rbs 4.0.3 → 4.1.0

data/src/lexer.re

@@ -58,7 +58,7 @@ rbs_token_t rbs_lexer_next_token(rbs_lexer_t *lexer) {
       "%a|" [^|\x00]* "|"  { return rbs_next_token(lexer, tANNOTATION); }
       "%a<" [^>\x00]* ">"  { return rbs_next_token(lexer, tANNOTATION); }
 
-      "#" (. \ [\x00])*    {
+      "#" (. \ [\x00\uFFFD])*    {
         return rbs_next_token(
           lexer,
           lexer->first_token_of_line ? tLINECOMMENT : tCOMMENT
@@ -83,6 +83,7 @@ rbs_token_t rbs_lexer_next_token(rbs_lexer_t *lexer) {
       "interface"     { return rbs_next_token(lexer, kINTERFACE); }
       "module"        { return rbs_next_token(lexer, kMODULE); }
       "module-alias"  { return rbs_next_token(lexer, kMODULEALIAS); }
+      "module-self"   { return rbs_next_token(lexer, kMODULESELF); }
       "nil"           { return rbs_next_token(lexer, kNIL); }
       "out"           { return rbs_next_token(lexer, kOUT); }
       "prepend"       { return rbs_next_token(lexer, kPREPEND); }

data/src/lexstate.c

@@ -68,6 +68,7 @@ static const char *RBS_TOKENTYPE_NAMES[] = {
     "kATRBS",        /* @rbs */
     "kSKIP",         /* skip */
     "kRETURN",       /* return */
+    "kMODULESELF",   /* module-self */
 
     "tLIDENT",  /* Identifiers starting with lower case */
     "tUIDENT",  /* Identifiers starting with upper case */
@@ -133,7 +134,15 @@ bool rbs_next_char(rbs_lexer_t *lexer, unsigned int *codepoint, size_t *byte_len
 
     *byte_len = lexer->encoding->char_width((const uint8_t *) start, (ptrdiff_t) (lexer->string.end - start));
 
-    if (*byte_len == 1) {
+    if (*byte_len == 0) {
+        // Invalid byte under the active encoding. Map it to a sentinel code
+        // point (U+FFFD) and advance one byte so the lexer always makes
+        // progress. Token rules that scan until a delimiter exclude this
+        // sentinel, so an invalid byte surfaces as an ErrorToken instead of
+        // being silently swallowed.
+        *byte_len = 1;
+        *codepoint = 0xFFFD;
+    } else if (*byte_len == 1) {
         *codepoint = (unsigned int) *start;
     } else {
         *codepoint = 12523; // Dummy data for "ル" from "ルビー" (Ruby) in Unicode

data/src/parser.c

@@ -3324,11 +3324,7 @@ bool rbs_parse_signature(rbs_parser_t *parser, rbs_signature_t **signature) {
         rbs_ast_directives_use_t *use_node;
         CHECK_PARSE(parse_use_directive(parser, &use_node));
 
-        if (use_node == NULL) {
-            rbs_node_list_append(dirs, NULL);
-        } else {
-            rbs_node_list_append(dirs, (rbs_node_t *) use_node);
-        }
+        rbs_node_list_append(dirs, (rbs_node_t *) use_node);
     }
 
     while (parser->next_token.type != pEOF) {
@@ -4052,6 +4048,60 @@ static bool parse_inline_leading_annotation(rbs_parser_t *parser, rbs_ast_ruby_a
             );
             return true;
         }
+        case kMODULESELF: {
+            rbs_parser_advance(parser);
+            rbs_range_t keyword_range = parser->current_token.range;
+
+            ADVANCE_ASSERT(parser, pCOLON);
+            rbs_range_t colon_range = parser->current_token.range;
+
+            rbs_parser_advance(parser);
+
+            rbs_range_t name_range;
+            rbs_type_name_t *type_name = NULL;
+            if (!parse_type_name(parser, (TypeNameKind) (CLASS_NAME | INTERFACE_NAME), &name_range, &type_name)) {
+                return false;
+            }
+
+            rbs_node_list_t *args = rbs_node_list_new(ALLOCATOR());
+            rbs_location_range open_bracket_loc = RBS_LOCATION_NULL_RANGE;
+            rbs_location_range close_bracket_loc = RBS_LOCATION_NULL_RANGE;
+            rbs_location_range_list_t *args_comma_locations = rbs_location_range_list_new(ALLOCATOR());
+            if (parser->next_token.type == pLBRACKET) {
+                rbs_parser_advance(parser);
+                open_bracket_loc = RBS_RANGE_LEX2AST(parser->current_token.range);
+                CHECK_PARSE(parse_type_list_with_commas(parser, pRBRACKET, args, args_comma_locations, true, false, false));
+                rbs_parser_advance(parser);
+                close_bracket_loc = RBS_RANGE_LEX2AST(parser->current_token.range);
+            }
+
+            rbs_location_range comment_loc = RBS_LOCATION_NULL_RANGE;
+            if (!parse_inline_comment(parser, &comment_loc)) {
+                return false;
+            }
+
+            rbs_range_t full_range = {
+                .start = rbs_range.start,
+                .end = parser->current_token.range.end
+            };
+
+            rbs_location_range full_loc = RBS_RANGE_LEX2AST(full_range);
+
+            *annotation = (rbs_ast_ruby_annotations_t *) rbs_ast_ruby_annotations_module_self_annotation_new(
+                ALLOCATOR(),
+                full_loc,
+                RBS_RANGE_LEX2AST(rbs_range),
+                RBS_RANGE_LEX2AST(keyword_range),
+                RBS_RANGE_LEX2AST(colon_range),
+                type_name,
+                args,
+                open_bracket_loc,
+                close_bracket_loc,
+                args_comma_locations,
+                comment_loc
+            );
+            return true;
+        }
         default: {
             rbs_parser_set_error(parser, parser->next_token, true, "unexpected token for @rbs annotation");
             return false;