fix: enable multi-byte char in block comment.

project-tsurugi · Jul 8, 2024 · e951283 · e951283
1 parent 28083c3
commit e951283
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 0 deletions.
diff --git a/src/mizugaki/parser/sql_scanner.ll b/src/mizugaki/parser/sql_scanner.ll
@@ -97,6 +97,7 @@ UTF8_CHAR {ASCII}|{UTF8_2}{U}|{UTF8_3}{U}{U}|{UTF8_4}{U}{U}{U}
         auto loc = exit_comment(true);
         return parser_type::make_UNCLOSED_BLOCK_COMMENT(loc(0, 2));
     }
+    {UTF8_CHAR} {}
     [\x00-\xff] {}
 }
 

diff --git a/test/mizugaki/parser/sql_parser_misc_test.cpp b/test/mizugaki/parser/sql_parser_misc_test.cpp
@@ -115,6 +115,16 @@ R"(
  */)");
 }
 
+TEST_F(sql_parser_misc_test, block_comment_multibyte) {
+    sql_parser parser;
+    auto result = parser("-", u8"/* \u3042 */");
+    ASSERT_TRUE(result) << diagnostics(result);
+
+    auto&& unit = **result;
+    ASSERT_EQ(unit.comments().size(), 1);
+    EXPECT_EQ(comment(unit, 0), u8"/* \u3042 */");
+}
+
 TEST_F(sql_parser_misc_test, block_comment_unclosed) {
     sql_parser parser;
     auto result = parser("-", "; /* unexpected EOF");