Merge pull request rust-lang#19864 from ChayimFriedman2/is-in-macro

fix: Properly implement `might_be_inside_macro_call()` using semantic information instead of syntactical hacks

Author: Veykril

src/tools/rust-analyzer/crates/hir/src/semantics.rs

@@ -407,14 +407,14 @@ impl<'db> SemanticsImpl<'db> {
         res
     }
 
-    pub fn expand_macro_call(&self, macro_call: &ast::MacroCall) -> Option<SyntaxNode> {
+    pub fn expand_macro_call(&self, macro_call: &ast::MacroCall) -> Option<InFile<SyntaxNode>> {
         let sa = self.analyze_no_infer(macro_call.syntax())?;
 
         let macro_call = InFile::new(sa.file_id, macro_call);
         let file_id = sa.expand(self.db, macro_call)?;
 
         let node = self.parse_or_expand(file_id.into());
-        Some(node)
+        Some(InFile::new(file_id.into(), node))
     }
 
     pub fn check_cfg_attr(&self, attr: &ast::TokenTree) -> Option<bool> {
@@ -468,10 +468,10 @@ impl<'db> SemanticsImpl<'db> {
     }
 
     /// If `item` has an attribute macro attached to it, expands it.
-    pub fn expand_attr_macro(&self, item: &ast::Item) -> Option<ExpandResult<SyntaxNode>> {
+    pub fn expand_attr_macro(&self, item: &ast::Item) -> Option<ExpandResult<InFile<SyntaxNode>>> {
         let src = self.wrap_node_infile(item.clone());
         let macro_call_id = self.with_ctx(|ctx| ctx.item_to_macro_call(src.as_ref()))?;
-        Some(self.expand(macro_call_id))
+        Some(self.expand(macro_call_id).map(|it| InFile::new(macro_call_id.into(), it)))
     }
 
     pub fn expand_derive_as_pseudo_attr_macro(&self, attr: &ast::Attr) -> Option<SyntaxNode> {
@@ -846,49 +846,35 @@ impl<'db> SemanticsImpl<'db> {
         res
     }
 
-    // FIXME: This isn't quite right wrt to inner attributes
-    /// Does a syntactic traversal to check whether this token might be inside a macro call
-    pub fn might_be_inside_macro_call(&self, token: &SyntaxToken) -> bool {
-        token.parent_ancestors().any(|ancestor| {
+    pub fn is_inside_macro_call(&self, token: InFile<&SyntaxToken>) -> bool {
+        // FIXME: Maybe `ancestors_with_macros()` is more suitable here? Currently
+        // this is only used on real (not macro) files so this is not a problem.
+        token.value.parent_ancestors().any(|ancestor| {
             if ast::MacroCall::can_cast(ancestor.kind()) {
                 return true;
             }
-            // Check if it is an item (only items can have macro attributes) that has a non-builtin attribute.
-            let Some(item) = ast::Item::cast(ancestor) else { return false };
-            item.attrs().any(|attr| {
-                let Some(meta) = attr.meta() else { return false };
-                let Some(path) = meta.path() else { return false };
-                if let Some(attr_name) = path.as_single_name_ref() {
-                    let attr_name = attr_name.text();
-                    let attr_name = Symbol::intern(attr_name.as_str());
-                    if attr_name == sym::derive {
-                        return true;
-                    }
-                    // We ignore `#[test]` and friends in the def map, so we cannot expand them.
-                    // FIXME: We match by text. This is both hacky and incorrect (people can, and do, create
-                    // other macros named `test`). We cannot fix that unfortunately because we use this method
-                    // for speculative expansion in completion, which we cannot analyze. Fortunately, most macros
-                    // named `test` are test-like, meaning their expansion is not terribly important for IDE.
-                    if attr_name == sym::test
-                        || attr_name == sym::bench
-                        || attr_name == sym::test_case
-                        || find_builtin_attr_idx(&attr_name).is_some()
-                    {
-                        return false;
-                    }
-                }
-                let mut segments = path.segments();
-                let mut next_segment_text = || segments.next().and_then(|it| it.name_ref());
-                // `#[core::prelude::rust_2024::test]` or `#[std::prelude::rust_2024::test]`.
-                if next_segment_text().is_some_and(|it| matches!(&*it.text(), "core" | "std"))
-                    && next_segment_text().is_some_and(|it| it.text() == "prelude")
-                    && next_segment_text().is_some()
-                    && next_segment_text()
-                        .is_some_and(|it| matches!(&*it.text(), "test" | "bench" | "test_case"))
-                {
-                    return false;
+
+            let Some(item) = ast::Item::cast(ancestor) else {
+                return false;
+            };
+            // Optimization to skip the semantic check.
+            if item.attrs().all(|attr| {
+                attr.simple_name()
+                    .is_some_and(|attr| find_builtin_attr_idx(&Symbol::intern(&attr)).is_some())
+            }) {
+                return false;
+            }
+            self.with_ctx(|ctx| {
+                if ctx.item_to_macro_call(token.with_value(&item)).is_some() {
+                    return true;
                 }
-                true
+                let adt = match item {
+                    ast::Item::Struct(it) => it.into(),
+                    ast::Item::Enum(it) => it.into(),
+                    ast::Item::Union(it) => it.into(),
+                    _ => return false,
+                };
+                ctx.has_derives(token.with_value(&adt))
             })
         })
     }

src/tools/rust-analyzer/crates/ide-completion/src/context.rs

@@ -8,8 +8,8 @@ use std::{iter, ops::ControlFlow};
 
 use base_db::RootQueryDb as _;
 use hir::{
-    DisplayTarget, HasAttrs, Local, ModuleDef, ModuleSource, Name, PathResolution, ScopeDef,
-    Semantics, SemanticsScope, Symbol, Type, TypeInfo,
+    DisplayTarget, HasAttrs, InFile, Local, ModuleDef, ModuleSource, Name, PathResolution,
+    ScopeDef, Semantics, SemanticsScope, Symbol, Type, TypeInfo,
 };
 use ide_db::{
     FilePosition, FxHashMap, FxHashSet, RootDatabase, famous_defs::FamousDefs,
@@ -751,7 +751,7 @@ impl<'a> CompletionContext<'a> {
             original_offset,
         } = expand_and_analyze(
             &sema,
-            original_file.syntax().clone(),
+            InFile::new(editioned_file_id.into(), original_file.syntax().clone()),
             file_with_fake_ident.syntax().clone(),
             offset,
             &original_token,

src/tools/rust-analyzer/crates/ide-completion/src/context/analysis.rs

@@ -1,7 +1,7 @@
 //! Module responsible for analyzing the code surrounding the cursor for completion.
 use std::iter;
 
-use hir::{ExpandResult, Semantics, Type, TypeInfo, Variant};
+use hir::{ExpandResult, InFile, Semantics, Type, TypeInfo, Variant};
 use ide_db::{RootDatabase, active_parameter::ActiveParameter};
 use itertools::Either;
 use syntax::{
@@ -50,7 +50,7 @@ pub(super) struct AnalysisResult {
 
 pub(super) fn expand_and_analyze(
     sema: &Semantics<'_, RootDatabase>,
-    original_file: SyntaxNode,
+    original_file: InFile<SyntaxNode>,
     speculative_file: SyntaxNode,
     offset: TextSize,
     original_token: &SyntaxToken,
@@ -72,7 +72,7 @@ pub(super) fn expand_and_analyze(
         relative_offset,
     )
     .unwrap_or(ExpansionResult {
-        original_file,
+        original_file: original_file.value,
         speculative_file,
         original_offset: offset,
         speculative_offset: fake_ident_token.text_range().start(),
@@ -125,7 +125,7 @@ fn token_at_offset_ignore_whitespace(file: &SyntaxNode, offset: TextSize) -> Opt
 /// the best we can do.
 fn expand_maybe_stop(
     sema: &Semantics<'_, RootDatabase>,
-    original_file: SyntaxNode,
+    original_file: InFile<SyntaxNode>,
     speculative_file: SyntaxNode,
     original_offset: TextSize,
     fake_ident_token: SyntaxToken,
@@ -142,17 +142,16 @@ fn expand_maybe_stop(
         return result;
     }
 
-    // This needs to come after the recursive call, because our "inside macro" detection is subtly wrong
-    // with regard to attribute macros named `test` that are not std's test. So hopefully we will expand
-    // them successfully above and be able to analyze.
-    // Left biased since there may already be an identifier token there, and we appended to it.
-    if !sema.might_be_inside_macro_call(&fake_ident_token)
-        && token_at_offset_ignore_whitespace(&original_file, original_offset + relative_offset)
-            .is_some_and(|original_token| !sema.might_be_inside_macro_call(&original_token))
+    // We can't check whether the fake expansion is inside macro call, because that requires semantic info.
+    // But hopefully checking just the real one should be enough.
+    if token_at_offset_ignore_whitespace(&original_file.value, original_offset + relative_offset)
+        .is_some_and(|original_token| {
+            !sema.is_inside_macro_call(original_file.with_value(&original_token))
+        })
     {
         // Recursion base case.
         Some(ExpansionResult {
-            original_file,
+            original_file: original_file.value,
             speculative_file,
             original_offset,
             speculative_offset: fake_ident_token.text_range().start(),
@@ -166,7 +165,7 @@ fn expand_maybe_stop(
 
 fn expand(
     sema: &Semantics<'_, RootDatabase>,
-    original_file: SyntaxNode,
+    original_file: InFile<SyntaxNode>,
     speculative_file: SyntaxNode,
     original_offset: TextSize,
     fake_ident_token: SyntaxToken,
@@ -176,7 +175,7 @@ fn expand(
 
     let parent_item =
         |item: &ast::Item| item.syntax().ancestors().skip(1).find_map(ast::Item::cast);
-    let original_node = token_at_offset_ignore_whitespace(&original_file, original_offset)
+    let original_node = token_at_offset_ignore_whitespace(&original_file.value, original_offset)
         .and_then(|token| token.parent_ancestors().find_map(ast::Item::cast));
     let ancestor_items = iter::successors(
         Option::zip(
@@ -249,7 +248,7 @@ fn expand(
     }
 
     // No attributes have been expanded, so look for macro_call! token trees or derive token trees
-    let orig_tt = ancestors_at_offset(&original_file, original_offset)
+    let orig_tt = ancestors_at_offset(&original_file.value, original_offset)
         .map_while(Either::<ast::TokenTree, ast::Meta>::cast)
         .last()?;
     let spec_tt = ancestors_at_offset(&speculative_file, fake_ident_token.text_range().start())
@@ -292,7 +291,7 @@ fn expand(
                 fake_mapped_tokens.into_iter().min_by_key(|(_, rank)| *rank)
             {
                 return Some(ExpansionResult {
-                    original_file,
+                    original_file: original_file.value,
                     speculative_file,
                     original_offset,
                     speculative_offset: fake_ident_token.text_range().start(),
@@ -349,7 +348,7 @@ fn expand(
                                 }
                                 let result = expand_maybe_stop(
                                     sema,
-                                    actual_expansion.clone(),
+                                    InFile::new(file.into(), actual_expansion.clone()),
                                     fake_expansion.clone(),
                                     new_offset,
                                     fake_mapped_token,

src/tools/rust-analyzer/crates/ide-completion/src/tests/expression.rs

@@ -2111,6 +2111,56 @@ fn foo() {
     );
 }
 
+#[test]
+fn cfg_attr_attr_macro() {
+    check(
+        r#"
+//- proc_macros: identity
+#[cfg_attr(test, proc_macros::identity)]
+fn foo() {
+    $0
+}
+    "#,
+        expect![[r#"
+            fn foo()  fn()
+            md proc_macros
+            bt u32     u32
+            kw async
+            kw const
+            kw crate::
+            kw enum
+            kw extern
+            kw false
+            kw fn
+            kw for
+            kw if
+            kw if let
+            kw impl
+            kw impl for
+            kw let
+            kw letm
+            kw loop
+            kw match
+            kw mod
+            kw return
+            kw self::
+            kw static
+            kw struct
+            kw trait
+            kw true
+            kw type
+            kw union
+            kw unsafe
+            kw use
+            kw while
+            kw while let
+            sn macro_rules
+            sn pd
+            sn ppd
+        "#]],
+    );
+}
+
 #[test]
 fn escaped_label() {
     check(

src/tools/rust-analyzer/crates/ide-db/src/search.rs

@@ -524,6 +524,7 @@ impl<'a> FindUsages<'a> {
     fn find_nodes<'b>(
         sema: &'b Semantics<'_, RootDatabase>,
         name: &str,
+        file_id: EditionedFileId,
         node: &syntax::SyntaxNode,
         offset: TextSize,
     ) -> impl Iterator<Item = SyntaxNode> + 'b {
@@ -534,7 +535,7 @@ impl<'a> FindUsages<'a> {
             })
             .into_iter()
             .flat_map(move |token| {
-                if sema.might_be_inside_macro_call(&token) {
+                if sema.is_inside_macro_call(InFile::new(file_id.into(), &token)) {
                     sema.descend_into_macros_exact(token)
                 } else {
                     <_>::from([token])
@@ -654,11 +655,14 @@ impl<'a> FindUsages<'a> {
                     let tree = LazyCell::new(move || sema.parse(file_id).syntax().clone());
 
                     for offset in FindUsages::match_indices(&file_text, &finder, search_range) {
-                        let usages =
-                            FindUsages::find_nodes(sema, &current_to_process, &tree, offset)
-                                .filter(|it| {
-                                    matches!(it.kind(), SyntaxKind::NAME | SyntaxKind::NAME_REF)
-                                });
+                        let usages = FindUsages::find_nodes(
+                            sema,
+                            &current_to_process,
+                            file_id,
+                            &tree,
+                            offset,
+                        )
+                        .filter(|it| matches!(it.kind(), SyntaxKind::NAME | SyntaxKind::NAME_REF));
                         for usage in usages {
                             if let Some(alias) = usage.parent().and_then(|it| {
                                 let path = ast::PathSegment::cast(it)?.parent_path();
@@ -813,7 +817,7 @@ impl<'a> FindUsages<'a> {
                 let tree = LazyCell::new(move || this.sema.parse(file_id).syntax().clone());
 
                 for offset in FindUsages::match_indices(&file_text, finder, search_range) {
-                    let usages = FindUsages::find_nodes(this.sema, name, &tree, offset)
+                    let usages = FindUsages::find_nodes(this.sema, name, file_id, &tree, offset)
                         .filter_map(ast::NameRef::cast);
                     for usage in usages {
                         let found_usage = usage
@@ -970,8 +974,8 @@ impl<'a> FindUsages<'a> {
                     return;
                 }
 
-                for name in
-                    Self::find_nodes(sema, name, &tree, offset).filter_map(ast::NameLike::cast)
+                for name in Self::find_nodes(sema, name, file_id, &tree, offset)
+                    .filter_map(ast::NameLike::cast)
                 {
                     if match name {
                         ast::NameLike::NameRef(name_ref) => self.found_name_ref(&name_ref, sink),
@@ -985,8 +989,8 @@ impl<'a> FindUsages<'a> {
             // Search for occurrences of the `Self` referring to our type
             if let Some((self_ty, finder)) = &include_self_kw_refs {
                 for offset in Self::match_indices(&text, finder, search_range) {
-                    for name_ref in
-                        Self::find_nodes(sema, "Self", &tree, offset).filter_map(ast::NameRef::cast)
+                    for name_ref in Self::find_nodes(sema, "Self", file_id, &tree, offset)
+                        .filter_map(ast::NameRef::cast)
                     {
                         if self.found_self_ty_name_ref(self_ty, &name_ref, sink) {
                             return;
@@ -1010,7 +1014,7 @@ impl<'a> FindUsages<'a> {
                 let tree = LazyCell::new(move || sema.parse(file_id).syntax().clone());
 
                 for offset in Self::match_indices(&text, finder, search_range) {
-                    for name_ref in Self::find_nodes(sema, "super", &tree, offset)
+                    for name_ref in Self::find_nodes(sema, "super", file_id, &tree, offset)
                         .filter_map(ast::NameRef::cast)
                     {
                         if self.found_name_ref(&name_ref, sink) {
@@ -1020,7 +1024,7 @@ impl<'a> FindUsages<'a> {
                 }
                 if let Some(finder) = &is_crate_root {
                     for offset in Self::match_indices(&text, finder, search_range) {
-                        for name_ref in Self::find_nodes(sema, "crate", &tree, offset)
+                        for name_ref in Self::find_nodes(sema, "crate", file_id, &tree, offset)
                             .filter_map(ast::NameRef::cast)
                         {
                             if self.found_name_ref(&name_ref, sink) {
@@ -1064,8 +1068,8 @@ impl<'a> FindUsages<'a> {
                 let finder = &Finder::new("self");
 
                 for offset in Self::match_indices(&text, finder, search_range) {
-                    for name_ref in
-                        Self::find_nodes(sema, "self", &tree, offset).filter_map(ast::NameRef::cast)
+                    for name_ref in Self::find_nodes(sema, "self", file_id, &tree, offset)
+                        .filter_map(ast::NameRef::cast)
                     {
                         if self.found_self_module_name_ref(&name_ref, sink) {
                             return;

src/tools/rust-analyzer/crates/ide-ssr/src/lib.rs

@@ -287,7 +287,7 @@ impl<'db> MatchFinder<'db> {
                 if let Some(expanded) = self.sema.expand_macro_call(&macro_call) {
                     if let Some(tt) = macro_call.token_tree() {
                         self.output_debug_for_nodes_at_range(
-                            &expanded,
+                            &expanded.value,
                             range,
                             &Some(self.sema.original_range(tt.syntax())),
                             out,