CI builds should always use the `--frozen-lockfile` option. It will fail the build if the lockfile is out-of-date:
> If you need reproducible dependencies, which is usually the case with the continuous integration systems, you should pass --frozen-lockfile flag.
(https://yarnpkg.com/en/docs/cli/install/)
* [schedule] Use linked list instead of queue and map for storing cbs
NOTE: This PR depends on https://github.com/facebook/react/pull/12880
and https://github.com/facebook/react/pull/12884
Please review those first, and after they land Flarnie will rebase on
top of them.
---
**what is the change?:**
See title
**why make this change?:**
This seems to make the code simpler, and potentially saves space of
having an array and object around holding references to the callbacks.
**test plan:**
Run existing tests
* minor style improvements
* refactor conditionals in cancelScheduledWork for increased clarity
* Remove 'canUseDOM' condition and fix some flow issues w/callbackID type
**what is the change?:**
- Removed conditional which fell back to 'setTimeout' when the
environment doesn't have DOM. This appears to be an old polyfill used
for test environments and we don't use it any more.
- Fixed type definitions around the callbackID to be more accurate in
the scheduler itself, and more loose in the React code.
**why make this change?:**
To get Flow passing, simplify the scheduler code, make things accurate.
**test plan:**
Run tests and flow.
* Rewrite 'cancelScheduledWork' so that Flow accepts it
**what is the change?:**
Adding verification that 'previousCallbackConfig' and
'nextCallbackConfig' are not null before accessing properties on them.
Slightly concerned because this implementation relies on these
properties being untouched and correct on the config which is passed to
'cancelScheduledWork' but I guess we already rely heavily on that for
this whole approach. :\
**why make this change?:**
To get Flow passing.
Not sure why it passed earlier and in CI, but now it's not.
**test plan:**
`yarn flow dom` and other flow tests, lint, tests, etc.
* ran prettier
* Put back the fallback implementation of scheduler for node environment
**what is the change?:**
We had tried removing the fallback implementation of `scheduler` but
tests reminded us that this is important for supporting isomorphic uses
of React.
Long term we will move this out of the `schedule` module but for now
let's keep things simple.
**why make this change?:**
Keep things working!
**test plan:**
Ran tests and flow
* Shorten properties stored in objects by sheduler
**what is the change?:**
`previousScheduledCallback` -> `prev`
`nextScheduledCallback` -> `next`
**why make this change?:**
We want this package to be smaller, and less letters means less code
means smaller!
**test plan:**
ran existing tests
* further remove extra lines in scheduler
* Moved actual time fields from Profiler stateNode to Fiber
* Record actual time for all Fibers within a ProfileMode tree
* Changed how profiler accumulates time
This change gives up on accumulating time across renders of different priority, but in exchange- simplifies how the commit phase (reset) code works, and perhaps also makes the profiling code more compatible with future resuming behavior
* Adding 'schedule' fixture
**what is the change?:**
We need to test the `schedule` module against real live browser APIs. As
a quick solution we're writing a fixture for using in manual testing.
Later we plan on adding automated browser testing, using this or a
similar fixture as the test page.
**why make this change?:**
To further solidify test coverage for `schedule` before making further
improvements/refactors to the module.
**test plan:**
`open fixtures/schedule/index.html` and inspect the results. It should
be clear that things pass.
We also temporarily broke the scheduler and verified that this fixture
demonstrates the problems.
**issue:**
Internal task T29442940
* Made fixture tests display red or green border depending on pass/fail
**what is the change?:**
Added red/green solid/dashed border for test results when using the
schedule fixture.
We also tweaked the timing of the last test because it was on the line
in terms of whether it passed or failed.
**why make this change?:**
To make it faster to use the fixture - it takes more time to read
through the results line by line and check that they match what is
expected.
**test plan:**
Looked at the fixture, and also tried modifying a test to show what it
looks like when something fails.
**what is the change?:**
Renamed some methods, and made a method to advance a frame in the test
environment.
**why make this change?:**
We often need to simulate a frame passing with some amount of idle time
or lack of idle time, and the new method makes it easier to write that
out.
**test plan:**
Run the updated tests.
Also temporarily tried breaking the scheduler and verified that the
tests will fail.
**issue:**
See internal task T29442940
* add legacy context APIs warning in strict mode
* refactor if statement and the warning message
* add other flags for type check
* add component stack tree and refactor wording
* fix the nits
* Rewrite to a switch
I find it a bit easier to follow than many comparison conditions.
* Remove unnecessary assignments
They are being assigned below anyway. This is likely a copypasta from the FOCUS/BLUR special case (which *does* need those assignments).
* Unify "cancel" and "close" cases
Their logic is identical.
* Don't listen to media events at the top
* Add a unit test for double-invoking form events
* Remove an unused case and document it in a test
The case I added was wrong (just like including this event in the top level list was always wrong).
In fact it never bubbles, even for <img>. And since we don't special case it in the <img> event
attachment logic when we create it, we never supported <img onLoadStart> at all.
We could fix it. But Chrome doesn't support it either: https://bugs.chromium.org/p/chromium/issues/detail?id=458851.
Nobody asked us for it yet. And supporting it would require attaching an extra listener to every <img>.
So maybe we don't need it? Let's document the existing state of things.
* Add a test verifying we don't attach unnecessary listeners
* Add a comment
* Add a test for submit (bubbles: false)
**what is the change?:**
Test coverage checking that callbacks are called when they time out.
This test surfaced a bug and this commit includes the fix.
I want to refine this approach, but basically we can simulate time outs
by controlling the return value of 'now()' and the argument passed to
the rAF callback.
Next we will write a browser fixture to further test this against real
browser APIs.
**why make this change?:**
Better tests will keep this module working smoothly while we continue
refactoring and improving it.
**test plan:**
Run the new tests, see that it fails without the bug fix.
We were adding a listener at the root when we weren't meant to. Blames to e96dc14059.
This now alerts once (at FORM) instead of twice (at FORM, #document):
```
var Hello = class extends React.Component {
render() {
return (
<form onSubmit={(e) => {e.preventDefault(); alert('hi ' + e.nativeEvent.currentTarget.nodeName);}}>
<button>hi</button>
</form>
);
}
};
```
* Added start time parameter to Profiler onRender callback
* Profiler also captures commit time
* Only init Profiler stateNode if enableProfilerTimer feature flag enabled
* Extract base Jest config
This makes it easier to change the source config without affecting the build test config.
* Statically import the host config
This changes react-reconciler to import HostConfig instead of getting it through a function argument.
Rather than start with packages like ReactDOM that want to inline it, I started with React Noop and ensured that *custom* renderers using react-reconciler package still work. To do this, I'm making HostConfig module in the reconciler look at a global variable by default (which, in case of the react-reconciler npm package, ends up being the host config argument in the top-level scope).
This is still very broken.
* Add scaffolding for importing an inlined renderer
* Fix the build
* ES exports for renderer methods
* ES modules for host configs
* Remove closures from the reconciler
* Check each renderer's config with Flow
* Fix uncovered Flow issue
We know nextHydratableInstance doesn't get mutated inside this function, but Flow doesn't so it thinks it may be null.
Help Flow.
* Prettier
* Get rid of enable*Reconciler flags
They are not as useful anymore because for almost all cases (except third party renderers) we *know* whether it supports mutation or persistence.
This refactoring means react-reconciler and react-reconciler/persistent third-party packages now ship the same thing.
Not ideal, but this seems worth how simpler the code becomes. We can later look into addressing it by having a single toggle instead.
* Prettier again
* Fix Flow config creation issue
* Fix imprecise Flow typing
* Revert accidental changes
**what is the change?:**
Fix a typo which caused timed out callbacks to not be called.
**why make this change?:**
This is a bug caught by tests I'm in the process of writing, and we
should fix it asap.
**test plan:**
Tests in a WIP PR - will push and share the WIP test in comments on this
PR.
* Generate Flow config on install
We'll need to do pre-renderer Flow passes with different configs.
This is the first step to get it working. We only want the original version checked in.
* Create multiple Flow configs from a template
* Run Flow per renderer
* Lint
* Revert the environment consolidation
I thought this would be a bit cleaner at first because we now have non-environment files in this directory.
But Sebastian is changing these files at the same time so I want to avoid conflicts and keep the PR more tightly scoped. Undo.
* Misc
* Lint for untyped imports and enable Flow typing in ReactFiber
* Re-enable Flow for ReactFiber and fix Flow issues
* Avoid an invariant in DEV-only code
I just introduced it, but on a second thought, it's better to keep it as a warning.
* Address review
* Temporary fix for grabbing wrong rAF polyfill in ReactScheduler
**what is the change?:**
For now...
We need to grab a slightly different implementation of rAF internally at
FB than in Open Source. Making rAF a dependency of the ReactScheduler
module allows us to fork the dependency at FB.
NOTE: After this lands we have an alternative plan to make this module
separate from React and require it before our Facebook timer polyfills
are applied. But want to land this now to keep master in a working state
and fix bugs folks are seeing at Facebook.
Thanks @sebmarkbage @acdlite and @sophiebits for discussing the options
and trade-offs for solving this issue.
**why make this change?:**
This fixes a problem we're running into when experimenting with
ReactScheduler internally at Facebook, **and* it's part of our long term
plan to use dependency injection with the scheduler to make it easier to
test and adjust.
**test plan:**
Ran tests, lint, flow, and will manually test when syncing into
Facebook's codebase.
**issue:**
See internal task T29442940
* ran prettier
Simen Bekkhus
Daniel Lo Nigro
Flarnie Marchan
Sebastian Markbåge
Brian Vaughn
Andrew Clark
Chang Yan
Toru Kobayashi
Dan Abramov
Sophie Alpert
Kevin Lamping
Kevin (Kun) "Kassimo" Qian
Royi Hagigi
Pete Nykänen
Gary Wang