Merge #4246

4246: Validate uses of self and super r=matklad a=djrenren

This change follows on the validation of the `crate` keyword in paths. It verifies the following things:

`super`:
 - May only be preceded by other `super` segments
 - If in a `UseItem` then all semantically preceding paths also consist only of `super`

`self`
 - May only be the start of a path


Just a note, a couple times while working on this I found myself really wanting a Visitor of some sort so that I could traverse descendants while skipping sub-trees that are unimportant. Iterators don't really work for this, so as you can see I reached for recursion. Considering paths are generally small a fancy debounced visitor probably isn't important but figured I'd say something in case we had something like this lying around and I wasn't using it.

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

8 files changed, 177 insertions, 74 deletions

diff --git a/crates/ra_syntax/src/ast/generated/nodes.rs b/crates/ra_syntax/src/ast/generated/nodes.rs
index 81260680f..3f16592b6 100644
--- a/crates/ra_syntax/src/ast/generated/nodes.rs
+++ b/crates/ra_syntax/src/ast/generated/nodes.rs
@@ -1241,6 +1241,8 @@ 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 self_token(&self) -> Option<SyntaxToken> { support::token(&self.syntax, T![self]) }
+    pub fn super_token(&self) -> Option<SyntaxToken> { support::token(&self.syntax, T![super]) }
     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 f0b3dec63..e075cd801 100644
--- a/crates/ra_syntax/src/validation.rs
+++ b/crates/ra_syntax/src/validation.rs
@@ -96,7 +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),
+                ast::PathSegment(it) => validate_path_keywords(it, &mut errors),
                 _ => (),
             }
         }
@@ -224,59 +224,82 @@ 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";
+fn validate_path_keywords(segment: ast::PathSegment, errors: &mut Vec<SyntaxError>) {
+    use ast::PathSegmentKind;
 
-    let crate_token = match segment.crate_token() {
-        None => return,
-        Some(it) => it,
-    };
+    let path = segment.parent_path();
+    let is_path_start = segment.coloncolon_token().is_none() && path.qualifier().is_none();
+
+    if let Some(token) = segment.self_token() {
+        if !is_path_start {
+            errors.push(SyntaxError::new(
+                "The `self` keyword is only allowed as the first segment of a path",
+                token.text_range(),
+            ));
+        }
+    } else if let Some(token) = segment.crate_token() {
+        if !is_path_start || use_prefix(path).is_some() {
+            errors.push(SyntaxError::new(
+                "The `crate` keyword is only allowed as the first segment of a path",
+                token.text_range(),
+            ));
+        }
+    } else if let Some(token) = segment.super_token() {
+        if !all_supers(&path) {
+            errors.push(SyntaxError::new(
+                "The `super` keyword may only be preceded by other `super`s",
+                token.text_range(),
+            ));
+            return;
+        }
 
-    // Disallow both ::crate and foo::crate
-    let mut 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;
+        let mut curr_path = path;
+        while let Some(prefix) = use_prefix(curr_path) {
+            if !all_supers(&prefix) {
+                errors.push(SyntaxError::new(
+                    "The `super` keyword may only be preceded by other `super`s",
+                    token.text_range(),
+                ));
+                return;
+            }
+            curr_path = prefix;
+        }
     }
 
-    // For expressions and types, validation is complete, but we still have
-    // to handle invalid UseItems like this:
-    //
-    //      use foo:{crate::bar::baz};
-    //
-    // To handle this we must inspect the parent `UseItem`s and `UseTree`s
-    // but right now we're looking deep inside the nested `Path` nodes because
-    // `Path`s are left-associative:
-    //
-    //   ((crate)::bar)::baz)
-    //       ^ current value of path
-    //
-    // So we need to climb to the top
-    while let Some(parent) = path.parent_path() {
-        path = parent;
+    fn use_prefix(mut path: ast::Path) -> Option<ast::Path> {
+        for node in path.syntax().ancestors().skip(1) {
+            match_ast! {
+                match node {
+                    ast::UseTree(it) => if let Some(tree_path) = it.path() {
+                        // Even a top-level path exists within a `UseTree` so we must explicitly
+                        // allow our path but disallow anything else
+                        if tree_path != path {
+                            return Some(tree_path);
+                        }
+                    },
+                    ast::UseTreeList(_it) => continue,
+                    ast::Path(parent) => path = parent,
+                    _ => return None,
+                }
+            };
+        }
+        return None;
     }
 
-    // Now that we've found the whole path we need to see if there's a prefix
-    // somewhere in the UseTree hierarchy. This check is arbitrarily deep
-    // because rust allows arbitrary nesting like so:
-    //
-    // use {foo::{{{{crate::bar::baz}}}}};
-    for node in path.syntax().ancestors().skip(1) {
-        match_ast! {
-            match node {
-                ast::UseTree(it) => if let Some(tree_path) = it.path() {
-                    // Even a top-level path exists within a `UseTree` so we must explicitly
-                    // allow our path but disallow anything else
-                    if tree_path != path {
-                        errors.push(SyntaxError::new(ERR_MSG, crate_token.text_range()));
-                    }
-                },
-                ast::UseTreeList(_it) => continue,
-                _ => return,
-            }
+    fn all_supers(path: &ast::Path) -> bool {
+        let segment = match path.segment() {
+            Some(it) => it,
+            None => return false,
         };
+
+        if segment.kind() != Some(PathSegmentKind::SuperKw) {
+            return false;
+        }
+
+        if let Some(ref subpath) = path.qualifier() {
+            return all_supers(subpath);
+        }
+
+        return true;
     }
 }
diff --git a/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rast b/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rast
new file mode 100644
index 000000000..d0360c467
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rast
@@ -0,0 +1,70 @@
+[email protected]
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected] "::"
+          [email protected] "super"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected] "a"
+        [email protected] "::"
+        [email protected]
+          [email protected] "super"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected] "super"
+          [email protected] "::"
+          [email protected]
+            [email protected]
+              [email protected] "a"
+        [email protected] "::"
+        [email protected]
+          [email protected] "super"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected] "a"
+      [email protected] "::"
+      [email protected]
+        [email protected] "{"
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected]
+                [email protected] "super"
+            [email protected] "::"
+            [email protected]
+              [email protected]
+                [email protected] "b"
+        [email protected] "}"
+    [email protected] ";"
+  [email protected] "\n"
+error 6..11: The `super` keyword may only be preceded by other `super`s
+error 20..25: The `super` keyword may only be preceded by other `super`s
+error 41..46: The `super` keyword may only be preceded by other `super`s
+error 56..61: The `super` keyword may only be preceded by other `super`s
diff --git a/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rs b/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rs
new file mode 100644
index 000000000..bd4d58042
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0041_illegal_super_keyword_location.rs
@@ -0,0 +1,4 @@
+use ::super;
+use a::super;
+use super::a::super;
+use a::{super::b};
diff --git a/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rast b/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rast
new file mode 100644
index 000000000..4f382b06c
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rast
@@ -0,0 +1,27 @@
+[email protected]
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected] "::"
+          [email protected] "self"
+    [email protected] ";"
+  [email protected] "\n"
+  [email protected]
+    [email protected] "use"
+    [email protected] " "
+    [email protected]
+      [email protected]
+        [email protected]
+          [email protected]
+            [email protected]
+              [email protected] "a"
+        [email protected] "::"
+        [email protected]
+          [email protected] "self"
+    [email protected] ";"
+  [email protected] "\n"
+error 6..10: The `self` keyword is only allowed as the first segment of a path
+error 19..23: The `self` keyword is only allowed as the first segment of a path
diff --git a/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rs b/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rs
new file mode 100644
index 000000000..b9e1d7d8b
--- /dev/null
+++ b/crates/ra_syntax/test_data/parser/err/0042_illegal_self_keyword_location.rs
@@ -0,0 +1,2 @@
+use ::self;
+use a::self;
diff --git a/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rast b/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rast
index a5a90df7b..05d9c05ad 100644
--- a/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rast
+++ b/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rast
@@ -1,4 +1,4 @@
-SOURCE_FILE@0..65
+SOURCE_FILE@0..38
   [email protected]
     [email protected] "use"
     [email protected] " "
@@ -31,27 +31,3 @@ [email protected]
             [email protected] "bar"
     [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] "::"
-                [email protected] "self"
-            [email protected] "::"
-            [email protected]
-              [email protected]
-                [email protected] "a"
-          [email protected] "::"
-          [email protected]
-            [email protected] "super"
-        [email protected] "::"
-        [email protected]
-          [email protected]
-            [email protected] "bar"
-    [email protected] ";"
-  [email protected] "\n"
diff --git a/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rs b/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rs
index faf6a42c7..9d9eb9917 100644
--- a/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rs
+++ b/crates/ra_syntax/test_data/parser/ok/0013_use_path_self_super.rs
@@ -1,3 +1,2 @@
 use self::foo;
 use super::super::bar;
-use ::self::a::super::bar;