Codestin Search App

aryanraj45 · 2026-02-25T18:37:11Z

Fixes #17063
Summary

In harmony (ESM) mode, HarmonyDetectionParserPlugin blocks AMD processing by returning true from a skipInHarmony tap on hooks.call.for("define"). This bail-out happens before the parser walks expression.arguments, so import bindings inside define() callbacks are never renamed causing a ReferenceError at runtime.

Example that was broken:

import { foo } from "./foo.js";

define(["dep"], function () {
    foo(); // ReferenceError: foo is not defined
});

Webpack correctly renames foo everywhere else, but skips it here because skipInHarmony bails walkCallExpression before it reaches the arguments.

Fix: add a stage: -1 tap on hooks.call.for("define") that walks expression.arguments before skipInHarmony (stage 0) fires. This ensures bindings are renamed, while AMD processing is still correctly skipped in harmony mode.

What kind of change does this PR introduce?

fix

Did you add tests for your changes?

Yes. Added test/cases/parsing/issue-17063/ — an ESM file that imports a binding and calls it inside define(). Without the fix this throws ReferenceError. With the fix it passes.

Also added test/configCases/target/web-webworker/ and test/configCases/target/web-webworker-auto-public-path/ which were previously missing coverage for the ["web", "webworker"] target combination.

Does this PR introduce a breaking change?

No. The fix only activates when HarmonyExports.isEnabled() is true (ESM files). Plain CJS files are unaffected.

If relevant, what needs to be documented once your changes are merged?

No documentation needed. This is a silent runtime bug fix with no API or config changes.]

changeset-bot · 2026-02-25T18:37:18Z

⚠️ No Changeset found

Latest commit: e061c66

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

alexander-akait · 2026-02-25T18:42:12Z

Wrong solution, no define in ECMA modules

aryanraj45 · 2026-02-25T19:51:27Z

yeah yeah totally Agree, define has no special meaning in ESM.
But the concern is how the parser handles it in ESM it's just a regular identifier, so shouldn't the parser still walk its arguments like any other function call?
Right now skipInHarmony bails before walking the AST, so import bindings inside the callback never get renamed and webpack silently emits broken output.

Would it make sense to just walk the arguments without treating it as AMD at all?

alexander-akait · 2026-02-25T20:05:49Z

We should not just handle it

aryanraj45 · 2026-02-25T20:30:59Z

okay how about we would a build warning like "define is not supported in EcmaScript modules" work? So webpack actively flags it instead of silently emitting broken output

alexander-akait · 2026-02-25T20:47:30Z

We should output nothing, define is just a function in ECMA modules, and we should not break code too

aryanraj45 · 2026-02-25T21:24:09Z

Thanks for the suggestion Alexander Since define behaves like a regular function in ESM, I’ve removed the Stage -1 hack and updated the logic to simply walk through its arguments so any import bindings inside it don’t break, and then return true to prevent the AMD plugin from processing it.

alexander-akait · 2026-02-25T21:46:49Z

test/configCases/target/web-webworker-auto-public-path/webpack.config.js

+	optimization: {
+		minimize: false
+	}
+};


Please avoid pushing not related things

so so sorry actually webworker one was my another test for the upcoming PR by mistake it came under this branch and pushed in this commit .will surely take care of this in future also I have fixed it and kept only [test/cases/parsing/issue-17063/]

aryanraj45 · 2026-02-26T15:33:09Z

@alexander-akait can u please review once again whenever u have a moment
thank you

codspeed-hq · 2026-02-27T13:16:20Z

Merging this PR will not alter performance

✅ 144 untouched benchmarks

_{Comparing aryanraj45:fix/define-harmony-import-binding (bd8eb04) with main (b8e9b05)}

alexander-akait · 2026-02-27T14:13:52Z

Tests are failed, we need improve logic here

aryanraj45 · 2026-02-28T13:01:10Z

Tests are failed, we need improve logic here

I think the CI is failing because we're blocking the AMD plugin from processing define. Would it be okay if I just stop skipping define in ESM and let the AMD plugin handle it naturally?

alexander-akait · 2026-02-28T14:51:07Z

I don't know, that is why it is a bug, it requires investigate

aryanraj45 · 2026-02-28T15:25:56Z

I know we usually avoid define in ESM, but blocking it actually breaks about 25+ existing tests that seem to mix both. Letting the AMD plugin handle it naturally fixes the bug (#17063) and keeps the CI green (coz i have tested locally both 45/45 test passed and even the new test case which I added in my latest push). Can we trigger a re-run to check? Thanks!

Copilot

Pull request overview

Fixes a Harmony (ESM) parsing edge case where define(...) calls were bailing out before their arguments were walked, preventing imported bindings used inside define() callbacks from being renamed and causing runtime ReferenceErrors.

Changes:

Add a pre-bail (stage -1) define call hook in HarmonyDetectionParserPlugin to walk expression.arguments in Harmony mode.
Add regression tests covering an ESM import referenced inside a define() callback.
Add a configCase covering the same scenario in the configCases harness.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
lib/dependencies/HarmonyDetectionParserPlugin.js	Walks `define()` call arguments in Harmony mode before the existing “skip in harmony” bail-out runs.
test/cases/parsing/issue-17063/index.js	Regression test ensuring an imported binding used inside `define()` callback is correctly renamed.
test/cases/parsing/issue-17063/foo.js	Fixture module returning a value used by the regression test.
test/configCases/amd/define-harmony-import/index.js	ConfigCase reproduction asserting the imported default binding is usable inside `define()` callback.
test/configCases/amd/define-harmony-import/lib.js	Fixture default export used by the configCase.
test/configCases/amd/define-harmony-import/webpack.config.js	Minimal config for the new configCase.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-01T18:16:41Z

lib/dependencies/HarmonyDetectionParserPlugin.js

+		parser.hooks.call
+			.for("define")
+			.tap({ name: PLUGIN_NAME, stage: -1 }, (expression) => {
+				if (HarmonyExports.isEnabled(parser.state) && expression.arguments) {
+					parser.walkExpressions(expression.arguments);
+				}
+			});


PR description says it also adds test/configCases/target/web-webworker/ and test/configCases/target/web-webworker-auto-public-path/, but those directories/files don't appear in the changes. Please either include those missing test additions or update the PR description so it matches what’s actually in the diff.

aryanraj45 force-pushed the fix/define-harmony-import-binding branch from 7601dc9 to ac46443 Compare February 25, 2026 21:23

alexander-akait reviewed Feb 25, 2026

View reviewed changes

aryanraj45 force-pushed the fix/define-harmony-import-binding branch from ac46443 to 8311efc Compare February 25, 2026 22:25

aryanraj45 requested a review from alexander-akait February 26, 2026 14:13

aryanraj45 force-pushed the fix/define-harmony-import-binding branch from 8311efc to bd8eb04 Compare February 28, 2026 13:30

fix: walk define arguments in harmony modules (fixes webpack#17063)

e061c66

aryanraj45 force-pushed the fix/define-harmony-import-binding branch from bd8eb04 to e061c66 Compare March 1, 2026 18:10

Copilot AI review requested due to automatic review settings March 1, 2026 18:10

Copilot started reviewing on behalf of aryanraj45 March 1, 2026 18:11 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Uh oh!

Conversation

aryanraj45 commented Feb 25, 2026

Uh oh!

changeset-bot bot commented Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚠️ No Changeset found

Uh oh!

alexander-akait commented Feb 25, 2026

Uh oh!

aryanraj45 commented Feb 25, 2026

Uh oh!

alexander-akait commented Feb 25, 2026

Uh oh!

aryanraj45 commented Feb 25, 2026

Uh oh!

alexander-akait commented Feb 25, 2026

Uh oh!

aryanraj45 commented Feb 25, 2026

Uh oh!

alexander-akait Feb 25, 2026

Choose a reason for hiding this comment

Uh oh!

aryanraj45 Feb 25, 2026

Choose a reason for hiding this comment

Uh oh!

aryanraj45 commented Feb 26, 2026

Uh oh!

codspeed-hq bot commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Uh oh!

alexander-akait commented Feb 27, 2026

Uh oh!

aryanraj45 commented Feb 28, 2026

Uh oh!

alexander-akait commented Feb 28, 2026

Uh oh!

aryanraj45 commented Feb 28, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

changeset-bot bot commented Feb 25, 2026 •

edited

Loading

codspeed-hq bot commented Feb 27, 2026 •

edited

Loading