fix(compiler-sfc): improve scoped style rewriting for nested and non-pseudo :is/:where selectors #12244

edison1105 · 2024-10-23T10:02:09Z

Summary by CodeRabbit

Bug Fixes
- Improved handling of scoped CSS when using nested :is() and :where() pseudo-classes, ensuring the scoped attribute is correctly applied in complex selector scenarios.
Tests
- Added new test cases to verify correct scoping behavior for nested and combined :is() and :where() selectors with ID prefixes.

github-actions · 2024-10-23T10:03:09Z

Size Report

Bundles

File	Size	Gzip	Brotli
runtime-dom.global.prod.js	101 kB	38.3 kB	34.5 kB
vue.global.prod.js	159 kB	58.5 kB	52 kB

Usages

Name	Size	Gzip	Brotli
createApp (CAPI only)	46.6 kB	18.2 kB	16.7 kB
createApp	54.5 kB	21.2 kB	19.4 kB
createSSRApp	58.8 kB	23 kB	20.9 kB
defineCustomElement	59.5 kB	22.8 kB	20.8 kB
overall	68.6 kB	26.4 kB	24.1 kB

pkg-pr-new · 2024-10-23T10:03:20Z

Open in StackBlitz

@vue/compiler-core

npm i https://pkg.pr.new/@vue/compiler-core@12244

@vue/compiler-dom

npm i https://pkg.pr.new/@vue/compiler-dom@12244

@vue/compiler-sfc

npm i https://pkg.pr.new/@vue/compiler-sfc@12244

@vue/compiler-ssr

npm i https://pkg.pr.new/@vue/compiler-ssr@12244

@vue/reactivity

npm i https://pkg.pr.new/@vue/reactivity@12244

@vue/runtime-core

npm i https://pkg.pr.new/@vue/runtime-core@12244

@vue/runtime-dom

npm i https://pkg.pr.new/@vue/runtime-dom@12244

@vue/server-renderer

npm i https://pkg.pr.new/@vue/server-renderer@12244

@vue/shared

npm i https://pkg.pr.new/@vue/shared@12244

vue

npm i https://pkg.pr.new/vue@12244

@vue/compat

npm i https://pkg.pr.new/@vue/compat@12244

commit: 4e2c116

zhangmo8 · 2024-10-24T09:04:39Z

This will cause pattern pollution and scoped will not work

zhangmo8 · 2024-10-24T09:24:01Z

Maybe it's another bug that I reproduced on the current official playground

I created a new issue to track it #12252

edison1105 · 2024-10-24T09:33:05Z

@zhangmo8

According to the document, I believe the behavior with this PR is proper.

zhangmo8 · 2024-10-24T09:59:26Z

This pr is correct, you are so cool @edison1105

skirtles-code · 2025-05-19T20:56:30Z

I'm wondering about nesting :is() inside :is():

Playground

A selector like :is(:is(.foo)) seems to be scoped correctly, but not #app :is(:is(.foo)).

coderabbitai · 2025-05-20T13:49:10Z

Walkthrough

The changes update the logic for handling scoped CSS selectors, specifically improving how nested :is() and :where() pseudo-classes are processed to ensure the scoped attribute is correctly injected. Additional tests were added to verify this behavior in complex selector scenarios, confirming accurate scoping within nested pseudo-classes.

Changes

File(s)	Change Summary
packages/compiler-sfc/src/style/pluginScoped.ts	Enhanced logic in `rewriteSelector` to better handle nested `:is()` and `:where()` pseudo-classes for scoping by adding a helper function and expanding conditions.
packages/compiler-sfc/tests/compileStyle.spec.ts	Added tests to verify correct scoping with nested and combined `:is()` and `:where()` selectors.

Sequence Diagram(s)

sequenceDiagram
    participant TestSuite
    participant Compiler
    participant SelectorRewriter

    TestSuite->>Compiler: Compile SFC with scoped CSS
    Compiler->>SelectorRewriter: Rewrite selectors for scoping
    SelectorRewriter->>SelectorRewriter: Check for :is()/:where() (including nesting)
    SelectorRewriter->>SelectorRewriter: Inject scoped attribute as needed
    SelectorRewriter-->>Compiler: Return rewritten selectors
    Compiler-->>TestSuite: Output compiled CSS

Assessment against linked issues

Objective	Addressed	Explanation
Ensure scoped CSS correctly handles pseudo-classes like :is()/:where(), including nested cases (#12241)	✅
Inject scoped attribute within selectors when using ID selector prefixes and pseudo-classes (#12241)	✅

Suggested reviewers

edison1105

Poem

In tangled CSS where selectors roam,
The rabbit hops through :is() and calls it home.
With scoping now clever, and nesting in view,
Each style finds its mark, as all rabbits knew.
So here’s to selectors, now handled with care—
No more lost styles, just harmony everywhere! 🐇✨

📜 Recent review details

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 759d5b4 and 4e2c116.

📒 Files selected for processing (2)

packages/compiler-sfc/__tests__/compileStyle.spec.ts (1 hunks)
packages/compiler-sfc/src/style/pluginScoped.ts (3 hunks)

🚧 Files skipped from review as they are similar to previous changes (2)

packages/compiler-sfc/src/style/pluginScoped.ts
packages/compiler-sfc/tests/compileStyle.spec.ts

✨ Finishing Touches

📝 Generate Docstrings

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Explain this complex logic.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai explain this code block.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and explain its main purpose.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

CodeRabbit Configuration File (`.coderabbit.yaml`)

You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
Please see the configuration documentation for more information.
If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

edison1105 · 2025-05-20T13:58:33Z

@skirtles-code
fixed via 8d9f49c

packages/compiler-sfc/src/style/pluginScoped.ts

Justineo · 2025-06-05T10:26:21Z

I think we may need to take a step back and consider a more comprehensive fix. It seems we’re not correctly identifying subject selectors in general when they don’t include element, class, or attribute selectors.

See this playground example.

The following input:

<style scoped>
:hover {
  color: red;
}

.main ~ :hover {
  color: red;
}

.main + :hover {
  color: red;
}

.main ::before {
  color: red;
}

.main :is(.title) {
  color: red;
}
</style>

is compiled to

[data-v-7ba5bd90]:hover {
  color: red;
}
.main[data-v-7ba5bd90] ~ :hover {
  color: red;
}
.main[data-v-7ba5bd90] + :hover {
  color: red;
}
.main[data-v-7ba5bd90] ::before {
  color: red;
}
.main[data-v-7ba5bd90] :is(.title) {
  color: red;
}

while it should be:

[data-v-7ba5bd90]:hover {
  color: red;
}
.main ~ [data-v-7ba5bd90]:hover {
  color: red;
}
.main + [data-v-7ba5bd90]:hover {
  color: red;
}
.main [data-v-7ba5bd90]::before {
  color: red;
}
.main [data-v-7ba5bd90]:is(.title) {
  color: red;
}

skirtles-code · 2025-06-05T11:05:32Z

@Justineo You may well be right. I did wonder about that when reviewing this PR, but I thought it was a small step in the right direction, which might prove easier than trying to make a big leap.

While testing this PR, I avoided testing with other pseudos, because they're known to be broken: #8868. I didn't want to increase the scope of this PR unnecessarily.

Different pseudos need different handling. For :is() and :where() we need to inject the scoping inside the argument, whereas for :not() and :has() or pseudos without arguments it needs to come before. With that in mind, I thought it was fine for this PR to just handle :is() and :where(), as those seemed to be a special case.

With hindsight, maybe I was missing a more important point. Maybe we already have all the special casing we need for :is() and :where(), and the underlying bug here isn't specific to :is() and :where() at all. Maybe it is just #8868.

Justineo · 2025-06-05T11:39:05Z

Why do we need to inject attribute selectors inside the arguments of :is() and :where()?

skirtles-code · 2025-06-05T16:31:38Z

@Justineo Good question.

I spent some time going through the history of this. I think the most enlightening was #8929. Long story short, it's for :deep() support.

The motivation was Tailwind's use of :is()/:where(), which it wraps around the selector from the code it's passed.

In Vue 3.5.16 (and other recent versions) we compile this:

:is(a b) {}
:is(a :deep(b)) {}

to this:

:is(a b[data-v-7ba5bd90]) {}
:is(a[data-v-7ba5bd90] b) {}

The first line could compile to [data-v-7ba5bd90]:is(a b) {} instead, but the second line can't be rewritten like that.

It isn't consistent, though. There are lots of edge cases, some of which are fixed by this PR, some of which aren't. e.g. div:where(.a :deep(.b)).

Getting the specificity right when combining :where() and :deep() is also a bit of a headache. It's not clear to me whether there's a good solution for that.

I wonder whether the :scoped() pseudo proposed in #13405 might help here too. I think the intention behind :deep() was to hide the details of how scoping is implemented from the user, keep everything abstract. But I'm not convinced that's working out in practice. Increasingly it feels like a leaky abstraction where you have to figure out what selector you really want and then work back to figure out how to trick the compiler into doing that for you. :scoped() would let you put the scoping exactly where you want it, a bit like :slotted(). (Sorry, may have gone off on a tangent...)

Justineo · 2025-06-06T05:06:55Z

Hmm… this is turning out to be more complex than I expected — especially when combining Vue’s selector extensions (:deep(), :slotted(), :global()) with :is(), :where(), :not(), and :has(). And that’s not even counting native CSS nesting.

I’ve put together a (very incomplete) table that only includes :deep() and :is():

Selector input	Current output	(My) Preferred output	Strict output^*
`.a .b .c`	`.a .b .c[data-v]`		`.a[data-v] .b[data-v] .c[data-v]`
`.a .b :is(.c)`	`.a .b[data-v] :is(.c)`	`.a .b :is(.c[data-v])`? `.a .b [data-v]:is(.c)`?	`.a[data-v] .b[data-v] :is(.c[data-v])`
`.a :is(.b .c)`	`.a[data-v] :is(.b .c)`	`.a :is(.b .c[data-v])`? `.a [data-v]:is(.b .c)`?	`.a[data-v] :is(.b[data-v] .c[data-v])`
`:is(.a .b) .c`	`:is(.a .b) .c[data-v]`		`:is(.a[data-v] .b[data-v]) .c[data-v]`
`:is(.a .b .c)`	`:is(.a .b .c[data-v])`	`[data-v]:is(.a .b .c)`?	`:is(.a[data-v] .b[data-v] .c[data-v])`
`.a .b :deep(.c)`	`.a .b[data-v] .c`		`.a[data-v] .b[data-v] .c`
`.a :deep(.b) .c`	`.a[data-v] .b .c`	`.a .b [data-v].c`	`.a[data-v] .b .c[data-v]`
`.a :deep(.b .c)`	`.a[data-v] .b .c`
`:deep(.a .b .c)`	`[data-v] .a .b .c`	`.a .b .c`?
`.a .b :is(:deep(.c))`	`.a .b[data-v] :is(:deep(.c))`	`.a .b[data-v] :is(.c)`	`.a[data-v] .b[data-v] :is(.c)`
`.a .b :deep(:is(.c))`	`.a .b[data-v] :is(.c)`		`.a[data-v] .b[data-v] :is(.c)`
`.a :is(.b :deep(.c))`	`.a[data-v] :is(.b :deep(.c))`	`.a :is(.b[data-v] .c)`	`.a[data-v] :is(.b[data-v] .c)`
`.a :deep(.b :is(.c))`	`.a[data-v] .b :is(.c)`
`.a :is(:deep(.b .c))`	`.a[data-v] :is(:deep(.b .c))`	`.a[data-v] :is(.b .c)`
`.a :deep(:is(.b .c))`	`.a[data-v] :is(.b .c)`
`:is(:deep(.a .b .c))`	`:is([data-v] .a .b .c)`	`:is(.a .b .c)`?
`:deep(:is(.a .b .c))`	`[data-v] :is(.a .b .c)`	`:is(.a .b .c)`?

*: for minimal specificity impact [data-v] would become :where([data-v]).

I hope there’s a way to establish clear guidelines that can help ensure we implement scoped styles as accurately as possible.

For reference, here's the compiled output of the above code by Svelte (Svelte only has a :global() extension): Svelte Playground

Spoiler: Svelte aligns with my mental model and consistently increases specificity by (0, 1, 0).

fix(compiler-sfc): properly rewrite :is with non-pseudo selector

afbda36

edison1105 added ready to merge The PR is ready to be merged. scope: sfc labels Oct 23, 2024

edison1105 added the 🔨 p3-minor-bug Priority 3: this fixes a bug, but is an edge case that only affects very specific usage. label Jan 7, 2025

test: add more tests

8d9f49c

edison1105 and others added 2 commits May 20, 2025 21:53

Merge branch 'main' into edison/fix/12241

47c0f03

[autofix.ci] apply automated fixes

74dfcf7

edison1105 changed the title ~~fix(compiler-sfc): properly rewrite :is with non-pseudo selector~~ fix(compiler-sfc): improve scoped style rewriting for nested and non-pseudo :is/:where selectors May 20, 2025

skirtles-code reviewed May 20, 2025

View reviewed changes

packages/compiler-sfc/src/style/pluginScoped.ts Outdated Show resolved Hide resolved

edison1105 and others added 3 commits May 21, 2025 08:22

test: add more tests

759d5b4

Merge branch 'main' into edison/fix/12241

ec7ff38

[autofix.ci] apply automated fixes

4e2c116

Uh oh!

fix(compiler-sfc): improve scoped style rewriting for nested and non-pseudo :is/:where selectors #12244

Are you sure you want to change the base?

fix(compiler-sfc): improve scoped style rewriting for nested and non-pseudo :is/:where selectors #12244

Uh oh!

Conversation

edison1105 commented Oct 23, 2024 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Uh oh!

github-actions bot commented Oct 23, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Size Report

Bundles

Usages

Uh oh!

pkg-pr-new bot commented Oct 23, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zhangmo8 commented Oct 24, 2024

Uh oh!

zhangmo8 commented Oct 24, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

edison1105 commented Oct 24, 2024

Uh oh!

zhangmo8 commented Oct 24, 2024

Uh oh!

skirtles-code commented May 19, 2025

Uh oh!

coderabbitai bot commented May 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Sequence Diagram(s)

Assessment against linked issues

Suggested reviewers

Poem

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

CodeRabbit Configuration File (.coderabbit.yaml)

Documentation and Community

Uh oh!

edison1105 commented May 20, 2025

Uh oh!

Uh oh!

Justineo commented Jun 5, 2025

Uh oh!

skirtles-code commented Jun 5, 2025

Uh oh!

Justineo commented Jun 5, 2025

Uh oh!

skirtles-code commented Jun 5, 2025

Uh oh!

Justineo commented Jun 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

edison1105 commented Oct 23, 2024 •

edited by coderabbitai bot

Loading

github-actions bot commented Oct 23, 2024 •

edited

Loading

pkg-pr-new bot commented Oct 23, 2024 •

edited

Loading

zhangmo8 commented Oct 24, 2024 •

edited

Loading

coderabbitai bot commented May 20, 2025 •

edited

Loading

CodeRabbit Configuration File (`.coderabbit.yaml`)

Justineo commented Jun 6, 2025 •

edited

Loading