[FEAT MERGE]charset latin1

src/sql/parser/ob_fast_parser.cpp

@@ -191,6 +191,8 @@ inline int64_t ObFastParserBase::is_identifier_flags(const int64_t pos)
     idf_pos = is_utf8_char(pos);
   } else if (CHARSET_GBK == charset_type_ || CHARSET_GB18030 == charset_type_) {
     idf_pos = is_gbk_char(pos);
+  } else if (CHARSET_LATIN1 == charset_type_) {
+    idf_pos = is_latin1_char(pos);
   }
   return idf_pos;
 }
@@ -501,6 +503,15 @@ int ObFastParserBase::process_interval()
   return ret;
 }
 
+inline int64_t ObFastParserBase::is_latin1_char(const int64_t pos)
+{
+  int64_t idf_pos = -1;
+  if (is_latin1(raw_sql_.char_at(pos))) {
+    idf_pos = pos + 1;
+  }
+  return idf_pos;
+}
+
 // ({U_2}{U}|{U_3}{U}{U}|{U_4}{U}{U}{U}
 inline int64_t ObFastParserBase::is_utf8_char(const int64_t pos)
 {
@@ -1154,6 +1165,8 @@ inline int64_t ObFastParserBase::is_first_identifier_flags(const int64_t pos)
     idf_pos = is_utf8_char(pos);
   } else if (CHARSET_GBK == charset_type_ || CHARSET_GB18030 == charset_type_) {
     idf_pos = is_gbk_char(pos);
+  } else if (CHARSET_LATIN1 == charset_type_) {
+    idf_pos = is_latin1_char(pos);
   }
   return idf_pos;
 }

src/sql/parser/ob_fast_parser.h

@@ -318,6 +318,8 @@ protected:
 		return is_valid_char(ch) && USER_VAR_CHAR[static_cast<uint8_t>(ch)];
 	}
 	void reset_parser_node(ParseNode *node);
+	//{U}
+	int64_t is_latin1_char(const int64_t pos);
 	// ({U_2}{U}|{U_3}{U}{U}|{U_4}{U}{U}{U}
 	int64_t is_utf8_char(const int64_t pos);
 	// NOTES: No boundary check, the caller guarantees safety!!!
@@ -375,7 +377,11 @@ protected:
 		return is_valid_char(ch) &&
 		static_cast<uint8_t>(ch) >= 0x40 && static_cast<uint8_t>(ch) <= 0xfe;
 	}
-
+    inline bool is_latin1(char ch)
+	{
+		return is_valid_char(ch) &&
+		static_cast<uint8_t>(ch) >= 0x00 && static_cast<uint8_t>(ch) <= 0xFF;
+	}
 	// [0-9]{n}
 	inline bool is_n_continuous_digits(const char *str,
 									   const int64_t pos,

src/sql/parser/parse_malloc.cpp

@@ -232,7 +232,10 @@ char *parse_strdup_with_replace_multi_byte_char(const char *str, int *connection
         case 45/*CS_TYPE_UTF8MB4_GENERAL_CI*/:
         case 46/*CS_TYPE_UTF8MB4_BIN*/:
         case 63/*CS_TYPE_BINARY*/:
-        case 224/*CS_TYPE_UTF8MB4_UNICODE_CI*/: {
+        case 224/*CS_TYPE_UTF8MB4_UNICODE_CI*/:
+        //case 8/*CS_TYPE_LATIN1_SWEDISH_CI*/:
+        //case 47/*CS_TYPE_LATIN1_BIN*/:
+        {
           if (i + 2 < dup_len) {
             if (str[i] == (char)0xe3 && str[i+1] == (char)0x80 && str[i+2] == (char)0x80) {
               //utf8 multi byte space

src/sql/parser/sql_parser_mysql_mode.l

@@ -112,6 +112,7 @@ _UTF8                         { REPUT_TOKEN_NEG_SIGN(_UTF8); }
 _UTF8MB4                      { REPUT_TOKEN_NEG_SIGN(_UTF8MB4); }
 _GBK                          { REPUT_TOKEN_NEG_SIGN(_GBK); }
 _GB18030                      { REPUT_TOKEN_NEG_SIGN(_GB18030); }
+_LATIN1                       { REPUT_TOKEN_NEG_SIGN(_LATIN1); }
 _BINARY                       { REPUT_TOKEN_NEG_SIGN(_BINARY); }
 _UTF16                        { REPUT_TOKEN_NEG_SIGN(_UTF16); }
 NOT {

src/sql/parser/sql_parser_mysql_mode.y

@@ -186,7 +186,7 @@ APPEND NO_GATHER_OPTIMIZER_STATISTICS GATHER_OPTIMIZER_STATISTICS
 NEG_SIGN
 
 %token /*can not be relation name*/
-_BINARY _UTF8 _UTF8MB4 _GBK _UTF16 _GB18030 CNNOP
+_BINARY _UTF8 _UTF8MB4 _GBK _UTF16 _GB18030 _LATIN1 CNNOP
 SELECT_HINT_BEGIN UPDATE_HINT_BEGIN DELETE_HINT_BEGIN INSERT_HINT_BEGIN REPLACE_HINT_BEGIN HINT_HINT_BEGIN HINT_END
 LOAD_DATA_HINT_BEGIN CREATE_HINT_BEGIN
 END_P SET_VAR DELIMITER
@@ -952,6 +952,15 @@ _UTF8
     YYABORT_NO_MEMORY;
   }
 }
+| _LATIN1
+{
+  malloc_terminal_node($$, result->malloc_pool_, T_CHARSET);
+  $$->str_value_ = parse_strdup("latin1", result->malloc_pool_, &($$->str_len_));
+  if (OB_UNLIKELY(NULL == $$->str_value_)) {
+    yyerror(NULL, result, "No more space for mallocing string");
+    YYABORT_NO_MEMORY;
+  }
+}
 | _GB18030
 {
   malloc_terminal_node($$, result->malloc_pool_, T_CHARSET);