* Use setProperty when setting style properties
setProperty is faster in all/most modern browsers. It also lets us support CSS variables.
* Only use setProperty when setting CSS variables
* Add test to ensure setting CSS variables do not warn
* Make this PR pretty again
* Run fiber test script
Use a ReactFeatureFlag instead. It won't be per-renderer, but we likely
won't need that.
When enableAsyncSubtreeAPI is false, unstable_asyncUpdates is ignored,
but does not warn or throw. That way if we discover a bug in async mode,
we can flip the flag and revert back to sync without code changes.
Shallow renderer and test utils bundles
Adds new bundles introduced with React 15.5 release to master (and 16 alpha)
react-dom/test-utils:
This new bundle contains what used to be react-addons-test-utils. This bundle shares things from react-dom rather than duplicates them.
A temporary createRenderer method has been left behind as a way to access the new shallow renderer. This is for the ReactNative release cycle only and should be going away before the final release.
react-test-renderer/shallow:
This new shallow renderer is almost entirely stand-alone (in that it doesn't use the React reconciler or scheduler). The only touch points are ReactElement and prop/context validation. This renderer is stack and fiber compatible.
* Fix tests to pass when we warn for missing keys in fragments
In most cases we just needed to add the 'key' prop.
This ignores the tests which are already failing on master when running
with ` REACT_DOM_JEST_USE_FIBER=1` - there are 8.
All tests should now pass with `npm run test`, and the 8 which fail when
running `REACT_DOM_JEST_USE_FIBER=1 npm run test` are the same 8 which
are failing on master.
* Added missing key warning for children in array fragments
After trying twice to reuse the code between the ReactChildFiber and
ReactElementValidator, I am thinking that it's simpler to just have some
duplication of code. The parts that are shared are interleaved with
parts which cannot be shared, either because of singleton modules that
must be required differently in 'isomorphic' and the 'renderers', or the
fact that 'warning' requires a hard coded string.
Test Plan:
- Added test to ReactChildren-test
- Manually tested via fixture that was not committed.
* commit updated "scripts/rollup/results.json"
* Make 'ReactChildren-test' more specific, and remove unneeded nesting
Based on helpful tips from @spicyj and @aweary's review
- Made the unit test for the warning on missing keys more specific
- Removed unneeded nesting in the code which generates missing key
warning
- Change test syntax to use JSX to be more consistent
Also fixes flow warning.
* Commit update of scripts/rollup/results.json
* run "scripts/fiber/record-tests"
* Add warning if rendering and HTMLUnknownElement
* Records fiber tests
* Fixes linting and server render tests
* Incorporates review comments
* Uses ownerDocument instead of document
* uses el instead of creating a new element
* Removes warning check of voidElementTags
* Add missing space to unknown element warning
* Only call isCustomComponent once
* Spy on console in menuitem test
Since menuitem is treated as an unknown element in jsdom it triggers the unknown element warning.
* Add unknown element warning to Fiber
* Replace instanceof with toString check
It is more resilient.
* Record tests
(Temporarily) re-adds getters with deprecation warnings for React.PropTypes, React.checkPropTypes, and React.createClass.
* 08bd020: Replace all references to React.PropTypes with prop-types to avoid triggering our own warning message.
* ef5b5c6: Removed several references to React.createClass that appeared after rebasing this branch. (reviewed by @flarnie)
* 524ce20: Added getters for createClass and PropTypes to the main React isomorphic object, behind one-time warning messages. (reviewed by @spicyj)
* db48f54: Fixed Rollup bundles to inline 'prop-types' and 'create-react-class' for UMD builds only. (reviewed by @spicyj, @trueadm )
* cf49cfd: Updated tests-passing.txt to remove tests that were deleted in this branch.
* d34109a: Responses to PR feedback from @spicyj. (Added package.json dependencies to packages/react and packages/react-dom. Renamed a var. Expanded on an inline comment.)
* 488c8d2: Added warning for moved package to React.checkPropTypes accessor too and updated build script.
* 83bcb29: Wordsmithing for deprecation notices (added fb.me links).
* afdc9d2: Tweaked legacy module inlining to remove order-of-deps constraint
* d1348b9: Removed $FlowFixMe.
* 7dbc3e7: More wordsmithing of deprecation notices based on Dan's feedback.
* Remove non-existent /lib/ from souce files
* Replace all shims with explicit access
This deletes shims and changes to access require('react').__DO_NOT_USE__ from renderers for global shared state.
I cloned flattenChildren() and traverseAllChildren() because they relied on CurrentOwner but were used both from Stack and Isomorphic. The stack implementations will die, and the isomorphic ones can be changed to be optimized for Children specifically in the future.
I also deleted UMD shims because they are now unnecessary. I moved the internals assignment to main modules since they're now used in tests, and made them direct UMD entry points.
* WIP
* fbjs support
* WIP
* dev/prod mode WIP
* More WIP
* builds a cjs bundle
* adding forwarding modules
* more progress on forwarding modules and FB config
* improved how certain modules get inlined for fb and cjs
* more forwarding modules
* added comments to the module aliasing code
* made ReactPerf and ReactTestUtils bundle again
* Use -core suffix for all bundles
This makes it easier to override things in www.
* Add a lazy shim for ReactPerf
This prevents a circular dependency between ReactGKJSModule and ReactDOM
* Fix forwarding module for ReactCurrentOwner
* Revert "Add a lazy shim for ReactPerf"
This reverts commit 723b402c07.
* Rename -core suffix to -fb for clarity
* Change forwarding modules to import from -fb
This is another, more direct fix for ReactPerf circular dependency
* should fix fb and cjs bundles for ReactCurrentOwner
* added provides module for ReactCurrentOwner
* should improve console output
* fixed typo with argument passing on functon call
* Revert "should improve console output"
This breaks the FB bundles.
This reverts commit 65f11ee64f.
* Work around internal FB transform require() issue
* moved ReactInstanceMap out of React and into ReactDOM and ReactDOMFiber
* Expose more internal modules to www
* Add missing modules to Stack ReactDOM to fix UFI
* Fix onlyChild module
* improved the build tool
* Add a rollup npm script
* Rename ReactDOM-fb to ReactDOMStack-fb
* Fix circular dependencies now that ReactDOM-fb is a GK switch
* Revert "Work around internal FB transform require() issue"
This reverts commit 0a50b6a90b.
* Bump rollup-plugin-commonjs to include a fix for rollup/rollup-plugin-commonjs#176
* Add more forwarding modules that are used on www
* Add even more forwarding modules that are used on www
* Add DOMProperty to hidden exports
* Externalize feature flags
This lets www specify them dynamically.
* Remove forwarding modules with implementations
Instead I'm adding them to react-fb in my diff.
* Add all injection necessary for error logging
* Add missing forwarding module (oops)
* Add ReactART builds
* Add ReactDOMServer bundle
* Fix UMD build of ReactDOMFiber
* Work in progress: start adding ReactNative bundle
* tidied up the options for bundles, so they can define what types they output and exclude
* Add a working RN build
* further improved and tidied up build process
* improved how bundles are built by exposing externals and making the process less "magical", also tidied up code and added more comments
* better handling of bundling ReactCurrentOwner and accessing it from renderer modules
* added NODE_DEV and NODE_PROD
* added NPM package creation and copying into build chain
* Improved UMD bundles, added better fixture testing and doc plus prod builds
* updated internal modules (WIP)
* removed all react/lib/* dependencies from appearing in bundles created on build
* added react-test-renderer bundles
* renamed bundles and paths
* fixed fixture path changes
* added extract-errors support
* added extractErrors warning
* moved shims to shims directory in rollup scripts
* changed pathing to use build rather than build/rollup
* updated release doc to reflect some rollup changes
* Updated ReactNative findNodeHandle() to handle number case (#9238)
* Add dynamic injection to ReactErrorUtils (#9246)
* Fix ReactErrorUtils injection (#9247)
* Fix Haste name
* Move files around
* More descriptive filenames
* Add missing ReactErrorUtils shim
* Tweak reactComponentExpect to make it standalone-ish in www
* Unflowify shims
* facebook-www shims now get copied over correctly to build
* removed unnecessary resolve
* building facebook-www/build is now all sync to prevent IO issues plus handles extra facebook-www src assets
* removed react-native-renderer package and made build make a react-native build dir instead
* 😭😭😭
* Add more SSR unit tests for elements and children. (#9221)
* Adding more SSR unit tests for elements and children.
* Some of my SSR tests were testing for react-text and react-empty elements that no longer exist in Fiber. Fixed the tests so that they expect correct markup in Fiber.
* Tweaked some test names after @gaearon review comment https://github.com/facebook/react/pull/9221#discussion_r107045673 . Also realized that one of the tests was essentially a direct copy of another, so deleted it.
* Responding to code review https://github.com/facebook/react/pull/9221#pullrequestreview-28996315 . Thanks @spicyj!
* ReactElementValidator uses temporary ReactNative View propTypes getter (#9256)
* Updating packages for 16.0.0-alpha.6 release
* Revert "😭😭😭"
This reverts commit 7dba33b2cf.
* Work around Jest issue with CurrentOwner shared state in www
* updated error codes
* splits FB into FB_DEV and FB_PROD
* Remove deps on specific builds from shims
* should no longer mangle FB_PROD output
* Added init() dev block to ReactTestUtils
* added shims for DEV only code so it does not get included in prod bundles
* added a __DEV__ wrapping code to FB_DEV
* added __DEV__ flag behind a footer/header
* Use right haste names
* keeps comments in prod
* added external babel helpers plugin
* fixed fixtures and updated cjs/umd paths
* Fixes Jest so it run tests correctly
* fixed an issue with stubbed modules not properly being replaced due to greedy replacement
* added a WIP solution for ReactCurrentOwner on FB DEV
* adds a FB_TEST bundle
* allows both ReactCurrentOwner and react/lib/ReactCurrentOwner
* adds -test to provides module name
* Remove TEST env
* Ensure requires stay at the top
* added basic mangle support (disbaled by default)
* per bundle property mangling added
* moved around plugin order to try and fix deadcode requires as per https://github.com/rollup/rollup/issues/855
* Fix flow issues
* removed gulp and grunt and moved tasks to standalone node script
* configured circleci to use new paths
* Fix lint
* removed gulp-extract-errors
* added test_build.sh back in
* added missing newline to flow.js
* fixed test coverage command
* changed permissions on test_build.sh
* fixed test_html_generations.sh
* temp removed html render test
* removed the warning output from test_build, the build should do this instead
* fixed test_build
* fixed broken npm script
* Remove unused ViewportMetrics shim
* better error output
* updated circleci to node 7 for async/await
* Fixes
* removed coverage test from circleci run
* circleci run tets
* removed build from circlci
* made a dedicated jest script in a new process
* moved order around of circlci tasks
* changing path to jest in more circleci tests
* re-enabled code coverage
* Add file header to prod bundles
* Remove react-dom/server.js (WIP: decide on the plan)
* Only UMD bundles need version header
* Merge with master
* disabled const evaluation by uglify for <script></script> string literal
* deal with ART modules for UMD bundles
* improved how bundle output gets printed
* fixed filesize difference reporting
* added filesize dep
* Update yarn lockfile for some reason
* now compares against the last run branch built on
* added react-dom-server
* removed un-needed comment
* results only get saved on full builds
* moved the rollup sized plugin into a plugins directory
* added a missing commonjs()
* fixed missing ignore
* Hack around to fix RN bundle
* Partially fix RN bundles
* added react-art bundle and a fixture for it
* Point UMD bundle to Fiber and add EventPluginHub to exported internals
* Make it build on Node 4
* fixed eslint error with resolve being defined in outer scope
* Tweak how build results are calculated and stored
* Tweak fixtures build to work on Node 4
* Include LICENSE/PATENTS and fix up package.json files
* Add Node bundle for react-test-renderer
* Revert "Hack around to fix RN bundle"
We'll do this later.
This reverts commit 59445a6259.
* Revert more RN changes
We'll do them separately later
* Revert more unintentional changes
* Revert changes to error codes
* Add accidentally deleted RN externals
* added RN_DEV/RN_PROD bundles
* fixed typo where RN_DEV and RN_PROD were the wrong way around
* Delete/ignore fixture build outputs
* Format scripts/ with Prettier
* tidied up the Rollup build process and split functions into various different files to improve readability
* Copy folder before files
* updated yarn.lock
* updated results and yarn dependencies to the latest versions
If your renderer doesn't use host context, you might prefer to return null. This used to give an error:
> Invariant Violation: Expected host context to exist. This error is likely caused by a bug in React. Please file an issue.
I use a sentinel value instead now.
The code in ReactFiberHostContext is a little complicated now. We could probably also just remove the invariants.
* Only re-assign defaultValue if it is different
* Do not set value if it is the same
* Properly cover defaultValue
* Use coercion to be smart about value assignment
* Add explanation of loose type checks in value assignment.
* Add test coverage for setAttribute update.
* Only apply loose value check to text inputs
* Fix case where empty switches to zero
* Handle zero case in controlled input
* Correct mistake with default value assignment after rebase
* Do not assign bad input to number input
* Only trigger number input value attribute updates on blur
* Remove reference to LinkedValueUtils
* Record new fiber tests
* Add tests for blurred number input behavior
* Replace onBlur wrapper with rule in ChangeEventPlugin
* Sift down to only number inputs
* Re-record fiber tests
* Add test case for updating attribute on uncontrolled inputs. Make related correction
* Handle uncontrolled inputs, integrate fiber
* Reorder boolean to mitigate DOM checks
* Only assign value if it is different
* Add number input browser test fixtures
During the course of the number input fix, we uncovered many edge
cases. This commit adds browser test fixtures for each of those instances.
* Address edge case preventing number precision lower than 1 place
0.0 coerces to 0, however they are not the same value when doing
string comparision. This prevented controlled number inputs from
inputing the characters `0.00`.
Also adds test cases.
* Accommodate lack of IE9 number input support
IE9 does not support number inputs. Number inputs in IE9 fallback to
traditional text inputs. This means that accessing `input.value` will
report the raw text, rather than parsing a numeric value.
This commit makes the ReactDOMInput wrapper check to see if the `type`
prop has been configured to `"number"`. In those cases, it will
perform a comparison based upon `parseFloat` instead of the raw input
value.
* Remove footnotes about IE exponent issues
With the recent IE9 fix, IE properly inserts `e` when it produces an
invalid number.
* Address exception in IE9/10 ChangeEventPlugin blur event
On blur, inputs have their values assigned. This is so that number
inputs do not conduct unexpected behavior in
Chrome/Safari. Unfortunately, there are cases where the target
instance might be undefined in IE9/10, raising an exception.
* Migrate over ReactDOMInput.js number input fixes to Fiber
Also re-record tests
* Update number fixtures to use latest components
* Add number input test case for dashes and negative numbers
* Replace trailing dash test case with replace with dash
Also run prettier
* Adding more SSR unit tests for elements and children.
* Some of my SSR tests were testing for react-text and react-empty elements that no longer exist in Fiber. Fixed the tests so that they expect correct markup in Fiber.
* Tweaked some test names after @gaearon review comment https://github.com/facebook/react/pull/9221#discussion_r107045673 . Also realized that one of the tests was essentially a direct copy of another, so deleted it.
* Responding to code review https://github.com/facebook/react/pull/9221#pullrequestreview-28996315 . Thanks @spicyj!
* Removed optimization for events without target in ReactNativeEventEmitter
This PR fixes the problem originally introduced in https://github.com/facebook/react/pull/6590
The problem is that `ResponderEventPlugin` (and `ResponderTouchHistoryStore`) relies on that fact that touch events are balanced.
So if one `startish` event happened, should be coming `endish` event. Otherwise there is no way to maintain internal `trackedTouchCount` counter. So, if we drop some events, we break this logic.
Moreover, that optimization clearly contradict with this statement from `ResponderEventPlugin`:
```
We must be resilient to `targetInst` being `null` on `touchMove` or
`touchEnd`. On certain platforms, this means that a native scroll has
assumed control and the original touch targets are destroyed.
```
This issue causes several major problems in React Native, and one of them (finally!) is easy to reproduce: https://github.com/facebook/react-native/issues/12976 .
The test also illustrates this problem.
* Prettier
* Added a handful of SSR unit tests, ported from a previous pull request.
* Fixing linting errors
* Fixed a test helper function to properly report errors.
* Un-nested the new rendering tests. Updated the fiber test passing/not passing lists.
* Edited to comply with the react/jsx-space-before-closing eslint rule, which will soon be merged into master.
* Response to code review from @spicyj. Moved tests to separate file, reworked wording of test names, corrected use of canUseDom, and simplified tests against tagName. Thanks for the help, @spicyj!
* Converted the unit tests to use async-await style.
* Moved async-await babel transform for tests from .babelrc to preprocessor.js.
* Response to code review in PR #9089. Thanks, @spicyj!
* Fixing some bugs in the SSR unit tests.
* Missed deleting some repeated code in the last commit.
* Adding unit tests for property to attribute mapping in SSR.
* Removing some redundant unit tests.
* Oops. I forgot to re-run record-tests after c593dbc; fixing that here.
* Reformatting for prettier so that the build will pass.
`scripts/prettier/index.js write` will run prettier on source files.
Run using `yarn prettier`.
`scripts/prettier/index.js` will throw if any source files are not
formatted with prettier. We'll use this to block CI.
Based on similar script in Jest repo.
* Fix version bump script when files are missing
* Add noop renderer to version bump script
Without this, grunt build fails due to the version mismatch.
* wip
* better
* better
* track commits
* better
* wip
* Fix
* Add some lifecycles
* wip
* Naming
* Moar emojis
* Remove stacks in favor of a flag
* Fix Flow
* Gate behind __DEV__
* Revert flag for testing
* Measure all lifecycles
* Push it to the limits
* Polish
* Indent
* Refactor and track cascading updates
* More prominent warnings
* Make mark names themselves readable
This is useful for RN Systrace which doesn't let us assign labels after the fact.
* Keep track of how many effects we call
* Fix typo
* Do less work to reduce the overhead
* Fix lint
* Remove closure
* Remove unintentional formatting changes
* Add tests
* Fix test regex and record tests
* Disable irrelevant tests needed for ReactPerf
* Fix typo
* Fix lint and flow
* Don't treat cWM or cWRP as cascading
* Whitespace
* Update tests
* Gate callComponentWillUnmountWithTimerInDev definition by DEV
* Added a handful of SSR unit tests, ported from a previous pull request.
* Fixing linting errors
* Fixed a test helper function to properly report errors.
* Un-nested the new rendering tests. Updated the fiber test passing/not passing lists.
* Edited to comply with the react/jsx-space-before-closing eslint rule, which will soon be merged into master.
* Response to code review from @spicyj. Moved tests to separate file, reworked wording of test names, corrected use of canUseDom, and simplified tests against tagName. Thanks for the help, @spicyj!
Added a guard in the global error event listener to prevent nested
errors from being captured higher up the stack.
Also found a bug where the DEV version of invokeGuardedCallback would
infinite loop if there were nested invocations with the same name. Fixed
by appending the depth to the fake event name. (I'm actually not sure
why this is necessary.)
We can use this more flexible version for error handling in Fiber.
The DEV mode version uses same fake event trick as before to preserve
"break on uncaught exception" behavior when debugging.
Stack uses a module called ReactInvalidSetStateWarningHook, but all it
does is warn about setState inside getChildContext. Doesn't seem worth
keeping around, so I didn't use it for Fiber, but if I'm mistaken I'll
change it.
> This PR adds a test (initially failing) for this case along with a fix for stack and fiber. The stack fix was copied from a Diff submitted by @sema. The fiber fix just required us to stop leaking properties for unmounted views.
> Longer term we may want to explicitly invoke a release event listener for a responder just before unmounting it. This PR does not do that.
ReactStatelessComponent-test.js fails due to warnings not showing up when functional components mount or update. This PR ports the existing warnings from stack and re-applies them to fiber functional components and attempts to re-use some logic in ReactFiberContext that currently exists for showing warnings when dealing with class components.
* Add toTree() method to stack and fiber TestRenderer
* Address PR feedback
* Refactor TestRenderer to use correct root
* Rebase off master and fix root node references
* Add flow types
* Add test for null rendering components
* Remove last remaining lint error
* Add missing test
* Fixed ReactNativeFiber event system regression introduced in b354db2
Also added test coverage to ReactNativeEvents-test for Fiber
* ReactNative tests now run against fiber renderer when env flag set
Updated the test-framework-setup file so that record-tests runs native tests against ReactNativeFiber. ReactComponentTreeHook native tests all now fail but that's expected.
* Avoid calling setChildren() if no children
We have a bunch of tests that are meant to test generic React behavior.
However, we've written the tests against the DOM renderer. This moves
tests that depend on ReactDOM or ReactTestUtils out of the shared repo.
This will let us sync the shared repo directly to environments that don't
support running these tests. Such as React Native.
* Revert "Revert "Warn if PropType function is called manually (#7132)""
This reverts commit be71f76ed8.
In other words, now we again have the warning if you attempt to call PropTypes manually.
It was removed in #8066 but we shouldn't have done this since we still want to avoid people accidentally calling them in production (and even more so since now it would throw).
Fixes#8080.
* Pass secret in ReactControlledValuePropTypes
* Record tests
The entries of a map are a two-element array of [key, value], which
React will treat as fragments with children. This is unlikely to ever
be the intended behavior, so we warn.
There are a few different issues:
* Updates result in unnecessary duplicate placements because it can't find the current fiber for continuations.
* When run together, coroutine update and unmounting tests appear to lock down in an infinite loop. They don't freeze in isolation.
I don't have a solution for this but just leaving it for future fixes.
It's not just events that read the current props. Controlled components
do as well. Since we're no longer updating the Fiber pointer during updates
we have to instead read from the node props to get the current props.
Since this method is no longer just used for events I renamed it.
* Allow renderers to return an update payload in prepareUpdate
This then gets stored on updateQueue so that the renderer doesn't need to
think about how to store this.
It then gets passed into commitUpdate during the commit phase.
This allows renderers to do the diffing during the time sliced path,
allocate a queue for changes and do only the absolute minimal work to
apply those changes in the commit phase.
If refs update we still schedule and update.
* Hack around the listener problem
* Diff ReactDOMFiber properties in prepareUpdate
We now take advantage of the new capability to diff properties early.
We do this by generating an update payload in the form of an array with
each property name and value that we're about to update.
* Add todo for handling wasCustomComponentTag
* Always force an update to wrapper components
Wrapper components have custom logic that gets applied at the commit phase
so we always need to ensure that we schedule an update for them.
* Remove rootContainerInstance from commitMount
No use case yet and I removed it from commitUpdate earlier.
* Use update signal object in test renderer
* Incorporate 8652 into new algorithm
* Fix comment
* Add failing test for flipping event handlers
This illustrates the problem that happens if we store a pointer to the
Fiber and then choose not to update that pointer when no properties change.
That causes an old Fiber to be retained on the DOM node. Then that Fiber
can be reused by the pooling mechanism which then will mutate that Fiber
with new event handlers, which makes them active before commit.
* Store current props in the RN instance cache and on the DOM node
This represents the current set of event listeners. By not relying on the
Fiber, it allows us to avoid doing any effects in the commit phase when
nothing changes.
This is a bit ugly. Not super happy how this all came together.
* Add a test for rendering SVG into a non-React SVG tree
It is failing in Fiber.
* Add a test for rendering HTML into non-React foreignObject
It happens to pass in Fiber.
* Determine root namespace by container namespace, not just tag
A tag alone is not enough to determine the tree namespace.
* Skip the test that exhibits jsdom bug in non-createElement mode
jsdom doesn't give HTML namespace to foreignObject.innerHTML children.
This problem doesn't exist in the browsers.
https://github.com/facebook/react/pull/8638#issuecomment-269349642
We will skip this test in non-createElement mode considering
that non-createElement mounting is effectively dead code now anyway.
A new module has been added (ReactFiberErrorLogger). This logs error information (call stack and component stack) to the console to make errors easier to debug. It also prompts users to use error boundaries if they are not already using them.
In the future, perhaps this will be injectable, enabling users to provide their own handler for custom processing/logging. For the time being, this should help with issues like this / #2461.
Previous (probably unintentional) behavior of Stack was to allow components to define childContextTypes without also supplying a getChildContext property. This PR updates Fiber to (temporarily) mimic that behavior. It also adds warning messages to both Fiber and Stack (along with a test).
For the time being, Fiber components with a missing getChildContext method will return the parent context as-is, after warning.
Fixes the case where there's an uncaught error and the root unmounts.
We implement this by rendering the root as if its child is null. Null is
not usually allowed at the top level, so we need to special case it.
Currently we update the memoized inputs (props, state) during the
complete phase, as we go back up the tree. That means we can't reuse
work until of its children have completed.
By moving memoization to the begin phase, we can do a shallow bailout,
reusing a unit of work even if there's still work to do in its children.
Memoization now happens whenever a fiber's `child` property is updated;
typically, right after reconciling. It's also updated when
`shouldComponentUpdate` returns false, because that indicates that the
given state and props are equal to the memoized state and props.
* fix failed tests on Windows #8737
* Use regexp literal instead of `RegExp` constructor so that we don't need to bother with escaping special character like `\` and `.`.
Note that by removing the path separator in regexp, I've relaxed the matching condition a little.Since there's little chance we have files like `XXXReact.d.ts`,it should not matter.
* ReactTestRenderer move current impl to stack dir
* ReactTestRenderer on fiber: commence!
* ReactTestRenderer: most non-ref/non-public-instance tests are passing
* Move ReactTestFiberComponent functions from Renderer to Component file
* test renderer: get rid of private root containers and root Maps
* TestRenderer: switch impl based on ReactDOMFeatureFlag.useFiber
* ReactTestRenderer: inline component creation
* ReactTestRenderer: return to pristine original glory (+ Fiber for error order difference)
* TestRendererFiber: use a simple class as TestComponentInstances
* Add `getPublicInstance` to support TestRenderer `createNodeMock`
* Rename files to end. Update for `mountContainer->createContainer` change
* test renderer return same object to prevent unnecessary context pushing/popping
* Fiber HostConfig add getPublicInstance. This should be the identity fn everywhere except the test renderer
* appease flow
* Initial cleanup from sleepy work
* unstable_batchedUpdates
* Stack test renderer: cache nodeMock to not call on unmount
* add public instance type parameter to the reconciler
* test renderer: set _nodeMock when mounted
* More cleanup
* Add test cases for root fragments and (maybe?) root text nodes
* Fix the npm package build
Explicitly require the Stack version by default.
Add a separate entry point for Fiber.
We don't add fiber.js to the package yet since it's considered internal until React 16.
* Relax the ref type from Object to mixed
This seems like the most straightforward way to support getPublicInstance for test renderer.
* Remove accidental newline
* test renderer: unify TestComponent and TestContainer, handle root updates
* Remove string/number serialization attempts since Fiber ensures all textInstances are strings
* Return full fragments in toJSON
* Test Renderer remove TestComponent instances for simple objects
* Update babylon for exact object type syntax
* Use $$typeof because clarity > punching ducks.
* Minor Flow annotation tweaks
* Tweak style, types, and naming
* Fix typo
* Verify that functional component ref warning is deduplicated
It's not a big problem for string refs because the ref stays the same and the warning code path runs once per mount.
However, it is super annoying for functional refs since they're often written as arrows, and thus the warning fires for every render.
Both tests are currently failing since we're mounting twice, so even the string ref case prints warnings twice.
* Extract the warning condition to the top level
We don't want to run getStackAddendumByID() unless we actually know we're going to print the warning.
This doesn't affect correctness. Just a performance fix.
* Deduplicate warnings about refs on functional components
This fixes the duplicate warnings and adds an additional test for corner cases.
Our goal is to print one warning per one offending call site, when possible.
We try to use the element source for deduplication first because it gives us the exact JSX call site location.
If the element source is not available, we try to use the owner name for deduplication.
If even owner name is unavailable, we try to use the functional component unique identifier for deduplication so that at least the warning is seen once per mounted component.
* Fiber: warn for refs on SFCs
* Stateless refs: update warning to use component stack
* Warn for function refs (stack and fiber)
* Add owner reference to ref warnings
* Centralize stack ref warnings in ReactRef.attachRef
* Fiber stateless comp ref warning should only do work when necessary
* ReactDebugCurrentFiber maybe FunctionalComponents should act this way instead
* (chore) scripts/fiber/record-tests
* Add component._compositeType to ReactInstance Flow definition
* Don't handle 'stack inside fiber' case in the warning
We don't have a test for it. It's easy to mess it up and print the wrong thing so instead of verifying it I'll just remove this bit.
* Revert the change to getCurrentFiberOwnerName
This happened to work, but it is just a coincidence. This change didn’t really match what the function was supposed to be doing.
I’m not sure what the correct fix would be yet so this commit breaks tests.
* Add component indirection to the tests using owner name
This passes in Stack. It helps ensure we supply the correct owner name.
* Invalid type invariant should read owner from element
This brings the Fiber behavior in line with how Stack does it. The Fiber test was passing accidentally but broke down in more complicated cases (such as when we have an <Indirection> between the owner and the failing element).
Now we can also remove the weird cases from getCurrentFiberOwnerName() that didn't really make sense but helped get the (incomplete) tests pass in the past.
* Fiber should throw on a string ref inside functional component
This will publish all packages in build/packages/*.tgz to the "next" tag. If the current version is "stable" (doesn't trigger semver.prerelease()) then it will also update the "latest" dist-tag to point at that version.
* Add manual build fixtures
* Inject ReactDOM into ReactWithAddons from ReactWithAddons
We used to read ReactDOM as a global inside ReactAddonsDOMDependenciesUMDShim.
This didn't work in AMD environments such as RequireJS and SystemJS.
Instead, I changed it so that ReactDOM gets injected into ReactWithAddons by ReactDOM itself.
This way we don't have to try to require it (which wouldn't work because AMD doesn't handle circular dependencies well).
This means you have to load ReactDOM first before using ReactDOM-dependent addons, but this was already the case before.
This commit makes all build fixtures pass.
* Memoize ReactDOM to avoid going into require on every access
* Add Brunch fixture
* Inline requires to work around Brunch bug
See #8556 and https://github.com/brunch/brunch/issues/1591#issuecomment-270742503 for context.
This appears to be a Brunch bug but we can keep a temporary fix until the next major.
Moved ReactFiberClassComponent validateCallback() helper function into a standalone util used by both fiber and stack implementations. Validation now happens in ReactFiberUpdateQueue so that non-DOM renderers will also benefit from it.
Reverses the effect of batchedUpdates by resetting the current
batching context.
Does not affect nested updates, which are always deferred regardless
of whether they are inside a batch.
The DOM renderer assumes that resetAfterCommit is called after
prepareForCommit without any nested commits in between. That may not
be the case now that syncUpdates forces a nested update.
To address, this changes the type of prepareForCommit to return a value
which is later passed to resetAfterCommit.
The finalizeInitialChildren HostConfig method now utilizes a boolean return type. Renderers can return true to indicate that custom effects should be processed at commit-time once host components have been mounted. This type of work is marked using the existing Update flag.
A new HostConfig method, commitMount, has been added as well for performing this type of work.
This change set is in support of the autoFocus prop.
* Push class context providers early
Previously we used to push them only after the instance was available. This caused issues in cases an error is thrown during componentWillMount().
In that case we never got to pushing the provider in the begin phase, but in complete phase the provider check returned true since the instance existed by that point. As a result we got mismatching context pops.
We solve the issue by making the context check independent of whether the instance actually exists. Instead we're checking the type itself.
This lets us push class context early. However there's another problem: we might not know the context value. If the instance is not yet created, we can't call getChildContext on it.
To fix this, we are introducing a way to replace current value on the stack, and a way to read the previous value. This also helps remove some branching and split the memoized from invalidated code paths.
* Add a now-passing test from #8604
Also rename another test to have a shorter name.
* Move isContextProvider() checks into push() and pop()
All uses of push() and pop() are guarded by it anyway.
This makes it more similar to how we use host context.
There is only one other place where isContextProvider() is used and that's legacy code needed for renderSubtree().
* Clarify why we read the previous context
* Use invariant instead of throwing an error
* Fix off-by-one in ReactFiberStack
* Manually keep track of the last parent context
The previous algorithm was flawed and worked by accident, as shown by the failing tests after an off-by-one was fixed.
The implementation of getPrevious() was incorrect because the context stack currently has no notion of a previous value per cursor.
Instead, we are caching the previous value directly in the ReactFiberContext in a local variable.
Additionally, we are using push() and pop() instead of adding a new replace() method.
* Add a failing test for recursion check
* Make requestIdleCallback() and requestAnimationFrame() shims async
They no longer need to be sync in tests because we already made DOM renderer sync by default.
This fixes a crash when testing incrementalness in DOM renderer itself.
* Style fix
* Fix lint
This was a bug because of the split between ClassComponent and IndeterminateComponent -- we didn't mark effects or push context when we come from an indeterminate component. Now they behave the same way.
The code is even clumsier than before but I'm pretty worried we'll screw these up in the future if we don't unify the paths.
Not many components exercise this path but Relay containers do so it's noticeable.
Otherwise you fall into an infinite loop where root fails -> host env
throws -> root fails...
This is a worst-case scenario that should only be possible if there's
a bug in the renderer.
We warn about getInitialState() usage in class components because users may have accidentally used it instead of state. If they have also specified state though then we should skip this warning.
Context: https://twitter.com/soprano/status/810003963286163456
* (WIP) Nesting warnings
* Remove indirection
* Add a note about namespace
* Fix Flow and make host context required
This makes it easier to avoid accidental nulls.
I also added a test for the production case to avoid regressing because of __DEV__ branches.
We should be able to support passing a function to replaceState, which
receives the accumulation of all the previously applied state updates.
Which means we shouldn't drop those updates from the queue. Technically,
we could drop them only when an object is passed to replaceState, but
that seems like more trouble than it's worth.
The update is scheduled as if the current processing update has already
been processed; if it has the same or higher priority, it will be
flushed in the same batch.
We also print a warning.
The queue maintains a pointer to the last progressed update in the list.
Updates that come after that pointer are pending. The pointer is set to
the end of the list during reconciliation.
Pending updates are sorted by priority then insertion. Progressed
updates are sorted by the order in which they were applied during
reconciliation, which may not be by priority: if a component bails out
before the updates are committed, in the next render, the progressed
updates are applied in the same order that they were previously, even if
a higher priority update comes in.
Once a progressed update is flushed/committed, it's removed from
the queue.
When resetting the priority in the complete phase, check the priority of
the update queue so that updates aren't dropped.
Updates inside render, child cWRP, etc are no longer dropped.
The next step is sort the queue by priority and only flush updates that
match the current priority level.
* Implement component stack for some warnings in Fiber
* Keep Fiber debug source up to date
When an element changes, we should copy the source and owner again.
Otherwise they can get stale since we're not reading them from the element.
* Remove outdated TODOs from tests
* Explicitly specify Fiber types to include in the stack
Fixes an accidental omission when both source and owner are null but displayName exists.
* Fix mised Stack+Fiber test to not expect extra warnings
When we're in Fiber mode we don't actually expect that warning being printed.
* Warn on passing different props to super()
* Implement duplicate key warnings
We keep known keys in a set in development. There is an annoying special case where we know we'll check it again because we break out of the loop early.
One test in the tree hook regresses to the failing state because it checks that the tree hook works without a Set available, but we started using Set in this code. It is not essential and we can clean this up later when we decide how to deal with polyfills.
* Move ReactTypeOfWork to src/shared
It needs to be available both to Fiber and Isomorphic because the tree hook lives in Isomorphic but pretty-prints Fiber stack.
* Add dev-only ReactDebugCurrentFiber for warnings
The goal is to use ReactCurrentOwner less and rely on ReactDebugCurrentFiber for warning owner name and stack.
* Make Stack invariant messages more consistent
Fiber used a helper so two tests had the same phrasing.
Stack also used a helper for most invariants but hardcoded a different phrase in one place.
I changed that invariant message to use a helper which made it consistent with what it prints in Fiber.
* Make CSSPropertyOperations use getCurrentFiberOwnerName()
This gets mount-time CSS warnings to be printed.
However update-time warnings are currently ignored because current fiber is not yet available during the commit phase.
We also regress on HostOperation hook tests but this doesn't matter because it's only used by ReactPerf and it doesn't work with Fiber yet anyway. We'll have to think more about it later.
* Set ReactDebugCurrentFiber during the commit phase
This makes it available during updates, fixing the last failing test in CSSPropertyOperations.
* Add DOM warnings by calling hooks directly
It is not clear if the old hook system is worth it in its generic incarnation. For now I am just hooking it up to the DOMFiber renderer directly.
* Add client-side counterparts for some warning tests
This helps us track which warnings are really failing in Fiber, and which ones depend on SSR.
This matches what we do in Fiber -- and doing it this way is the only way we can prepare new views in the background before unmounting old ones.
In particular, this breaks this pattern:
```js
class Child1 extends React.Component {
render() { ... }
componentWillMount() {
this.props.registerChild(this);
}
componentWillUnmount() {
this.props.unregisterChild();
}
}
class Child2 extends React.Component {
render() { ... }
componentWillMount() {
this.props.registerChild(this);
}
componentWillUnmount() {
this.props.unregisterChild();
}
}
class Parent extends React.Component {
render() {
return (
showChild1 ?
<Child1
registerChild={(child) => this.registered = child}
unregisterChild={() => this.registered = null}
/> :
<Child2
registerChild={(child) => this.registered = child}
unregisterChild={() => this.registered = null}
/>
);
}
}
```
Previously, `this.registered` would always be set -- now, after a rerender, `this.registered` gets stuck at null because the old child's componentWillUnmount runs *after* the new child's componentWillMount.
A correct fix here is to use componentDidMount rather than componentWillMount. (In general, componentWillMount should not have side effects.) If Parent stored a list or set of registered children instead, there would also be no issue.
This matches the behavior in Fiber. Normally we would change Fiber
to match Stack to minimize breaking changes for the initial release.
However, in this case it would require too large a compromise to change
Fiber to act like Stack.
We introduced runtime validation of tag names because we used to generate
HTML that was supposed to be inserted into a HTML string which could've
been an XSS attack.
However, these days we use document.createElement in most cases. That
already does its internal validation in the browser which throws. We're now
double validating it. Stack still has a path where innerHTML is used and
we still need it there. However in Fiber we can remove it completely.
In #2756 we ended up using toLowerCase to allow case insensitive HTML tags.
However, this requires extra processing every time we access the tag or
at least we need to process it for mount and store it in an extra field
which wastes memory.
So instead, we can just enforce case sensitivity for HTML since this might
matter for the XML namespaces like SVG anyway.
* Enable additional (failing) portal tests
* Fix portal unmounting
When unmount a portal, we need to unmount *its* children from itself.
This is similar to what we would do for a root if we allowed deleting roots.
* Skip portals when looking for host siblings
A portal is not part of that host tree despite being a child.
* Fix comment typo
* Add a failing test for portal child reconciliation
It is failing because portal bails out of update, seeing null in pendingProps.
It is null because we set pendingProps to nextPortal.children, which is null in this test.
* Fix the bug when switching to a null portal child
If pendingProps is null, we do a bailout in beginWork.
This prevents unmounting of the existing child when the new child is null.
We fix this by changing portal fiber's pendingProps to be the portal object itself instead of its children.
This way, it is never null, and thus doesn't cause a false positive in the bailout condition.
* Add a comment about HostPortal in getHostSibling
* Revert the fix because I don't know why it worked
unmountHostComponents() should have worked despite finding the wrong parent because it should have changed the parent when pushing and popping the portals.
* Don't call commitDeletion recursively
This leads to a "Cannot find host parent" bug because commitDeletion() clears the return field.
When we're inside the loop, we assign node.sibling.return to node.return but by this moment node.return has already been nulled.
As a solution we inline code from commitDeletion() without the nulling.
It still fails tests but for a different reason (unrelated bug).
* Skip portal children in commitNestedUnmounts()
We are currently already visiting them in commitUnmount() portal case since it's recursive.
This condition avoids visting them twice.
* Set node.child.return before going deeper
It doesn't seem to influence existing tests but we have this protection in all other similar places.
It protects against infinite loops.
* Revert "Fix the bug when switching to a null portal child"
This reverts commit ed9747deed.
I'll solve this by using an array in place of null instead.
* Use [] for empty Portal pendingProps
This avoids a false positive bailout with pendingProps == null when portal is empty.
* Test that SVG elements get created with the right namespace
* Pass root to the renderer methods
* Keep track of host instances and containers
* Keep instances instead of fibers on the stack
* Create text instances in begin phase
* Create instance before bailing on offscreen children
Otherwise, the parent gets skipped next time.
We could probably create it later but this seems simpler.
* Tweak magic numbers in incremental tests
I don't understand why they changed but probably related to us moving some work into begin phase?
* Only push newly created nodes on the parent stack
Previously I was pushing nodes on the parent stack regardless of whether they were already in current or not.
As a result, insertions during updates were duplicated, and nodes were added to existing parents before commit phase.
Luckily we have a test that caught that.
* Fix lint
* Fix Flow
I had to wrap HostContext API into a closure so that it's parameterizeable with I and C.
* Use the same destructuring style in scheduler as everywhere else
* Remove branches that don't seem to run anymore
I'm not 100% sure this is right but I can't get tests to fail.
* Be explicit about the difference between type and tag
I was confused by th HACK comment so I learned how DOM and SVG work with casing and tried to write a more descriptive comment.
It also seems like passing fiber.type into finalizeInitialChildren() is a potential problem because DOM code assumes tag is lowercase.
So I added a similar "hack" to finalizeInitialChildren() that is identical to the one we have prepareUpdate() so if we fix them later, we fix both.
* Save and restore host context when pushing and popping portals
* Revert parent context and adding children in the begin phase
We can address this later separately as it is a more hot path.
This doesn't affect correctness of SVG container behavior.
* Add a test for SVG updates
This tests the "jump" reuse code path in particular.
* Record tests
* Read ownerDocument from the root container instance
This way createInstance() depends on the innermost container only for reading the namespace.
* Track namespaces instead of creating instances early
While we might want to create instance in the begin phase, we shouldn't let DOM guide reconciler design.
Instead, we are adding a new concept of "host context". In case of ReactDOMFiber, it's just the current namespace.
We are keeping a stack of host context values, ignoring those that are referentially equal.
The renderer receives the parent context and type, and can return a new context.
* Pop child context before reading own context and clarify API
It wasn't quite clear from the API which context was being returned by the renderer. Changed the API to specifically ask for child context, and thus to pop it before getting the current context.
This fixes the case with <foreignObject> to which I intended to give SVG namespace.
* Give SVG namespace to <svg> itself
* Don't allocate unnecessarily when reconciling portals
We create stacks lazily so that if portal doesn't contain <svg>s, we don't need to allocate.
We also reuse the same object for portal host context state instead of creating a new one every time.
* Add more tests for edge cases
* Fix up math namespace
* Maintain a separate container stack
* Fix rebase mistakes
* Unwind context on errors
* Reset the container state when reusing the object
* Add getChildHostContext() to ReactART
* Record tests
If a ref throws while detaching, it should not prevent
componentWillUnmount from being called.
Additionally, errors thrown by removeChild should not be ignored, even
as the result of unmounting a failed subtree.
This fixes the issue where using the .return pointer isn't guaranteed to
return the current Fiber so we might read the wrong props when we try
to get the current event.
When we perform an update to the event handler we properly update the
immediate Fiber pointer of a child to be the current. However, when we
bubble events we use the return pointer which is not guaranteed to point
to the current Fiber even if we start from the current.
This manifests itself when we bailout in a parent. So I made the tests
use a PureComponent to illustrate this scenario. There is already a failing
case but I'm adding another one too.
This strategy finds the current fiber. It traverses back up to the root if
the two trees are completely separate and determines which one is current.
If the two trees converge anywhere along the way, we assume that is the
current tree. We find the current child by searching the converged child
set.
This could fail if there's any way for both return pointers to point
backwards to the work in progress. I don't think there is but I could be
wrong.
This may also fail on coroutines where we have reparenting situations.
Abhay Nikam