines a
width/height up front to avoid layout trash. Then we estimate that it
takes about 0.25 bytes per pixel which is somewhat conservative
estimate.
This is somewhat conservative given that the image could've been
preloaded and be better compressed.
So it really only kicks in for high end connections that are known to
load fast.
In a follow up, we can add an additional wait for View Transitions that
does the same estimate but only for the images that turn out to be in
viewport.
Currently suspensey images doesn't account for how long we've already
been waiting. This means that you can for example wait for 300ms for the
throttle + 500ms for the images. If a Transition takes a while to
complete you can also wait that time + an additional 500ms for the
images.
This tracks the start time of a Transition so that we can count the
timeout starting from when the user interacted or when the last fallback
committed (which is where the 300ms throttle is computed from). Creating
a single timeline.
This also moves the timeout to a central place which I'll use in a
follow up.
Fixes#34098.
There's an issue in Chrome where the `InvalidStateError` always has the
same error message. The spec doesn't specify the error message to use
but it's more useful to have a specific one for each case like Safari
does.
One reason it's better to have a specific error message is because the
browser console is not the main surface that people look for errors.
Chrome relies on a separate log also in the console. Frameworks has
built-in error dialogs that pop up first and that's where you see the
error and that dialog can't show something specific. Additionally, these
errors can't log something specific to servers in production logging. So
this is a bad strategy.
It's not good to have those error dialogs pop up for non-actionable
errors like when it doesn't start because the document was hidden. Since
we don't have more specific information we have no choice but to hide
all of them. This includes actionable things like duplicate names
(although we also have a React specific warning for that in the common
case).
This adds `experimental_scrollIntoView(alignToTop)`. It doesn't yet
support `scrollIntoView(options)`.
Cases:
- No host children: Without host children, we represent the virtual
space of the Fragment by attempting to scroll to the nearest edge by
using its siblings. If the preferred sibling is not found, we'll try the
other side, and then the parent.
- 1 or more host children: In order to handle the case of children
spread between multiple scroll containers, we scroll to each child in
reverse order based on the `alignToTop` flag.
Due to the complexity of multiple scroll containers and dealing with
portals, I've added this under a separate feature flag with an
experimental prefix. We may stabilize it along with the other APIs, but
this allows us to not block the whole feature on it.
This PR was previously implementing a much more complex approach to
handling multiple scroll containers and portals. We're going to start
with the simple loop and see if we can find any concrete use cases where
that doesn't suffice. 01f31d43013ba7f6f54fd8a36990bbafc3c3cc68 is the
diff between approaches here.
This update was a bit more involved.
- `React$Component` was removed, I replaced it with Flow component
types.
- Flow removed shipping the standard library. This adds the environment
libraries back from `flow-typed` which seemed to have changed slightly
(probably got more precise and less `any`s). Suppresses some new type
errors.
After an easy couple version with #34252, this version is less flexible
(and safer) on inferring exported types mainly.
We require to annotate some exported types to differentiate between
`boolean` and literal `true` types, etc.
Before the first rAF, we don't know if there has been other paints
before this and if so when. (We could get from performance observer.) We
can assume that it's not earlier than 0 so we used delay up until the
throttle time starting from zero but if the first paint is about to
happen that can be very soon after.
Instead, this reveals it during the next paint which should let us be
able to get into the first paint. If we can trust `rel="expect"` to have
done its thing we should schedule our raf before first paint but ofc
browsers can cheat and paint earlier if they want to.
If we're wrong, this is at least more batched than doing it
synchronously. However it will mean that things might get more flashy
than it should be if it would've been throttled. An alternative would be
to always throttle first reveal.
The theory here is that when we reveal a boundary coming from the server
we want to paint that before hydrating it. Hydration gets scheduled in a
macrotask with the scheduler but it's in theory possible that it runs
before the paint. If that's the case, then the JS that runs before
yielding during hydration might slightly delay the paint and we might
miss a window to skip the previous paint.
Stacked on https://github.com/facebook/react/pull/34069
Same basic semantics as the react-dom for determining document position
of a Fragment compared to a given node. It's simpler here because we
don't have to deal with inserted nodes or portals. So we can skip a
bunch of the validation logic.
The logic for handling empty fragments is the same so I've split out
`compareDocumentPositionForEmptyFragment` into a shared module. There
doesn't seem to be a great place to put shared DOM logic between Fabric
and DOM configs at the moment. There may be more of this coming as we
add more and more DOM APIs to RN.
For testing I've written Fantom tests internally which pass the basic
cases on this build. The renderer we have configured for Fabric tests in
the repo doesn't support the Element APIs we need like
`compareDocumentPosition`.
Found a couple of issues while integrating
FragmentInstance#compareDocumentPosition into Fabric.
1. Basic checks of nested host instances were inaccurate. For example,
checking the first child of the first child of the Fragment would not
return CONTAINED_BY.
2. Then fixing that logic exposed issues with Portals. The DOM
positioning relied on the assumption that the first and last top-level
children were in the same order as the Fiber tree. I added additional
checks against the parent's position in the DOM, and special cased a
portaled Fragment by getting its DOM parent from the child instance,
rather than taking the instance from the Fiber return. This should be
accurate in more cases. Though its still a guess and I'm not sure yet
I've covered every variation of this. Portals are hard to deal with and
we may end up having to push more results towards
IMPLEMENTATION_SPECIFIC if accuracy is an issue.
View Transitions has this annoying quirk where it adds `width` and
`height` to keyframes automatically when generating keyframes even when
it's not needed. This causes them to deopt from running on the
compositor thread in both Chrome and Safari. @bramus has a [good article
on
it](https://www.bram.us/2025/02/07/view-transitions-applied-more-performant-view-transition-group-animations/).
In React we can automatically rewrite the keyframes when we're starting
a View Transition to drop the `width` and `height` from the keyframes
when they have the same value and the same value as the pseudo element.
To compare it against the pseudo element we first apply the new
keyframes without the width/height and then read it back to see if it
has changed. For gestures, we have already cancelled the previous
animation so we can just read out from that.
The React API is just that we now accept this protocol as an alternative
to a native `AnimationTimeline` to be passed to
`startGestureTransition`. This is specifically the DOM version.
```js
interface CustomTimeline {
currentTime: number;
animate(animation: Animation): void | (() => void);
}
```
Instead, of passing this to the `Animation` that we start to control the
View Transition keyframes, we instead inverse the control and pass the
`Animation` to this one. It lets any custom implementation drive the
updates. It can do so by updating the time every frame or letting it run
a time based animation (such as momentum scroll).
In this case I added a basic polyfill for `ScrollTimeline` in the
example but we'll need a better one.
Adding throttling or delaying on images, can obviously impact metrics.
However, it's all in the name of better actual user experience overall.
(Note that it's not strictly worse even for metric. Often it's actually
strictly better due to less work being done overall thanks to batching.)
Metrics can impact things like search ranking but I believe this is on a
curve. If you're already pretty good, then a slight delay won't suddenly
make you rank in a completely different category. Similarly, if you're
already pretty bad then a slight delay won't make it suddenly way worse.
It's still in the same realm. It's just one weight of many. I don't
think this will make a meaningful practical impact and if it does,
that's probably a bug in the weights that will get fixed.
However, because there's a race to try to "make everything green" in
terms of web vitals, if you go from green to yellow only because of some
throttling or suspensey images, it can feel bad. Therefore this
implements a heuristic where if the only reason we'd miss a specific
target is because of throttling or suspensey images, then we shorten the
timeout to hit the metric. This is a worse user experience because it
can lead to extra flashing but feeling good about "green" matters too.
If you then have another reveal that happens to be the largest
contentful paint after that, then that's throttled again so that it
doesn't become flashy after that. If you've already missed the deadline
then you're not going to hit your metric target anyway. It can affect
average but not median.
This is mainly about LCP. It doesn't affect FCP since that doesn't have
a throttle. If your LCP is the same as your FCP then it also doesn't
matter.
We assume that `performance.now()`'s zero point starts at the "start of
the navigation" which makes this simple. Even if we used the
`PerformanceNavigationTiming` API it would just tell us the same thing.
This only implements for Fizz since these metrics tend to currently only
by tracked for initial loads, but with soft navs tracking we could
consider implementing the same for Fiber throttles.
We want to make sure that we can block the reveal of a well designed
complete shell reliably. In the Suspense model, client transitions don't
have any way to implicitly resolve. This means you need to use Suspense
or SuspenseList to explicitly split the document. Relying on implicit
would mean you can't add a Suspense boundary later where needed. So we
highly encourage the use of them around large content.
However, if you have constructed a too large shell (e.g. by not adding
any Suspense boundaries at all) then that might take too long to render
on the client. We shouldn't punish users (or overzealous metrics
tracking tools like search engines) in that scenario.
This opts out of render blocking if the shell ends up too large to be
intentional and too slow to load. Instead it deopts to showing the
content split up in arbitrary ways (browser default). It only does this
for SSR, and not client navs so it's not reliable.
In fact, we issue an error to `onError`. This error is recoverable in
that the document is still produced. It's up to your framework to decide
if this errors the build or just surface it for action later.
What should be the limit though? There's a trade off here. If this limit
is too low then you can't fit a reasonably well built UI within it
without getting errors. If it's too high then things that accidentally
fall below it might take too long to load.
I came up with 512kB of uncompressed shell HTML. See the comment in code
for the rationale for this number. TL;DR: Data and theory indicates that
having this much content inside `rel="expect"` doesn't meaningfully
change metrics. Research of above-the-fold content on various websites
indicate that this can comfortable fit all of them which should be
enough for any intentional initial paint.
Block the view transition on suspensey images Up to 500ms just like the
client.
We can't use `decode()` because a bug in Chrome where those are blocked
on `startViewTransition` finishing we instead rely on sync decoding but
also that the image is live when it's animating in and we assume it
doesn't start visible.
However, we can block the View Transition from starting on the `"load"`
or `"error"` events.
The nice thing about blocking inside `startViewTransition` is that we
have already done the layout so we can only wait on images that are
within the viewport at this point. We might want to do that in Fiber
too. If many image doesn't have fixed size but need to load first, they
can all end up in the viewport. We might consider only doing this for
images that have a fixed size or only a max number that doesn't have a
fixed size.
Alternative to #33421. The difference is that this also adds an
underscore between the "R" and the ID.
The reason we wanted to use special characters is because we use the
full spectrum of A-Z 0-9 in our ID generation so we can basically
collide with any common word (or anyone using a similar algorithm,
base64 or even base16). It's a little less likely that someone would put
`_R_` specifically unless you generate like two IDs separated by
underscore.
fixes https://github.com/facebook/react/issues/32449
This is my first time touching this code. There are multiple systems in
place here and I wouldn't be surprised to learn that this has to be
handled in some other areas too. I have found some other style-related
code areas but I had no time yet to double-check them.
cc @gnoff
This is the same technique we do for the client except we don't check
whether this is newly created font loading to keep code small.
Unfortunately, we can't use this technique for Suspensey images. They'll
need to block before we call `startViewTransition` in a separate
refactor. This is due to a bug in Chrome where `img.decode()` doesn't
resolve until `startViewTransition` does.
Stacked on #33330.
This walks the element tree to activate the various classes under
different scenarios. There are some edge case things that are a little
different since we can't express every scenario without virtual nodes.
The main thing that's still missing though is avoiding animating updates
if it can be contained to a layout or enter/exit/share if they're out of
the viewport. I.e. layout stuff.
Follow up to #33293.
This solves a race condition when boundaries are added to the batch
after the `startViewTransition` call.
This doesn't matter yet but it will once we start assigning names before
the `startViewTransition` call.
A possible alternative solution might be to ensure the names are added
synchronously in the event that adds to the batch. It's possible to keep
adding to a batch until the snapshot has happened.
I believe that these mean the same thing. We don't have to emit the
attribute if it's `none` for these cases because if there is no matching
scenario we won't apply the animation in this case.
The only case where we have to emit `none` in the attribute is for
`vt-update` because those can block updates from propagating upwards.
When a new child of a fragment instance is inserted, we need to notify
the instance to keep any relevant tracking up to date. For example, we
automatically observe the new child with any active
IntersectionObserver.
For mutable renderers (DOM), we reuse the existing traversal in
`commitPlacement` that does the insertions for HostComponents. Immutable
renderers (Fabric) exit this path before the traversal though, so
currently we can't notify the fragment instances.
Here I've created a separate traversal in `commitPlacement`,
specifically for immutable renders when `enableFragmentRefs` is on.
We have many cases internally where the `containerInstance` resolves to
a comment node. `restoreRootViewTransitionName` is called when
`enableViewTransition` is on, even without introducing a
`<ViewTransition />`. So that means it can crash pages because
`containerInstance.style` is `undefined` just by turning on the flag.
This skips cancel/restore of root view transition name if a comment node is the root.
Stacked on #33308.
For "together" mode, we can be a self-blocking row that adds all its
boundaries to the blocked set, but there's no parent row that unblocks
it.
A particular quirk of this mode is that it's not enough to just unblock
them all on the server together. Because if one boundary downloads all
its html and then issues a complete instruction it'll appear before the
others while streaming in. What we actually want is to reveal them all
in a single batch.
This implementation takes a short cut by unblocking the rows in
`flushPartialBoundary`. That ensures that all the segments of every
boundary has a chance to flush before we start emitting any of the
complete boundary instructions. Once the last one unblocks, all the
complete boundary instructions are queued. Ideally this would be a
single `<script>` tag so that they can't be split up even if we get a
chunk containing some of them.
~A downside of this approach is that we always outline these boundaries.
We could inline them if they all complete before the parent flushes.
E.g. by checking if the row is blocked only by its own boundaries and if
all the boundaries would fit without getting outlined, then we can
inline them all at once.~ I went ahead and did this because it solves an
issue with `renderToString` where it doesn't support the script runtime
so it can only handle this if inlined.
When needed.
For the external runtime we always include this wrapper.
For others, we only include it if we have an ViewTransitions affecting.
If we discover the ViewTransitions late, then we can upgrade an already
emitted instruction.
This doesn't yet do anything useful with it, that's coming in a follow
up. This is just the mechanism for how it gets installed.
Stacked on #33194 and #33200.
When Suspense boundaries reveal during streaming, the Fizz runtime will
be responsible for animating the reveal if necessary (not in this PR).
However, for the future runtime to know what to do it needs to know
about the `<ViewTransition>` configuration to apply.
Ofc, these are virtual nodes that disappear from the HTML. We could
model them as comments like we do with other virtual nodes like Suspense
and Activity. However, that doesn't let us target them with
querySelector and CSS (for no-JS transitions). We also don't have to
model every ViewTransition since not every combination can happen using
only the server runtime. So instead this collapses `<ViewTransition>`
and applies the configuration to the inner DOM nodes.
```js
<ViewTransition name="hi">
<div />
<div />
</ViewTransition>
```
Becomes:
```html
<div vt-name="hi" vt-update="auto"></div>
<div vt-name="hi_1" vt-update="auto"></div>
```
I use `vt-` prefix as opposed to `data-` to keep these virtual
attributes away from user specific ones but we're effectively claiming
this namespace.
There are four triggers `vt-update`, `vt-enter`, `vt-exit` and
`vt-share`. The server resolves which ones might apply to this DOM node.
The value represents the class name (after resolving
view-transition-type mappings) or `"auto"` if no specific class name is
needed but this is still a trigger.
The value can also be `"none"`. This is different from missing because
for example an `vt-update="none"` will block mutations inside it from
triggering the boundary where as a missing `vt-update` would bubble up
to be handled by a parent.
`vt-name` is technically only necessary when `vt-share` is specified to
find a pair. However, since an explicit name can also be used to target
specific CSS selectors, we include it even for other cases.
We want to exclude as many of these annotations as possible.
`vt-enter` can only affect the first DOM node inside a Suspense
boundary's content since the reveal would cause it to enter but nothing
deeper inside. Similarly `vt-exit` can only affect the first DOM node
inside a fallback. So for every other case we can exclude them. (For
future MPA ViewTransitions of the whole document it might also be
something we annotate to children inside the `<body>` as well.) Ideally
we'd only include `vt-enter` for Suspense boundaries that actually
flushed a fallback but since we prepare all that content earlier it's
hard to know.
`vt-share` can be anywhere inside an fallback or content. Technically we
don't have to include it outside the root most Suspense boundary or for
boundaries that are inlined into the root shell. However, this is tricky
to detect. It would also not be correct for future MPA ViewTransitions
because in that case the shared scenario can affect anything in the two
documents so it needs to be in every node everywhere which is
effectively what we do. If a `share` class is specified but it has no
explicit name, we can exclude it since it can't match anything.
`vt-update` is only necessary if something below or a sibling might
update like a Suspense boundary. However, since we don't know when
rendering a segment if it'll later asynchronously add a Suspense
boundary later we have to assume that anywhere might have a child. So
these are always included. We collapse to use the inner most one when
directly nested though since that's the one that ends up winning.
There are some weird edge cases that can't be fully modeled by the lack
of virtual nodes.
Removes the `isFallback` flag on Tasks and tracks it on the
formatContext instead.
Less memory and avoids passing and tracking extra arguments to all the
pushStartInstance branches that doesn't need it.
We'll need to be able to track more Suspense related contexts on this
for View Transitions anyway.
