crates/biome_service/src/file_handlers/css.rs (1)

616-625: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚖️ Poor tradeoff

Same module_db cfg incantation copy-pasted eight times.

This exact block appears across lint, code_actions, and three fix_all sites here (and again in html.rs). Consider exposing a single always-compiled helper on WorkspaceDb so the cfg lives in one place:

♻️ Suggested helper (in biome_workspace_db)

impl WorkspaceDb {
    /// Returns the module DB when the `module_graph` feature is enabled, else `None`.
    pub fn optional_module_db(&self) -> Option<Rc<dyn ModuleDb>> {
        #[cfg(feature = "module_graph")]
        { Some(self.rc_module_db()) }
        #[cfg(not(feature = "module_graph"))]
        { None }
    }
}

Each call site then collapses to module_db: params.workspace_db.optional_module_db(),. Purely optional, but it trims eight copies that would otherwise have to be kept in lockstep forever.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@crates/biome_service/src/file_handlers/css.rs` around lines 616 - 625, The
`module_db` feature-gating logic is duplicated across multiple call sites,
including the `css.rs` handlers and related `html.rs` paths. Add a single
always-compiled helper on `WorkspaceDb` (for example, `optional_module_db`) that
encapsulates the `#[cfg(feature = "module_graph")]` / `#[cfg(not(feature =
"module_graph"))]` split and returns the optional module DB. Then update the
`module_db:` assignments in `lint`, `code_actions`, and the `fix_all` sites to
call that helper instead of repeating the cfg block.

crates/biome_service/Cargo.toml (1)

170-172: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low value

Two unresolved TODOs about feature dependencies.

Lines 170 and 202 leave open questions on whether module_graph and type_inference need extra lang_* deps. Worth resolving before merge so the matrix isn't a guess.

Happy to draft a verification script that maps the cfg(feature=...) usages in the relevant source files to confirm the correct dependency set — just say the word.

Also applies to: 202-203

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@crates/biome_service/Cargo.toml` around lines 170 - 172, Resolve the
unresolved feature-dependency TODOs in Cargo.toml by verifying which lang_*
crates are actually required for module_graph and type_inference. Check the
cfg(feature=...) usage in the relevant code paths and update the feature lists
for module_graph, type_inference, and any related entries so the declared
dependencies match the real usage instead of leaving the matrix as a guess.

crates/biome_service/src/file_handlers/javascript.rs (1)

1536-1547: 🩺 Stability & Availability | 🔵 Trivial | 💤 Low value

Prefer a graceful fallback over panic! here, to match the HTML handler.

The #[cfg(not(feature = "js_embeds"))] stub panics, whereas file_handlers/html.rs falls back to plain format(...) in the same situation. It's currently unreachable (the non-js_embeds parse_embedded_nodes returns empty, so embedded_nodes is never non-empty), but mirroring HTML keeps the two handlers symmetric and removes a latent foot-gun if capability wiring ever changes.

♻️ Fall back to format

     #[cfg(not(feature = "js_embeds"))]
     {
-        let _ = (
-            biome_path,
-            document_file_source,
-            parse,
-            settings,
-            embedded_nodes,
-            workspace_db,
-        );
-        panic!("formatting embedded JavaScript snippets requires the `js_embeds` feature")
+        let _ = embedded_nodes;
+        format(biome_path, document_file_source, parse, settings, workspace_db)
     }

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@crates/biome_service/src/file_handlers/javascript.rs` around lines 1536 -
1547, The `#[cfg(not(feature = "js_embeds"))]` branch in `parse_embedded_nodes`
should not `panic!`; mirror the HTML handler’s behavior and fall back to the
normal `format(...)` path instead. Update the JavaScript file handler in
`javascript.rs` so the non-`js_embeds` stub remains a graceful no-op/formatting
fallback, using the same surrounding logic and symbols like
`parse_embedded_nodes`, `embedded_nodes`, and `format` to keep the handlers
symmetric.

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes