fix: enable multi-byte char in line comment.

project-tsurugi · Jul 8, 2024 · 37c6132 · 37c6132
1 parent e951283
commit 37c6132
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
@@ -115,6 +115,7 @@ UTF8_CHAR {ASCII}|{UTF8_2}{U}|{UTF8_3}{U}{U}|{UTF8_4}{U}{U}{U}
         driver.add_comment(exit_comment(true));
         return parser_type::make_END_OF_FILE(location(true));
     }
+    {UTF8_CHAR} {}
     . {}
 }
 

diff --git a/test/mizugaki/parser/sql_parser_misc_test.cpp b/test/mizugaki/parser/sql_parser_misc_test.cpp
@@ -76,6 +76,16 @@ TEST_F(sql_parser_misc_test, simple_comment_crlf) {
     EXPECT_EQ(comment(unit, 1), "--b");
 }
 
+TEST_F(sql_parser_misc_test, simple_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) {
     sql_parser parser;
     auto result = parser("-",