MDEV-19239 ERROR 1300 (HY000): Invalid utf8 character string in 10.3.13-MariaDB

A sequence of <digits>e<mbhead><mbtail>, e.g.: SELECT 123eXYzzz FROM t1; was not scanned correctly (where XY is a multi-byte character). The multi-byte head byte X was appended to 123e separately from the multi-byte tail byte Y, so a pointer to "Yzzz" was passed into scan_ident_start(), which failed on a bad multi-byte sequence. After this change, scan_ident_start() gets a pointer to "XYzzz", so it correctly sees the whole multi-byte character.
2019-04-21 12:07:30 +04:00 · 2019-04-21 12:07:30 +04:00 · 765ae6e821
commit 765ae6e821
parent f4019f5b35
3 changed files with 162 additions and 1 deletions
--- a/mysql-test/main/ctype_utf8.result
+++ b/mysql-test/main/ctype_utf8.result
@ -11300,5 +11300,73 @@ t1	CREATE TABLE `t1` (
 DROP TABLE t1;
 SET sql_mode=DEFAULT;
 #
+# MDEV-19239 ERROR 1300 (HY000): Invalid utf8 character string in 10.3.13-MariaDB
+#
+SET NAMES utf8;
+SELECT
+x.消息ID,
+x.消息TITLE,
+x.消息类型,
+x.发送时间,
+x.阅读时间,x.老师ID,
+IF(x.四天内最近一次登录时间='2100-01-01 00:00:00','',x.四天内最近一次登录时间) 四天内最近一次登录时间
+FROM (
+SELECT
+msg.*,
+CASE
+WHEN login.login_time BETWEEN msg.发送时间 AND DATE_ADD(msg.发送时间,INTERVAL 4 DAY)
+THEN login.login_time
+WHEN (login.login_time NOT BETWEEN msg.发送时间 AND DATE_ADD(msg.发送时间,INTERVAL 4 DAY)) AND login.login_time>0
+THEN '2100-01-01 00:00:00' ELSE ''
+      END 四天内最近一次登录时间
+FROM (
+SELECT
+me.id 消息ID,
+me.title 消息TITLE,
+CASE
+WHEN me.type=1
+THEN 'Interview Message'
+        WHEN me.type=2
+THEN 'Orientation Message'
+        WHEN me.type=3
+THEN 'Warning Message'
+        WHEN me.type=4
+THEN 'Fail Message'
+        WHEN me.type=5
+THEN 'FM Message'
+        WHEN me.type=6
+THEN 'Training Message'
+        WHEN me.type=7
+THEN 'TUrgent Message'
+      END 消息类型,
+FROM_UNIXTIME(me.sending_time) 发送时间,
+IF(tar.is_read=1,FROM_UNIXTIME(tar.read_time),'') 阅读时间,
+tar.tid 老师ID
+FROM ebk_message me
+LEFT JOIN ebk_message_target tar
+ON me.id=tar.msg_id
+WHERE
+FROM_UNIXTIME(me.sending_time,'%Y-%m-%d') BETWEEN 'start' AND 'end' AND me.status=1 AND tar.tid>0
+GROUP BY
+tar.tid,
+me.sending_time,me.id) msg
+LEFT JOIN (
+SELECT tid,FROM_UNIXTIME(login_time) login_time
+FROM ebk_teacher_login_log
+WHERE FROM_UNIXTIME(login_time,'%Y-%m-%d') BETWEEN 'start' AND DATE_ADD('end',INTERVAL 4 DAY)
+ORDER BY tid,FROM_UNIXTIME(login_time)) login
+ON
+msg.老师ID=login.tid
+ORDER BY msg.消息ID,msg.发送时间,msg.老师ID,login_time) x
+GROUP BY x.消息ID,x.发送时间,x.老师ID;
+ERROR 42S02: Table 'test.ebk_message' doesn't exist
+SET NAMES utf8;
+CREATE TABLE t1 (x INT);
+INSERT INTO t1 VALUES (1);
+SELECT x AS 5天内最近一次登录时间 FROM t1;
+5天内最近一次登录时间
+1
+DROP TABLE t1;
+#
 # End of 10.3 tests
 #
--- a/mysql-test/main/ctype_utf8.test
+++ b/mysql-test/main/ctype_utf8.test
@ -2180,6 +2180,81 @@ DROP TABLE t1;
 SET sql_mode=DEFAULT;


+--echo #
+--echo # MDEV-19239 ERROR 1300 (HY000): Invalid utf8 character string in 10.3.13-MariaDB
+--echo #
+
+#
+# Test that the following query does not fail on "Invalid utf8 character string"
+#
+
+SET NAMES utf8;
+--error ER_NO_SUCH_TABLE
+SELECT
+  x.消息ID,
+  x.消息TITLE,
+  x.消息类型,
+  x.发送时间,
+  x.阅读时间,x.老师ID,
+  IF(x.四天内最近一次登录时间='2100-01-01 00:00:00','',x.四天内最近一次登录时间) 四天内最近一次登录时间
+FROM (
+  SELECT
+    msg.*,
+    CASE
+      WHEN login.login_time BETWEEN msg.发送时间 AND DATE_ADD(msg.发送时间,INTERVAL 4 DAY)
+        THEN login.login_time
+      WHEN (login.login_time NOT BETWEEN msg.发送时间 AND DATE_ADD(msg.发送时间,INTERVAL 4 DAY)) AND login.login_time>0
+        THEN '2100-01-01 00:00:00' ELSE ''
+      END 四天内最近一次登录时间
+  FROM (
+    SELECT
+      me.id 消息ID,
+      me.title 消息TITLE,
+      CASE
+        WHEN me.type=1
+          THEN 'Interview Message'
+        WHEN me.type=2
+          THEN 'Orientation Message'
+        WHEN me.type=3
+          THEN 'Warning Message'
+        WHEN me.type=4
+          THEN 'Fail Message'
+        WHEN me.type=5
+          THEN 'FM Message'
+        WHEN me.type=6
+          THEN 'Training Message'
+        WHEN me.type=7
+          THEN 'TUrgent Message'
+      END 消息类型,
+      FROM_UNIXTIME(me.sending_time) 发送时间,
+      IF(tar.is_read=1,FROM_UNIXTIME(tar.read_time),'') 阅读时间,
+      tar.tid 老师ID
+    FROM ebk_message me
+    LEFT JOIN ebk_message_target tar
+    ON me.id=tar.msg_id
+      WHERE
+        FROM_UNIXTIME(me.sending_time,'%Y-%m-%d') BETWEEN 'start' AND 'end' AND me.status=1 AND tar.tid>0
+      GROUP BY
+        tar.tid,
+        me.sending_time,me.id) msg
+    LEFT JOIN (
+      SELECT tid,FROM_UNIXTIME(login_time) login_time
+      FROM ebk_teacher_login_log
+      WHERE FROM_UNIXTIME(login_time,'%Y-%m-%d') BETWEEN 'start' AND DATE_ADD('end',INTERVAL 4 DAY)
+      ORDER BY tid,FROM_UNIXTIME(login_time)) login
+    ON
+      msg.老师ID=login.tid
+    ORDER BY msg.消息ID,msg.发送时间,msg.老师ID,login_time) x
+    GROUP BY x.消息ID,x.发送时间,x.老师ID;
+
+
+SET NAMES utf8;
+CREATE TABLE t1 (x INT);
+INSERT INTO t1 VALUES (1);
+SELECT x AS 5天内最近一次登录时间 FROM t1;
+DROP TABLE t1;
+
+
 --echo #
 --echo # End of 10.3 tests
 --echo #
--- a/sql/sql_lex.cc
+++ b/sql/sql_lex.cc
@ -1587,9 +1587,27 @@ int Lex_input_stream::lex_one_token(YYSTYPE *yylval, THD *thd)
            return(FLOAT_NUM);
          }
        }
+        /*
+          We've found:
+          - A sequence of digits
+          - Followed by 'e' or 'E'
+          - Followed by some byte XX which is not a known mantissa start,
+            and it's known to be a valid identifier part.
+            XX can be either a 8bit identifier character, or a multi-byte head.
+        */
        yyUnget();
+        return scan_ident_start(thd, &yylval->ident_cli);
      }
-      // fall through
+      /*
+        We've found:
+        - A sequence of digits
+        - Followed by some character XX, which is neither 'e' nor 'E',
+          and it's known to be a valid identifier part.
+          XX can be a 8bit identifier character, or a multi-byte head.
+      */
+      yyUnget();
+      return scan_ident_start(thd, &yylval->ident_cli);
+
    case MY_LEX_IDENT_START:                    // We come here after '.'
      return scan_ident_start(thd, &yylval->ident_cli);