feat(html): improved parsing spread attributes

[ematipico]

Summary

Closes #8887
Part of #8590
Closes #8851

I believe we're almost there, and we are stable enough to finally say that Svelte is definitely more stable and usable.

Test Plan

Added new tests

Docs

N/A

[changeset-bot]

🦋 Changeset detected

Latest commit: d1930a3

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 13 packages

Name	Type
@biomejs/biome	Patch
@biomejs/cli-win32-x64	Patch
@biomejs/cli-win32-arm64	Patch
@biomejs/cli-darwin-x64	Patch
@biomejs/cli-darwin-arm64	Patch
@biomejs/cli-linux-x64	Patch
@biomejs/cli-linux-arm64	Patch
@biomejs/cli-linux-x64-musl	Patch
@biomejs/cli-linux-arm64-musl	Patch
@biomejs/wasm-web	Patch
@biomejs/wasm-bundler	Patch
@biomejs/wasm-nodejs	Patch
@biomejs/backend-jsonrpc	Patch

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

[codspeed-hq]

CodSpeed Performance Report

Merging this PR will not alter performance

_{Comparing feat/spread-html-attribute (d1930a3) with next (fda74af)}

Summary

✅ 4 untouched benchmarks
⏩ 152 skipped benchmarks¹

Footnotes

1. 152 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[coderabbitai]

Walkthrough

This PR adds parsing support for spread attributes in Svelte and Astro files, enabling syntax like {...props} on HTML elements and components. It introduces new parser functions for both frameworks, updates the HTML syntax grammar to include HtmlSpreadAttribute, adds corresponding formatter implementations, and extends test coverage with both valid and error cases across multiple test suites.

Possibly related PRs

• feat(parse/html): parse svelte shorthand props and spread props #7872: Implements parsing and formatting for Svelte/Astro spread and shorthand attributes, adding HTML_SPREAD_ATTRIBUTE and related parser functions.
• feat(core): improved bindings detection #8642: Adds tokenizer/parser/AST and formatter support for three-dot spread syntax in HTML/Svelte/Astro with HTML_SPREAD_ATTRIBUTE and DOT3 handling.
• fix(html): improved svelte parsing #8555: Modifies Svelte HTML parsing for single-text/spread expressions, affecting the parser flow that this PR extends.

Suggested reviewers

• dyc3

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: adding improved parsing support for spread attributes in HTML-like files (Svelte/Astro).
Linked Issues check	✅ Passed	The PR successfully implements spread attribute parsing for Svelte and Astro, addressing the core requirement from issue #8887.
Out of Scope Changes check	✅ Passed	All changes are directly scoped to implementing spread attribute parsing: syntax definitions, parsers, formatters, and tests.
Description check	✅ Passed	The PR description clearly relates to the changeset by explaining that spread attributes parsing support is being added for Svelte and Astro, with references to specific closed/related issues.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 4

🤖 Fix all issues with AI agents

In `@crates/biome_html_parser/src/syntax/mod.rs`:
- Around line 739-758: The current condition in parse_name misuses
should_bail_if and causes tokens to be consumed when callers pass |_| true;
change the early-return to bail when a name is missing AND should_bail_if(p) is
true (i.e., replace `if !p.at(IDENT) && !should_bail_if(p) { return Absent; }`
with `if !p.at(IDENT) && should_bail_if(p) { return Absent; }`) so missing names
fail instead of being consumed, and update any call sites that relied on the old
inverted semantics of the should_bail_if closure.
- Around line 465-472: The match arm ordering currently checks
SingleTextExpressions before Astro, so when both flags are enabled (Astro
frontmatter + SingleTextExpressions) the parse_astro_spread_or_expression branch
is shadowed; swap the two guards so the T!['{'] if Astro.is_supported(p) =>
parse_astro_spread_or_expression(p) arm comes before the T!['{'] if
SingleTextExpressions.is_supported(p) => parse_svelte_spread_or_expression(p)
arm (leave the fallback T!['{'] exclusive parse_exclusive_syntax entry
unchanged) to ensure Astro-specific parsing runs first.

In `@crates/biome_html_parser/src/syntax/svelte.rs`:
- Around line 359-389: parse_svelte_spread_or_expression currently calls
parse_name after consuming "..." which only accepts a single identifier; replace
that with the general expression parser so spread attributes accept arbitrary JS
expressions. Specifically, in parse_svelte_spread_or_expression, instead of
parse_name(p, HtmlLexContext::Svelte, ...), invoke the module's general
expression parsing entry (the same parser used for text expressions—e.g., the
logic behind parse_single_text_expression or a parse_expression(p,
HtmlLexContext::Svelte) helper) to parse any valid expression, keep the existing
diagnostic (.or_add_diagnostic(...)) if parsing fails, and then expect the
closing '}' as before before completing HTML_SPREAD_ATTRIBUTE.

In `@xtask/codegen/html.ungram`:
- Line 212: The grammar rule for spread attribute currently restricts the
argument to an identifier via "HtmlName = 'ident'" / "argument: HtmlName", which
rejects valid JS expressions (e.g., {...foo.bar}, {...getProps()}, {...cond ? a
: b}); change the grammar so the spread attribute argument uses a generic
expression production instead of HtmlName (e.g., use the existing Expression or
JsExpression nonterminal) and update the parser code handling the spread
attribute (the parser branch that consumes "argument" in the spread-attribute
handling code) to parse and store a full expression node rather than only an
identifier; apply the same change to the other spread-attribute occurrences
referenced (the second spread-attribute production area) so both grammar and
parser accept arbitrary expressions.

[dyc3]

I tested this on a private repo I have and https://github.com/huntabyte/shadcn-svelte. This definitely resolves a lot of the parsing errors, but IMO there's still too many false positives for the rules to fully close #8590.

[ematipico]

I tested this on a private repo I have and huntabyte/shadcn-svelte. This definitely resolves a lot of the parsing errors, but IMO there's still too many false positives for the rules to fully close #8590.

Yeah definitely. I updated the PR description, and we could use that issue to collect issues/bugs

[coderabbitai]

Actionable comments posted: 0

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

crates/biome_html_parser/src/syntax/mod.rs (1)

427-448: Remove duplicate match arm.

The second T!["{{"] arm (lines 438-448) is unreachable—Rust's match evaluates top-to-bottom, so the first arm at lines 427-436 always wins. This looks like a merge artefact.

🐛 Proposed fix

         T!["{{"] => {
             let m = p.start();
             DoubleTextExpressions
                 .parse_exclusive_syntax(
                     p,
                     |p| parse_double_text_expression(p, HtmlLexContext::InsideTag),
                     |p, marker| disabled_interpolation(p, marker.range(p)),
                 )
                 .ok();
             Present(m.complete(p, HTML_ATTRIBUTE))
         }
-        T!["{{"] => {
-            let m = p.start();
-            HtmlSyntaxFeatures::DoubleTextExpressions
-                .parse_exclusive_syntax(
-                    p,
-                    |p| parse_double_text_expression(p, HtmlLexContext::InsideTag),
-                    |p, marker| disabled_interpolation(p, marker.range(p)),
-                )
-                .ok();
-
-            Present(m.complete(p, HTML_ATTRIBUTE))
-        }

🧹 Nitpick comments (1)

crates/biome_html_parser/src/syntax/astro.rs (1)

46-77: LGTM!

The implementation correctly handles Astro spread attributes and mirrors the Svelte implementation. The comment explaining why HtmlLexContext::Svelte is used for lexing ... is helpful.

The logic is nearly identical to parse_svelte_spread_or_expression in svelte.rs. If spread parsing grows more complex, consider extracting a shared helper that takes the feature check as a parameter.

[dyc3]

I believe this should also close #8851

When I get a chance, I'll start filing issues related to #8590