Merge #4178

4178: Validate the location of `crate` in paths r=matklad a=djrenren

**This solution does not fully handle `use` statements. See below**

This pull requests implements simple validation of usages of the `crate` keyword in `Path`s. Specifically it validates that:

- If a `PathSegment` is starts with the `crate` keyword, it is also the first segment of the `Path`
- All other usages of `crate` in `Path`s are considered errors.

This aligns with `rustc`'s rules. Unlike rustc this implementation does not issue a special error message in the case of `::crate` but it does catch the error.

Furthermore, this change does not cover all error cases. Specifically the following is not caught:

```rust
use foo::{crate}
```

This is because this check is context sensitive. From an AST perspective, `crate` is the root of the `Path`. Only by inspecting the full `UseItem` do we see that it is not in fact the root. This problem becomes worse because `UseTree`s are allowed to be arbitrarily nested:

```rust
use {crate, {{crate, foo::{crate}}}
```

So this is a hard problem to solve without essentially a breadth-first search. In a traditional compiler, I'd say this error is most easily found during the AST -> HIR conversion pass but within rust-analyzer I'm not sure where it belongs.  

Under the implementation in this PR, such errors are ignored so we're *more correct* just not *entirely correct*. 

Co-authored-by: John Renner <[email protected]>

4 files changed, 120 insertions, 0 deletions

diff --git a/crates/ra_syntax/src/ast/generated/nodes.rs b/crates/ra_syntax/src/ast/generated/nodes.rs
index 2cb3ad011..3b5e05af9 100644
--- a/crates/ra_syntax/src/ast/generated/nodes.rs
+++ b/crates/ra_syntax/src/ast/generated/nodes.rs
@@ -1249,6 +1249,7 @@ pub struct PathSegment {
 }
 impl PathSegment {
     pub fn coloncolon_token(&self) -> Option<SyntaxToken> { support::token(&self.syntax, T![::]) }
+    pub fn crate_token(&self) -> Option<SyntaxToken> { support::token(&self.syntax, T![crate]) }
     pub fn l_angle_token(&self) -> Option<SyntaxToken> { support::token(&self.syntax, T![<]) }
     pub fn name_ref(&self) -> Option<NameRef> { support::child(&self.syntax) }
     pub fn type_arg_list(&self) -> Option<TypeArgList> { support::child(&self.syntax) }
diff --git a/crates/ra_syntax/src/validation.rs b/crates/ra_syntax/src/validation.rs
index 5e93895ec..a30bc97bb 100644
--- a/crates/ra_syntax/src/validation.rs
+++ b/crates/ra_syntax/src/validation.rs
@@ -96,6 +96,7 @@ pub(crate) fn validate(root: &SyntaxNode) -> Vec<SyntaxError> {
                 ast::RecordField(it) => validate_numeric_name(it.name_ref(), &mut errors),
                 ast::Visibility(it) => validate_visibility(it, &mut errors),
                 ast::RangeExpr(it) => validate_range_expr(it, &mut errors),
+                ast::PathSegment(it) => validate_crate_keyword_in_path_segment(it, &mut errors),
                 _ => (),
             }
         }
@@ -222,3 +223,41 @@ fn validate_range_expr(expr: ast::RangeExpr, errors: &mut Vec<SyntaxError>) {
         ));
     }
 }
+
+fn validate_crate_keyword_in_path_segment(
+    segment: ast::PathSegment,
+    errors: &mut Vec<SyntaxError>,
+) {
+    const ERR_MSG: &str = "The `crate` keyword is only allowed as the first segment of a path";
+
+    let crate_token = match segment.crate_token() {
+        None => return,
+        Some(it) => it,
+    };
+
+    // Disallow both ::crate and foo::crate
+    let path = segment.parent_path();
+    if segment.coloncolon_token().is_some() || path.qualifier().is_some() {
+        errors.push(SyntaxError::new(ERR_MSG, crate_token.text_range()));
+        return;
+    }
+
+    // We now know that the path variable describes a complete path.
+    // For expressions and types, validation is complete, but we still have
+    // to handle UseItems like this:
+    //      use foo:{crate};
+    // so we crawl upwards looking for any preceding paths on `UseTree`s
+    for node in path.syntax().ancestors().skip(1) {
+        match_ast! {
+            match node {
+                ast::UseTree(it) => if let Some(tree_path) = it.path() {
+                    if tree_path != path {
+                        errors.push(SyntaxError::new(ERR_MSG, crate_token.text_range()));
+                    }
+                },
+                ast::UseTreeList(_it) => continue,
+                _ => return,
+            }
+        };
+    }
+}
diff --git a/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rast b/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rast
new file mode 100644
index 000000000..8306f7361
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rast
@@ -0,0 +1,76 @@
+[email protected]
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected] "::"
+          [email protected] "crate"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected] "{"
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected] "crate"
+        [email protected] ","
+        [email protected] " "
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected]
+                [email protected] "foo"
+          [email protected] "::"
+          [email protected]
+            [email protected] "{"
+            [email protected]
+              [email protected]
+                [email protected]
+                  [email protected] "crate"
+            [email protected] "}"
+        [email protected] "}"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected] "hello"
+        [email protected] "::"
+        [email protected]
+          [email protected] "crate"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected]
+                [email protected] "hello"
+          [email protected] "::"
+          [email protected]
+            [email protected] "crate"
+        [email protected] "::"
+        [email protected]
+          [email protected]
+            [email protected] "there"
+    [email protected] ";"
+  [email protected] "\n"
+error 6..11: The `crate` keyword is only allowed as the first segment of a path
+error 31..36: The `crate` keyword is only allowed as the first segment of a path
+error 51..56: The `crate` keyword is only allowed as the first segment of a path
+error 69..74: The `crate` keyword is only allowed as the first segment of a path
diff --git a/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rs b/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rs
new file mode 100644
index 000000000..bead4c0b6
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0040_illegal_crate_kw_location.rs
@@ -0,0 +1,4 @@
+use ::crate;
+use {crate, foo::{crate}};
+use hello::crate;
+use hello::crate::there;