Add a frozen string flag

Author: kddnewton

config.yml

@@ -369,6 +369,10 @@ flags:
         comment: "u - forces the UTF-8 encoding"
       - name: ONCE
         comment: "o - only interpolates values into the regular expression once"
+  - name: StringFlags
+    values:
+      - name: FROZEN
+        comment: "frozen by virtue of a frozen_string_literal comment"
 nodes:
   - name: AliasGlobalVariableNode
     fields:
@@ -2249,6 +2253,9 @@ nodes:
           ^^^^^^^^^^^
   - name: StringNode
     fields:
+      - name: flags
+        type: flags
+        kind: StringFlags
       - name: opening_loc
         type: location?
       - name: content_loc

src/yarp.c

@@ -4197,10 +4197,16 @@ yp_string_concat_node_create(yp_parser_t *parser, yp_node_t *left, yp_node_t *ri
 static yp_string_node_t *
 yp_string_node_create(yp_parser_t *parser, const yp_token_t *opening, const yp_token_t *content, const yp_token_t *closing) {
     yp_string_node_t *node = YP_ALLOC_NODE(parser, yp_string_node_t);
+    yp_node_flags_t flags = 0;
+
+    if (parser->frozen_string_literal) {
+        flags = YP_NODE_FLAG_STATIC_LITERAL | YP_STRING_FLAGS_FROZEN;
+    }
 
     *node = (yp_string_node_t) {
         {
             .type = YP_STRING_NODE,
+            .flags = flags,
             .location = {
                 .start = (opening->type == YP_TOKEN_NOT_PROVIDED ? content->start : opening->start),
                 .end = (closing->type == YP_TOKEN_NOT_PROVIDED ? content->end : closing->end)
@@ -4361,10 +4367,16 @@ yp_string_node_to_symbol_node(yp_parser_t *parser, yp_string_node_t *node, const
 static yp_string_node_t *
 yp_symbol_node_to_string_node(yp_parser_t *parser, yp_symbol_node_t *node) {
     yp_string_node_t *new_node = YP_ALLOC_NODE(parser, yp_string_node_t);
+    yp_node_flags_t flags = 0;
+
+    if (parser->frozen_string_literal) {
+        flags = YP_NODE_FLAG_STATIC_LITERAL | YP_STRING_FLAGS_FROZEN;
+    }
 
     *new_node = (yp_string_node_t) {
         {
             .type = YP_STRING_NODE,
+            .flags = flags,
             .location = node->base.location
         },
         .opening_loc = node->opening_loc,
@@ -5111,7 +5123,8 @@ parser_lex_frozen_string_literal_comment(yp_parser_t *parser) {
 
     while ((cursor = yp_memchr(cursor, 'f', (size_t) (cursor_limit - cursor), parser->encoding_changed, &parser->encoding)) != NULL) {
         if (memcmp(cursor, "frozen_string_literal", key_length) == 0) {
-            cursor += yp_strspn_inline_whitespace(cursor + key_length, end - (cursor + key_length));
+            cursor += key_length;
+            cursor += yp_strspn_inline_whitespace(cursor, end - cursor);
 
             if (*cursor == ':' || *cursor == '=') {
                 cursor++;
@@ -6174,7 +6187,7 @@ parser_lex(yp_parser_t *parser) {
                 /* fallthrough */
                 case '\r':
                 case '\n': {
-                    parser->semantic_token_seen = semantic_token_seen;
+                    parser->semantic_token_seen = semantic_token_seen & 0x1;
                     size_t eol_length = match_eol_at(parser, parser->current.end - 1);
 
                     if (eol_length) {

test/yarp/errors_test.rb

@@ -601,7 +601,7 @@ def test_do_not_allow_trailing_commas_in_lambda_parameters
     end
 
     def test_do_not_allow_multiple_codepoints_in_a_single_character_literal
-      expected = StringNode(Location(), Location(), nil, "\u0001\u0002")
+      expected = StringNode(0, Location(), Location(), nil, "\u0001\u0002")
 
       assert_errors expected, '?\u{0001 0002}', [
         ["Invalid Unicode escape sequence; multiple codepoints are not allowed in a character literal", 9..12]
@@ -615,15 +615,15 @@ def test_invalid_hex_escape
     end
 
     def test_do_not_allow_more_than_6_hexadecimal_digits_in_u_Unicode_character_notation
-      expected = StringNode(Location(), Location(), Location(), "\u0001")
+      expected = StringNode(0, Location(), Location(), Location(), "\u0001")
 
       assert_errors expected, '"\u{0000001}"', [
         ["Invalid Unicode escape sequence; maximum length is 6 digits", 4..11],
       ]
     end
 
     def test_do_not_allow_characters_other_than_0_9_a_f_and_A_F_in_u_Unicode_character_notation
-      expected = StringNode(Location(), Location(), Location(), "\u0000z}")
+      expected = StringNode(0, Location(), Location(), Location(), "\u0000z}")
 
       assert_errors expected, '"\u{000z}"', [
         ["Invalid Unicode escape sequence", 7..7],