Bug#31063981: ASSERTION "MYSQLLEX"

The parser doesn't support multi-byte character sets with
minimum length greater than 1. Hence the functions that call
the parser shouldn't either.

Change-Id: I3afe847a4e7749513540ee88e77a51cc6ebaaac5

Author: Martin Hansson

include/m_ctype.h

@@ -1,4 +1,4 @@
-/* Copyright (c) 2000, 2019, Oracle and/or its affiliates. All rights reserved.
+/* Copyright (c) 2000, 2020, Oracle and/or its affiliates. All rights reserved.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License, version 2.0,
@@ -743,4 +743,8 @@ static inline uint my_ismbchar(const CHARSET_INFO *cs, const uchar *str,
   ((s)->cset->strntoull((s), (a), (b), (c), (d), (e)))
 #define my_strntod(s, a, b, c, d) ((s)->cset->strntod((s), (a), (b), (c), (d)))
 
+static inline bool is_supported_parser_charset(const CHARSET_INFO *cs) {
+  return (cs->mbminlen == 1);
+}
+
 #endif  // M_CTYPE_INCLUDED

share/messages_to_clients.txt

@@ -9373,6 +9373,9 @@ ER_MDL_OUT_OF_RESOURCES
 ER_IMPLICIT_COMPARISON_FOR_JSON
   eng "Evaluating a JSON value in SQL boolean context does an implicit comparison against JSON integer 0; if this is not what you want, consider converting JSON to a SQL numeric type with JSON_VALUE RETURNING"
 
+ER_FUNCTION_DOES_NOT_SUPPORT_CHARACTER_SET
+  eng "The function %s does not support the character set '%s'."
+
 #
 #  End of 8.0 error messages (server-to-client).
 #  Do NOT add messages intended for the error log above!

sql/item_strfunc.cc

@@ -46,6 +46,9 @@
 
 #include "base64.h"  // base64_encode_max_arg_length
 #include "decimal.h"
+#include "field_types.h"  // MYSQL_TYPE_BIT
+#include "lex_string.h"   // LEX_CSTRING
+#include "m_ctype.h"      // is_supported_parser_charset
 #include "m_string.h"
 #include "my_aes.h"  // MY_AES_IV_SIZE
 #include "my_byteorder.h"
@@ -926,7 +929,12 @@ String *Item_func_statement_digest::val_str_ascii(String *buf) {
   {
     THD *thd = current_thd;
     Thd_parse_modifier thd_mod(thd, m_token_buffer);
-
+    const CHARSET_INFO *cs = args[0]->charset_for_protocol();
+    if (!is_supported_parser_charset(cs)) {
+      my_error(ER_FUNCTION_DOES_NOT_SUPPORT_CHARACTER_SET, myf(0), func_name(),
+               cs->name);
+      return error_str();
+    }
     if (parse(thd, args[0], statement_string)) return error_str();
     compute_digest_hash(&thd->m_digest->m_digest_storage, digest);
   }
@@ -956,7 +964,12 @@ String *Item_func_statement_digest_text::val_str(String *buf) {
 
   THD *thd = current_thd;
   Thd_parse_modifier thd_mod(thd, m_token_buffer);
-
+  const CHARSET_INFO *cs = args[0]->charset_for_protocol();
+  if (!is_supported_parser_charset(cs)) {
+    my_error(ER_FUNCTION_DOES_NOT_SUPPORT_CHARACTER_SET, myf(0), func_name(),
+             cs->name);
+    return error_str();
+  }
   if (parse(thd, args[0], statement_string)) return error_str();
 
   compute_digest_text(&thd->m_digest->m_digest_storage, buf);

sql/sql_parse.h

@@ -127,10 +127,6 @@ bool show_precheck(THD *thd, LEX *lex, bool lock);
 extern uint sql_command_flags[];
 extern const LEX_CSTRING command_name[];
 
-inline bool is_supported_parser_charset(const CHARSET_INFO *cs) {
-  return (cs->mbminlen == 1);
-}
-
 bool sqlcom_can_generate_row_events(enum enum_sql_command command);
 
 bool all_tables_not_ok(THD *thd, TABLE_LIST *tables);

sql/sys_vars.cc

@@ -1749,7 +1749,10 @@ static Sys_var_struct<CHARSET_INFO, Get_csname> Sys_character_set_database(
 static bool check_cs_client(sys_var *self, THD *thd, set_var *var) {
   if (check_charset_not_null(self, thd, var)) return true;
 
-  // Currently, UCS-2 cannot be used as a client character set
+  // We don't currently support any variable-width character set with a minumum
+  // length greater than 1. If we ever do, we have to revisit
+  // is_supported_parser_charset(). See Item_func_statement_digest::val_str()
+  // and Item_func_statement_digest_text::val_str().
   return (static_cast<const CHARSET_INFO *>(var->save_result.ptr))->mbminlen >
          1;
 }