-
Notifications
You must be signed in to change notification settings - Fork 46.9k
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
Delete flushSuspenseFallbacksInTests flag #18596
Delete flushSuspenseFallbacksInTests flag #18596
Conversation
This was meant to be a temporary hack to unblock the `act` work, but it quickly spread throughout our tests. What it's meant to do is force fallbacks to flush inside `act` even in Concurrent Mode. It does this by wrapping the `setTimeout` call in a check to see if it's in an `act` context. If so, it skips the delay and immediately commits the fallback. Really this is only meant for our internal React tests that need to incrementally render. Nobody outside our team (and Relay) needs to do that, yet. Even if/when we do support that, it may or may not be with the same `flushAndYield` pattern we use internally. However, even for our internal purposes, the behavior isn't right because a really common reason we flush work incrementally is to make assertions on the "suspended" state, before the fallback has committed. There's no way to do that from inside `act` with the behavior of this flag, because it causes the fallback to immediately commit. This has led us to *not* use `act` in a lot of our tests, or to write code that doesn't match what would actually happen in a real environment. What we really want is for the fallbacks to be flushed at the *end` of the `act` scope. Not within it. This only affects the noop and test renderer versions of `act`, which are implemented inside the reconciler. Whereas `ReactTestUtils.act` is implemented in "userspace" for backwards compatibility. This is fine because we didn't have any DOM Suspense tests that relied on this flag; they all use test renderer or noop. In the future, we'll probably want to move always use the reconciler implementation of `act`. It will not affect the prod bundle, because we currently only plan to support `act` in dev. Though we still haven't completely figured that out. However, regardless of whether we support a production `act` for users, we'll still need to write internal React tests in production mode. For that use case, we'll likely add our own internal version of `act` that assumes a mock Scheduler and might rely on hacks that don't 100% align up with the public one.
This pull request is automatically built and testable in CodeSandbox. To see build info of the built libraries, click here or the icon next to each commit SHA. Latest deployment of this branch, based on commit 3a12db4:
|
if (flushMockScheduler !== undefined) { | ||
const prevIsFlushing = isFlushingAct; | ||
isFlushingAct = true; | ||
try { | ||
return flushMockScheduler(); | ||
} finally { | ||
isFlushingAct = prevIsFlushing; | ||
} | ||
} else { | ||
// No mock scheduler available. However, the only type of pending work is | ||
// passive effects, which we control. So we can flush that. | ||
const prevIsFlushing = isFlushingAct; | ||
isFlushingAct = true; | ||
try { | ||
let didFlushWork = false; | ||
while (flushPassiveEffects()) { | ||
didFlushWork = true; | ||
} | ||
return didFlushWork; | ||
} finally { | ||
isFlushingAct = prevIsFlushing; | ||
} | ||
} | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is the relevant change. I set isFlushingAct
while the work is being flushed at the end. Then I check this where we used to check the "force fallback" feature flag.
Details of bundled changes.Comparing: f3f3d77...3a12db4 react-art
react-test-renderer
react-dom
react-native-renderer
react-reconciler
ReactDOM: size: 0.0%, gzip: -0.0% Size changes (experimental) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Out of curiosity, why do we have 3 different versions of act
?
Details of bundled changes.Comparing: f3f3d77...3a12db4 react-art
react-test-renderer
react-native-renderer
react-dom
react-reconciler
ReactDOM: size: 0.0%, gzip: -0.0% Size changes (stable) |
913b818
to
2d24402
Compare
@lunaruan One is built into the reconciler and relies on some internals. That's the implementation that's used by React Test Renderer and React Noop. There's another one implemented inside ReactTestUtils. It has to account for backwards compatibility with older versions of React. So it has lots of fallback mechanisms to try to work as best as possible for existing users. It's really only used to test components that include In Concurrent Mode, it goes further and flushes all pending React tasks, including rendering and suspended fallbacks. That can't be properly implemented without relying on internals. So they diverge slightly. Long term plan is to move everyone to the built-in implementation (the one I changed in this PR). |
2d24402
to
7e27e5d
Compare
c87f73e
to
3a12db4
Compare
This was meant to be a temporary hack to unblock the
act
work, but it quickly spread throughout our tests.What it's meant to do is force fallbacks to flush inside
act
even in Concurrent Mode. It does this by wrapping thesetTimeout
call in a check to see if it's in anact
context. If so, it skips the delay and immediately commits the fallback.Really this is only meant for our internal React tests that need to incrementally render. Nobody outside our team (and Relay) needs to do that, yet. Even if/when we do support that, it may or may not be with the same
flushAndYield
pattern we use internally.However, even for our internal purposes, the behavior isn't right because a really common reason we flush work incrementally is to make assertions on the "suspended" state, before the fallback has committed. There's no way to do that from inside
act
with the behavior of this flag, because it causes the fallback to immediately commit. This has led us to not useact
in a lot of our tests, or to write code that doesn't match what would actually happen in a real environment.What we really want is for the fallbacks to be flushed at the end of the
act
scope. Not within it.This only affects the noop and test renderer versions of
act
, which are implemented inside the reconciler. WhereasReactTestUtils.act
is implemented in "userspace" for backwards compatibility. This is fine because we didn't have any DOM Suspense tests that relied on this flag; they all use test renderer or noop.In the future, we'll probably want to move always use the reconciler implementation of
act
. It will not affect the prod bundle, because we currently only plan to supportact
in dev. Though we still haven't completely figured that out. However, regardless of whether we support a productionact
for users, we'll still need to write internal React tests in production mode. For that use case, we'll likely add our own internal version ofact
that assumes a mock Scheduler and might rely on hacks that don't 100% align up with the public one.