rust: track inline module scopes for module file resolution

The set of inline modules is required to find the expected location of a
module file. Track this information with an RAII object
(`InlineModuleStackScope`) and pass it down to any out-of-line modules
so that, when requested, the set of inline modules can be added to the
search path.

Signed-off-by: Ben Boeckel <[email protected]>

Author: mathstuf

gcc/rust/ast/rust-ast-full-test.cc

@@ -4057,6 +4057,13 @@ Module::process_file_path ()
     current_directory_name
       = including_fname.substr (0, dir_slash_pos) + file_separator;
 
+  // Handle inline module declarations adding path components.
+  for (auto const &name : module_scope)
+    {
+      current_directory_name.append (name);
+      current_directory_name.append (file_separator);
+    }
+
   auto path_string = filename_from_path_attribute (get_outer_attrs ());
   if (!path_string.empty ())
     {
@@ -4070,12 +4077,13 @@ Module::process_file_path ()
   // current file is titled `mod.rs`.
 
   // First, we search for <directory>/<module_name>.rs
-  bool file_mod_found
-    = file_exists (current_directory_name + expected_file_path);
+  std::string file_mod_path = current_directory_name + expected_file_path;
+  bool file_mod_found = file_exists (file_mod_path);
 
   // Then, search for <directory>/<module_name>/mod.rs
-  current_directory_name += module_name + file_separator;
-  bool dir_mod_found = file_exists (current_directory_name + expected_dir_path);
+  std::string dir_mod_path
+    = current_directory_name + module_name + file_separator + expected_dir_path;
+  bool dir_mod_found = file_exists (dir_mod_path);
 
   bool multiple_candidates_found = file_mod_found && dir_mod_found;
   bool no_candidates_found = !file_mod_found && !dir_mod_found;
@@ -4093,8 +4101,7 @@ Module::process_file_path ()
   if (no_candidates_found || multiple_candidates_found)
     return;
 
-  module_file = file_mod_found ? expected_file_path
-			       : current_directory_name + expected_dir_path;
+  module_file = std::move (file_mod_found ? file_mod_path : dir_mod_path);
 }
 
 void

gcc/rust/ast/rust-item.h

@@ -991,6 +991,8 @@ class Module : public VisItem
   std::vector<Attribute> inner_attrs;
   // bool has_items;
   std::vector<std::unique_ptr<Item>> items;
+  // Names of including inline modules (immediate parent is last in the list)
+  std::vector<std::string> module_scope;
 
   // Filename the module refers to. Empty string on LOADED modules or if an
   // error occured when dealing with UNLOADED modules
@@ -1013,11 +1015,12 @@ class Module : public VisItem
   // Unloaded module constructor
   Module (Identifier module_name, Visibility visibility,
 	  std::vector<Attribute> outer_attrs, Location locus,
-	  std::string outer_filename)
+	  std::string outer_filename, std::vector<std::string> module_scope)
     : VisItem (std::move (visibility), std::move (outer_attrs)),
       module_name (module_name), locus (locus), kind (ModuleKind::UNLOADED),
       outer_filename (outer_filename), inner_attrs (std::vector<Attribute> ()),
-      items (std::vector<std::unique_ptr<Item>> ())
+      items (std::vector<std::unique_ptr<Item>> ()),
+      module_scope (std::move (module_scope))
   {}
 
   // Loaded module constructor, with items
@@ -1035,7 +1038,8 @@ class Module : public VisItem
   // Copy constructor with vector clone
   Module (Module const &other)
     : VisItem (other), module_name (other.module_name), locus (other.locus),
-      kind (other.kind), inner_attrs (other.inner_attrs)
+      kind (other.kind), inner_attrs (other.inner_attrs),
+      module_scope (other.module_scope)
   {
     // We need to check whether we are copying a loaded module or an unloaded
     // one. In the second case, clear the `items` vector.
@@ -1054,6 +1058,7 @@ class Module : public VisItem
     locus = other.locus;
     kind = other.kind;
     inner_attrs = other.inner_attrs;
+    module_scope = other.module_scope;
 
     // Likewise, we need to clear the `items` vector in case the other module is
     // unloaded

gcc/rust/parse/rust-parse-impl.h

@@ -2078,6 +2078,78 @@ Parser<ManagedTokenSource>::parse_visibility ()
     }
 }
 
+static std::string
+extract_module_path (const AST::AttrVec &inner_attrs, const AST::AttrVec &outer_attrs, const std::string &name)
+{
+  AST::Attribute path_attr = AST::Attribute::create_empty ();
+  for (const auto &attr : inner_attrs)
+    {
+      if (attr.get_path ().as_string () == "path")
+	{
+	  path_attr = attr;
+	  break;
+	}
+    }
+
+  // Here, we found a path attribute, but it has no associated string. This is
+  // invalid
+  if (!path_attr.is_empty () && !path_attr.has_attr_input ())
+    {
+      rust_error_at (
+	path_attr.get_locus (),
+	// Split the format string so that -Wformat-diag does not complain...
+	"path attributes must contain a filename: '%s'", "#![path = \"file\"]");
+      return name;
+    }
+
+  for (const auto &attr : outer_attrs)
+    {
+      if (attr.get_path ().as_string () == "path")
+	{
+	  path_attr = attr;
+	  break;
+	}
+    }
+
+  // We didn't find a path attribute. This is not an error, there simply isn't
+  // one present
+  if (path_attr.is_empty ())
+    return name;
+
+  // Here, we found a path attribute, but it has no associated string. This is
+  // invalid
+  if (!path_attr.has_attr_input ())
+    {
+      rust_error_at (
+	path_attr.get_locus (),
+	// Split the format string so that -Wformat-diag does not complain...
+	"path attributes must contain a filename: '%s'", "#[path = \"file\"]");
+      return name;
+    }
+
+  auto path_value = path_attr.get_attr_input ().as_string ();
+
+  // At this point, the 'path' is of the following format: '= "<file.rs>"'
+  // We need to remove the equal sign and only keep the actual filename.
+  // In order to do this, we can simply go through the string until we find
+  // a character that is not an equal sign or whitespace
+  auto filename_begin = path_value.find_first_not_of ("=\t ");
+
+  auto path = path_value.substr (filename_begin);
+
+  // On windows, the path might mix '/' and '\' separators. Replace the
+  // UNIX-like separators by MSDOS separators to make sure the path will resolve
+  // properly.
+  //
+  // Source: rustc compiler
+  // (https://github.com/rust-lang/rust/blob/9863bf51a52b8e61bcad312f81b5193d53099f9f/compiler/rustc_expand/src/module.rs#L174)
+#if defined(HAVE_DOS_BASED_FILE_SYSTEM)
+  path.replace ('/', '\\');
+#endif /* HAVE_DOS_BASED_FILE_SYSTEM */
+
+  return path;
+}
+
 // Parses a module - either a bodied module or a module defined in another file.
 template <typename ManagedTokenSource>
 std::unique_ptr<AST::Module>
@@ -2104,14 +2176,18 @@ Parser<ManagedTokenSource>::parse_module (AST::Visibility vis,
       // Construct an external module
       return std::unique_ptr<AST::Module> (
 	new AST::Module (std::move (name), std::move (vis),
-			 std::move (outer_attrs), locus,
-			 lexer.get_filename ()));
+			 std::move (outer_attrs), locus, lexer.get_filename (),
+			 inline_module_stack));
       case LEFT_CURLY: {
 	lexer.skip_token ();
 
 	// parse inner attributes
 	AST::AttrVec inner_attrs = parse_inner_attributes ();
 
+	std::string module_path_name = extract_module_path (inner_attrs, outer_attrs, name);
+	rust_debug ("pushing inline module path for %s: %s", name.c_str (), module_path_name.c_str ());
+	InlineModuleStackScope scope (*this, std::move (module_path_name));
+
 	// parse items
 	std::vector<std::unique_ptr<AST::Item>> items;
 	const_TokenPtr tok = lexer.peek_token ();

gcc/rust/parse/rust-parse.h

@@ -622,6 +622,21 @@ template <typename ManagedTokenSource> class Parser
   ManagedTokenSource lexer;
   // The error list.
   std::vector<Error> error_table;
+  // The names of inline modules while parsing.
+  std::vector<std::string> inline_module_stack;
+
+  class InlineModuleStackScope
+  {
+  private:
+    Parser &parser;
+
+  public:
+    InlineModuleStackScope (Parser &parser, std::string name) : parser (parser)
+    {
+      parser.inline_module_stack.emplace_back (std::move (name));
+    }
+    ~InlineModuleStackScope () { parser.inline_module_stack.pop_back (); }
+  };
 };
 } // namespace Rust
 

gcc/testsuite/rust/compile/missing_middle/explicit.not.rs

@@ -0,0 +1 @@
+mod other;

gcc/testsuite/rust/compile/missing_middle/other.rs

@@ -0,0 +1,3 @@
+pub fn f() -> u32 {
+    2
+}

gcc/testsuite/rust/compile/missing_middle/sub/mod.rs

@@ -0,0 +1,3 @@
+pub fn f() -> u32 {
+    1
+}

gcc/testsuite/rust/compile/mod_missing_middle.rs

@@ -0,0 +1,30 @@
+// { dg-additional-options "-w" }
+
+mod missing_middle {
+    mod sub;
+
+    #[path = "explicit.not.rs"]
+    mod explicit;
+}
+
+#[path = "missing_middle"]
+mod with_outer_path_attr {
+    #[path = "outer_path.rs"]
+    mod inner;
+}
+
+// FIXME: the parser doesn't accept inner attributes for inline modules.
+// mod with_inner_path_attr {
+//     #![path = "missing_middle"]
+
+//     #[path = "explicit.not.rs"]
+//     mod inner;
+// }
+
+// #![path = "missing_middle"]
+// mod with_both_path_attr {
+//     #![path = "this_is_ignored"]
+
+//     #[path = "explicit.not.rs"]
+//     mod inner;
+// }