refactor: reduce EXPECTED_ANY in stats hook and printer types#21198
Merged
Conversation
Recover the real inner-hook and result types in StatsFactory and StatsPrinter via HookMap/Hook inference instead of EXPECTED_ANY for the level-hook caches and _forEachLevel result generics.
Cast the compilation printer to its real signature like the sibling asset/issuer/profile dispatch loops; the remaining loops dispatch over heterogeneous printer unions that have no callable common signature.
Replace the EXPECTED_ANY FactoryData typedef in _forEachLevelWaterfall and _forEachLevelFilter with a generic D template defaulting to EXPECTED_ANY, so the caller's item type flows through; also fixes _forEachLevelFilter declaring its return as the hook result type (boolean | void) instead of the item type.
Cast the heterogeneous printer dispatches to SimplePrinter instead of a whole-function EXPECTED_ANY, so the printer context and return value are type-checked; only the printed value stays EXPECTED_ANY.
countWithChildren only reads .length, so type getItems as
() => { length: number } instead of a generic T[]; this lets the
warnings counter return its real Error[]/StatsError[] values and drops
the EXPECTED_ANY[] cast that previously bridged them.
🦋 Changeset detectedLatest commit: 50dcb92 The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Contributor
|
This PR is packaged and the instant preview is available (b15a94c). Install it locally:
npm i -D webpack@https://pkg.pr.new/webpack@b15a94c
yarn add -D webpack@https://pkg.pr.new/webpack@b15a94c
pnpm add -D webpack@https://pkg.pr.new/webpack@b15a94c |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #21198 +/- ##
==========================================
- Coverage 92.70% 92.69% -0.01%
==========================================
Files 588 588
Lines 64118 64120 +2
Branches 17797 17798 +1
==========================================
- Hits 59440 59438 -2
- Misses 4678 4682 +4
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
The no-filter branch of warningsCount was guarded by `!warningsFilter && warningsFilter.length === 0`, which is never true because warningsFilter is always normalized to an array. Drop the dead `!warningsFilter` term so the empty-filter case returns the raw warnings for counting instead of running factory.create().filter(); warningsCount is unchanged (stats snapshots pass).
Merging this PR will improve performance by 30.5%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Memory | benchmark "side-effects-reexport", scenario '{"name":"mode-development-rebuild","mode":"development","watch":true}' |
766.9 KB | 1,205.2 KB | -36.37% |
| ❌ | Memory | benchmark "many-modules-esm", scenario '{"name":"mode-production","mode":"production"}' |
7.8 MB | 10 MB | -22.59% |
| ⚡ | Memory | benchmark "asset-modules-inline", scenario '{"name":"mode-development-rebuild","mode":"development","watch":true}' |
1,293.8 KB | 324.5 KB | ×4 |
| ⚡ | Memory | benchmark "css-modules", scenario '{"name":"mode-production","mode":"production"}' |
10 MB | 6.8 MB | +47.68% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing claude/stats-expected-any-7e1x0g (50dcb92) with main (0db8bbb)
Covers the warnings-count branch fixed in this PR: emits warnings with no warningsFilter and shows only warningsCount, asserting the raw count is reported.
Contributor
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
lib/stats/usedEXPECTED_ANY(alias forany) in many spots where a real type is recoverable. This PR tightens the recoverable ones while keepingEXPECTED_ANYonly where it is genuinely existential:StatsFactory/StatsPrinterhook plumbing: infer the real inner-hook and result types for the level-hook caches and_forEachLevel*generics (also fixes theresultcache being typed asSyncBailHookinstead ofSyncWaterfallHook).StatsFactorylevel helpers: thread a@template [D=EXPECTED_ANY]item type through_forEachLevelWaterfall/_forEachLevelFilter(this also fixes_forEachLevelFilterdeclaring its return as the hook resultboolean | voidinstead of the item type).DefaultStatsPrinterPlugindispatch loops: type the compilation loop with its real signature and the remaining loops asSimplePrinter<EXPECTED_ANY, …>so the printer context and return value are checked (only the printed value staysEXPECTED_ANY).countWithChildren: it only reads.length, so it now takes() => { length: number }instead of a genericT[], dropping theEXPECTED_ANY[]cast that bridgedError[]/StatsError[].While dropping that cast, the
warningsCountno-filter branch was found to be dead: it was guarded by!warningsFilter && warningsFilter.length === 0, which is never true becausewarningsFilteris always normalized to an array. The guard is corrected towarningsFilter.length === 0, so the empty-filter case returns the raw warnings for counting instead of runningfactory.create(...).filter(...)— a small perf win with an unchangedwarningsCount(stats snapshots pass).The remaining
EXPECTED_ANYusages are intentional: plugin-extensible stats records (Known* & Record<string, EXPECTED_ANY>), existential generics (Comparator,GroupConfig), dynamic factory/printer data, and runtime-string hook dispatch. No public type surface (types.d.tsunchanged).What kind of change does this PR introduce?
refactor (primarily a type cleanup; includes one small
perffix to the dead warnings-count branch surfaced by the refactor)Did you add tests for your changes?
Yes — added
test/statsCases/warnings-countassertingwarningsCountis reported without awarningsFilter(covering the fixed branch). The type changes are also exercised by the existingtest/StatsTestCases(now 132 cases, all passing).Does this PR introduce a breaking change?
No.
If relevant, what needs to be documented once your changes are merged or what have you already documented?
n/a
Use of AI
AI (Claude Code) was used to inventory the
EXPECTED_ANYusages, implement the type changes and the warnings-count fix, add the test, and verify each step withyarn tscand the stats tests; all changes were reviewed.