bug in my_strnncoll_utf8 (and friends) fixed

cleanups
better, charset-dependent, ft_max_len_for_sort value

myisam/ftdefs.h

@@ -27,7 +27,7 @@
 #define misc_word_char(X)	((X)=='\'')
 #define word_char(s,X)		(true_word_char(s,X) || misc_word_char(X))
 
-#define FT_MAX_WORD_LEN_FOR_SORT 20
+#define FT_MAX_WORD_LEN_FOR_SORT 31
 
 #define COMPILE_STOPWORDS_IN
 

myisam/mi_check.c

@@ -2020,12 +2020,14 @@ int mi_repair_by_sort(MI_CHECK *param, register MI_INFO *info,
 
     if (sort_param.keyinfo->flag & HA_FULLTEXT)
     {
+      uint ft_max_word_len_for_sort=FT_MAX_WORD_LEN_FOR_SORT*
+                                    sort_param.keyinfo->seg->charset->mbmaxlen;
       sort_info.max_records=
-        (ha_rows) (sort_info.filelength/FT_MAX_WORD_LEN_FOR_SORT+1);
+        (ha_rows) (sort_info.filelength/ft_max_word_len_for_sort+1);
 
       sort_param.key_read=sort_ft_key_read;
       sort_param.key_write=sort_ft_key_write;
-      sort_param.key_length+=FT_MAX_WORD_LEN_FOR_SORT-HA_FT_MAXBYTELEN;
+      sort_param.key_length+=ft_max_word_len_for_sort-HA_FT_MAXBYTELEN;
     }
     else
     {
@@ -2425,7 +2427,11 @@ int mi_repair_parallel(MI_CHECK *param, register MI_INFO *info,
     total_key_length+=sort_param[i].key_length;
 
     if (sort_param[i].keyinfo->flag & HA_FULLTEXT)
-      sort_param[i].key_length+=FT_MAX_WORD_LEN_FOR_SORT-HA_FT_MAXBYTELEN;
+    {
+      uint ft_max_word_len_for_sort=FT_MAX_WORD_LEN_FOR_SORT*
+                                    sort_param[i].keyinfo->seg->charset->mbmaxlen;
+      sort_param[i].key_length+=ft_max_word_len_for_sort-HA_FT_MAXBYTELEN;
+    }
   }
   sort_info.total_keys=i;
   sort_param[0].master= 1;
@@ -2634,7 +2640,6 @@ static int sort_key_read(MI_SORT_PARAM *sort_param, void *key)
   DBUG_RETURN(sort_write_record(sort_param));
 } /* sort_key_read */
 
-
 static int sort_ft_key_read(MI_SORT_PARAM *sort_param, void *key)
 {
   int error;
@@ -3950,25 +3955,28 @@ static ha_checksum mi_byte_checksum(const byte *buf, uint length)
   return crc;
 }
 
-/*
-  Deactive all not unique index that can be recreated fast
-  These include packed keys on which sorting will use more temporary
-  space than the max allowed file length or for which the unpacked keys
-  will take much more space than packed keys.
-  Note that 'rows' may be zero for the case when we don't know how many
-  rows we will put into the file.
- */
-
 static my_bool mi_too_big_key_for_sort(MI_KEYDEF *key, ha_rows rows)
 {
   uint key_maxlength=key->maxlength;
   if (key->flag & HA_FULLTEXT)
-    key_maxlength+=FT_MAX_WORD_LEN_FOR_SORT-HA_FT_MAXBYTELEN;
+  {
+    uint ft_max_word_len_for_sort=FT_MAX_WORD_LEN_FOR_SORT*
+                                  key->seg->charset->mbmaxlen;
+    key_maxlength+=ft_max_word_len_for_sort-HA_FT_MAXBYTELEN;
+  }
   return (key->flag & (HA_BINARY_PACK_KEY | HA_VAR_LENGTH_KEY | HA_FULLTEXT) &&
 	  ((ulonglong) rows * key_maxlength >
 	   (ulonglong) myisam_max_temp_length));
 }
 
+/*
+  Deactivate all not unique index that can be recreated fast
+  These include packed keys on which sorting will use more temporary
+  space than the max allowed file length or for which the unpacked keys
+  will take much more space than packed keys.
+  Note that 'rows' may be zero for the case when we don't know how many
+  rows we will put into the file.
+ */
 
 void mi_disable_non_unique_index(MI_INFO *info, ha_rows rows)
 {

myisam/myisamchk.c

@@ -154,7 +154,7 @@ enum options_mc {
   OPT_KEY_CACHE_BLOCK_SIZE, OPT_MYISAM_BLOCK_SIZE,
   OPT_READ_BUFFER_SIZE, OPT_WRITE_BUFFER_SIZE, OPT_SORT_BUFFER_SIZE,
   OPT_SORT_KEY_BLOCKS, OPT_DECODE_BITS, OPT_FT_MIN_WORD_LEN,
-  OPT_FT_MAX_WORD_LEN, OPT_FT_MAX_WORD_LEN_FOR_SORT, OPT_FT_STOPWORD_FILE,
+  OPT_FT_MAX_WORD_LEN, OPT_FT_STOPWORD_FILE,
   OPT_MAX_RECORD_LENGTH
 };
 

myisam/sort.c

@@ -204,7 +204,7 @@ int _create_index_by_sort(MI_SORT_PARAM *info,my_bool no_messages,
 	reinit_io_cache(&tempfile,READ_CACHE,0L,0,0))
       goto err;					/* purecov: inspected */
     if (!no_messages)
-      puts("  - Last merge and dumping keys\n"); /* purecov: tested */
+      printf("  - Last merge and dumping keys\n"); /* purecov: tested */
     if (merge_index(info,keys,sort_keys,dynamic_element(&buffpek,0,BUFFPEK *),
                     maxbuffer,&tempfile))
       goto err;					/* purecov: inspected */
@@ -219,6 +219,8 @@ int _create_index_by_sort(MI_SORT_PARAM *info,my_bool no_messages,
     uint     keyno=info->key;
     uint     key_length, ref_length=index->s->rec_reflength;
 
+    if (!no_messages)
+      printf("  - Adding exceptions\n"); /* purecov: tested */
     if (flush_io_cache(&tempfile_for_exceptions) ||
 	reinit_io_cache(&tempfile_for_exceptions,READ_CACHE,0L,0,0))
       goto err;

sql/mysql_priv.h

@@ -783,10 +783,10 @@ int key_cmp(KEY_PART_INFO *key_part, const byte *key, uint key_length);
 bool init_errmessage(void);
 void sql_perror(const char *message);
 
-void vprint_msg_to_log( enum loglevel level, const char *format, va_list args );
-void sql_print_error( const char *format, ... );
-void sql_print_warning( const char *format, ...); 
-void sql_print_information( const char *format, ...); 
+void vprint_msg_to_log(enum loglevel level, const char *format, va_list args);
+void sql_print_error(const char *format, ...);
+void sql_print_warning(const char *format, ...);
+void sql_print_information(const char *format, ...);
 
 
 

sql/mysqld.cc

@@ -762,7 +762,7 @@ void kill_mysql(void)
     abort_loop=1;
     if (pthread_create(&tmp,&connection_attrib, kill_server_thread,
 			   (void*) 0))
-      sql_print_error("Error: Can't create thread to kill server");
+      sql_print_error("Can't create thread to kill server");
   }
 #endif
   DBUG_VOID_RETURN;
@@ -791,7 +791,7 @@ static void __cdecl kill_server(int sig_ptr)
   abort_loop=1;					// This should be set
   signal(sig,SIG_IGN);
   if (sig == MYSQL_KILL_SIGNAL || sig == 0)
-    sql_print_error(ER(ER_NORMAL_SHUTDOWN),my_progname);
+    sql_print_information(ER(ER_NORMAL_SHUTDOWN),my_progname);
   else
     sql_print_error(ER(ER_GOT_SIGNAL),my_progname,sig); /* purecov: inspected */
 
@@ -806,7 +806,7 @@ static void __cdecl kill_server(int sig_ptr)
 #ifdef __NETWARE__
   pthread_join(select_thread, NULL);		// wait for main thread
 #endif /* __NETWARE__ */
-  
+
   pthread_exit(0);				/* purecov: deadcode */
 
 #endif /* EMBEDDED_LIBRARY */
@@ -834,7 +834,7 @@ extern "C" sig_handler print_signal_warning(int sig)
   if (!DBUG_IN_USE)
   {
     if (global_system_variables.log_warnings)
-      sql_print_error("Warning: Got signal %d from thread %d",
+      sql_print_warning("Got signal %d from thread %d",
 		      sig,my_thread_id());
   }
 #ifdef DONT_REMEMBER_SIGNAL
@@ -961,7 +961,7 @@ void clean_up(bool print_message)
 #endif
 
   if (print_message && errmesg)
-    sql_print_error(ER(ER_SHUTDOWN_COMPLETE),my_progname);
+    sql_print_information(ER(ER_SHUTDOWN_COMPLETE),my_progname);
 #if !defined(__WIN__) && !defined(EMBEDDED_LIBRARY)
   if (!opt_bootstrap)
     (void) my_delete(pidfile_name,MYF(0));	// This may not always exist
@@ -1062,8 +1062,8 @@ static void set_user(const char *user)
       struct passwd *user_info= getpwnam(user);
       if ((!user_info || user_id != user_info->pw_uid) &&
 	  global_system_variables.log_warnings)
-	fprintf(stderr,
-		"Warning: One can only use the --user switch if running as root\n");
+        sql_print_warning(
+                    "One can only use the --user switch if running as root\n");
     }
     return;
   }
@@ -1183,7 +1183,7 @@ static void server_init(void)
     if (listen(ip_sock,(int) back_log) < 0)
     {
       sql_perror("Can't start server: listen() on TCP/IP port");
-      sql_print_error("Error:  listen() on TCP/IP failed with error %d",
+      sql_print_error("listen() on TCP/IP failed with error %d",
 		      socket_errno);
       unireg_abort(1);
     }
@@ -1278,7 +1278,7 @@ static void server_init(void)
     (void) chmod(mysqld_unix_port,S_IFSOCK);	/* Fix solaris 2.6 bug */
 #endif
     if (listen(unix_sock,(int) back_log) < 0)
-      sql_print_error("Warning:  listen() on Unix socket failed with error %d",
+      sql_print_warning("listen() on Unix socket failed with error %d",
 		      socket_errno);
   }
 #endif
@@ -1870,7 +1870,7 @@ static void init_signals(void)
     struct rlimit rl;
     rl.rlim_cur = rl.rlim_max = RLIM_INFINITY;
     if (setrlimit(RLIMIT_CORE, &rl) && global_system_variables.log_warnings)
-      sql_print_error("Warning: setrlimit could not change the size of core files to 'infinity';  We may not be able to generate a core file on signals");
+      sql_print_warning("setrlimit could not change the size of core files to 'infinity';  We may not be able to generate a core file on signals");
   }
 #endif
   (void) sigemptyset(&set);
@@ -2024,7 +2024,7 @@ extern "C" void *signal_hand(void *arg __attribute__((unused)))
     case SIGQUIT:
     case SIGKILL:
 #ifdef EXTRA_DEBUG
-      sql_print_error("Got signal %d to shutdown mysqld",sig);
+      sql_print_information("Got signal %d to shutdown mysqld",sig);
 #endif
       DBUG_PRINT("info",("Got signal: %d  abort_loop: %d",sig,abort_loop));
       if (!abort_loop)
@@ -2036,7 +2036,7 @@ extern "C" void *signal_hand(void *arg __attribute__((unused)))
 	  my_pthread_attr_setprio(&connection_attrib,INTERRUPT_PRIOR);
 	if (pthread_create(&tmp,&connection_attrib, kill_server_thread,
 			   (void*) sig))
-	  sql_print_error("Error: Can't create thread to kill server");
+	  sql_print_error("Can't create thread to kill server");
 #else
 	kill_server((void*) sig);	// MIT THREAD has a alarm thread
 #endif
@@ -2060,7 +2060,7 @@ extern "C" void *signal_hand(void *arg __attribute__((unused)))
 #endif
     default:
 #ifdef EXTRA_DEBUG
-      sql_print_error("Warning: Got signal: %d  error: %d",sig,error); /* purecov: tested */
+      sql_print_warning("Got signal: %d  error: %d",sig,error); /* purecov: tested */
 #endif
       break;					/* purecov: tested */
     }
@@ -2339,11 +2339,11 @@ static int init_common_variables(const char *conf_file_name, int argc,
 		   ("Changed limits: max_open_files: %u  max_connections: %ld  table_cache: %ld",
 		    files, max_connections, table_cache_size));
 	if (global_system_variables.log_warnings)
-	  sql_print_error("Warning: Changed limits: max_open_files: %u  max_connections: %ld  table_cache: %ld",
+	  sql_print_warning("Changed limits: max_open_files: %u  max_connections: %ld  table_cache: %ld",
 			files, max_connections, table_cache_size);
       }
       else if (global_system_variables.log_warnings)
-	sql_print_error("Warning: Could not increase number of max_open_files to more than %u (request: %u)", files, wanted_files);
+	sql_print_warning("Could not increase number of max_open_files to more than %u (request: %u)", files, wanted_files);
     }
     open_files_limit= files;
   }
@@ -2523,8 +2523,8 @@ static int init_server_components()
   }
   else if (opt_log_slave_updates)
   {
-      sql_print_error("\
-Warning: you need to use --log-bin to make --log-slave-updates work. \
+      sql_print_warning("\
+you need to use --log-bin to make --log-slave-updates work. \
 Now disabling --log-slave-updates.");
   }
 
@@ -2532,7 +2532,7 @@ Now disabling --log-slave-updates.");
   if (opt_log_slave_updates && replicate_same_server_id)
   {
     sql_print_error("\
-Error: using --replicate-same-server-id in conjunction with \
+using --replicate-same-server-id in conjunction with \
 --log-slave-updates is impossible, it would lead to infinite loops in this \
 server.");
     unireg_abort(1);
@@ -2561,12 +2561,12 @@ server.");
   if (opt_innodb_safe_binlog)
   {
     if (have_innodb != SHOW_OPTION_YES)
-      sql_print_error("Warning: --innodb-safe-binlog is meaningful only if "
+      sql_print_warning("--innodb-safe-binlog is meaningful only if "
                       "the InnoDB storage engine is enabled in the server.");
 #ifdef HAVE_INNOBASE_DB
     if (innobase_flush_log_at_trx_commit != 1)
     {
-      sql_print_error("Warning: --innodb-safe-binlog is meaningful only if "
+      sql_print_warning("--innodb-safe-binlog is meaningful only if "
                       "innodb_flush_log_at_trx_commit is 1; now setting it "
                       "to 1.");
       innobase_flush_log_at_trx_commit= 1;
@@ -2578,14 +2578,14 @@ server.");
         good (especially "littlesync", and on Windows... see
         srv/srv0start.c).
       */
-      sql_print_error("Warning: --innodb-safe-binlog requires that "
+      sql_print_warning("--innodb-safe-binlog requires that "
                       "the innodb_flush_method actually synchronizes the "
                       "InnoDB log to disk; it is your responsibility "
                       "to verify that the method you chose does it.");
     }
     if (sync_binlog_period != 1)
     {
-      sql_print_error("Warning: --innodb-safe-binlog is meaningful only if "
+      sql_print_warning("--innodb-safe-binlog is meaningful only if "
                       "the global sync_binlog variable is 1; now setting it "
                       "to 1.");
       sync_binlog_period= 1;
@@ -2624,7 +2624,7 @@ server.");
     if (mlockall(MCL_CURRENT))
     {
       if (global_system_variables.log_warnings)
-	sql_print_error("Warning: Failed to lock memory. Errno: %d\n",errno);
+	sql_print_warning("Failed to lock memory. Errno: %d\n",errno);
       locked_in_memory= 0;
     }
   }
@@ -2650,7 +2650,7 @@ static void create_maintenance_thread()
   {
     pthread_t hThread;
     if (pthread_create(&hThread,&connection_attrib,handle_manager,0))
-      sql_print_error("Warning: Can't create thread to manage maintenance");
+      sql_print_warning("Can't create thread to manage maintenance");
   }
 }
 
@@ -2662,7 +2662,7 @@ static void create_shutdown_thread()
   hEventShutdown=CreateEvent(0, FALSE, FALSE, shutdown_event_name);
   pthread_t hThread;
   if (pthread_create(&hThread,&connection_attrib,handle_shutdown,0))
-    sql_print_error("Warning: Can't create thread to handle shutdown requests");
+    sql_print_warning("Can't create thread to handle shutdown requests");
 
   // On "Stop Service" we have to do regular shutdown
   Service.SetShutdownEvent(hEventShutdown);
@@ -2671,7 +2671,7 @@ static void create_shutdown_thread()
   pthread_cond_init(&eventShutdown, NULL);
   pthread_t hThread;
   if (pthread_create(&hThread,&connection_attrib,handle_shutdown,0))
-    sql_print_error("Warning: Can't create thread to handle shutdown requests");
+    sql_print_warning("Can't create thread to handle shutdown requests");
 #endif
 #endif // EMBEDDED_LIBRARY 
 }
@@ -2702,7 +2702,7 @@ static void handle_connections_methods()
     if (pthread_create(&hThread,&connection_attrib,
 		       handle_connections_namedpipes, 0))
     {
-      sql_print_error("Warning: Can't create thread to handle named pipes");
+      sql_print_warning("Can't create thread to handle named pipes");
       handler_count--;
     }
   }
@@ -2713,7 +2713,7 @@ static void handle_connections_methods()
     if (pthread_create(&hThread,&connection_attrib,
 		       handle_connections_sockets, 0))
     {
-      sql_print_error("Warning: Can't create thread to handle TCP/IP");
+      sql_print_warning("Can't create thread to handle TCP/IP");
       handler_count--;
     }
   }
@@ -2724,7 +2724,7 @@ static void handle_connections_methods()
     if (pthread_create(&hThread,&connection_attrib,
 		       handle_connections_shared_memory, 0))
     {
-      sql_print_error("Warning: Can't create thread to handle shared memory");
+      sql_print_warning("Can't create thread to handle shared memory");
       handler_count--;
     }
   }
@@ -2784,7 +2784,7 @@ int main(int argc, char **argv)
     if (stack_size && stack_size < thread_stack)
     {
       if (global_system_variables.log_warnings)
-	sql_print_error("Warning: Asked for %ld thread stack, but got %ld",
+	sql_print_warning("Asked for %ld thread stack, but got %ld",
 			thread_stack, stack_size);
       thread_stack= stack_size;
     }
@@ -2807,8 +2807,8 @@ int main(int argc, char **argv)
     if (lower_case_table_names_used)
     {
       if (global_system_variables.log_warnings)
-	sql_print_error("\
-Warning: You have forced lower_case_table_names to 0 through a command-line \
+	sql_print_warning("\
+You have forced lower_case_table_names to 0 through a command-line \
 option, even though your file system '%s' is case insensitive.  This means \
 that you can corrupt a MyISAM table by accessing it with different cases. \
 You should consider changing lower_case_table_names to 1 or 2",
@@ -2817,7 +2817,7 @@ You should consider changing lower_case_table_names to 1 or 2",
     else
     {
       if (global_system_variables.log_warnings)
-	sql_print_error("Warning: Setting lower_case_table_names=2 because file system for %s is case insensitive", mysql_real_data_home);
+	sql_print_warning("Setting lower_case_table_names=2 because file system for %s is case insensitive", mysql_real_data_home);
       lower_case_table_names= 2;
     }
   }
@@ -2850,14 +2850,14 @@ You should consider changing lower_case_table_names to 1 or 2",
 #ifdef EXTRA_DEBUG
     switch (server_id) {
     case 1:
-      sql_print_error("\
-Warning: You have enabled the binary log, but you haven't set server-id to \
+      sql_print_warning("\
+You have enabled the binary log, but you haven't set server-id to \
 a non-zero value: we force server id to 1; updates will be logged to the \
 binary log, but connections from slaves will not be accepted.");
       break;
     case 2:
-      sql_print_error("\
-Warning: You should set server-id to a non-0 value if master_host is set; \
+      sql_print_warning("\
+You should set server-id to a non-0 value if master_host is set; \
 we force server id to 2, but this MySQL server will not act as a slave.");
       break;
     }
@@ -3197,7 +3197,7 @@ static int bootstrap(FILE *file)
   if (pthread_create(&thd->real_id,&connection_attrib,handle_bootstrap,
 		     (void*) thd))
   {
-    sql_print_error("Warning: Can't create thread to handle bootstrap");
+    sql_print_warning("Can't create thread to handle bootstrap");
     DBUG_RETURN(-1);
   }
   /* Wait for thread to die */
@@ -5606,7 +5606,7 @@ get_one_option(int optid, const struct my_option *opt __attribute__((unused)),
     if (!mysqld_user || !strcmp(mysqld_user, argument))
       mysqld_user= argument;
     else
-      fprintf(stderr, "Warning: Ignoring user change to '%s' because the user was set to '%s' earlier on the command line\n", argument, mysqld_user);
+      sql_print_warning("Ignoring user change to '%s' because the user was set to '%s' earlier on the command line\n", argument, mysqld_user);
     break;
   case 'L':
     strmake(language, argument, sizeof(language)-1);
@@ -6391,7 +6391,7 @@ static int test_if_case_insensitive(const char *dir_name)
   (void) my_delete(buff2, MYF(0));
   if ((file= my_create(buff, 0666, O_RDWR, MYF(0))) < 0)
   {
-    sql_print_error("Warning: Can't create test file %s", buff);
+    sql_print_warning("Can't create test file %s", buff);
     DBUG_RETURN(-1);
   }
   my_close(file, MYF(0));

sql/sql_acl.cc

@@ -251,9 +251,9 @@ my_bool acl_init(THD *org_thd, bool dont_read_acl_tables)
     {
       global_system_variables.old_passwords= 1;
       pthread_mutex_unlock(&LOCK_global_system_variables);
-      sql_print_error("mysql.user table is not updated to new password format; "
-                      "Disabling new password usage until "
-                      "mysql_fix_privilege_tables is run");
+      sql_print_warning("mysql.user table is not updated to new password format; "
+                        "Disabling new password usage until "
+                        "mysql_fix_privilege_tables is run");
     }
     thd->variables.old_passwords= 1;
   }

sql/tztime.cc

@@ -1560,8 +1560,8 @@ my_tz_init(THD *org_thd, const char *default_tzname, my_bool bootstrap)
   if (open_tables(thd, tables_buff, &counter) ||
       lock_tables(thd, tables_buff, counter))
   {
-    sql_print_error("Warning: Can't open and lock time zone table: %s "
-                    "trying to live without them", thd->net.last_error);
+    sql_print_warning("Can't open and lock time zone table: %s "
+                      "trying to live without them", thd->net.last_error);
     /* We will try emulate that everything is ok */
     return_val= time_zone_tables_exist= 0;
     goto end_with_setting_default_tz;
@@ -1740,8 +1740,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
   if (!(alloc_buff= alloc_root(&tz_storage, sizeof(TIME_ZONE_INFO) +
                                tz_name->length() + 1)))
   {
-    sql_print_error("Error: Out of memory while loading time zone "
-                    "description");
+    sql_print_error("Out of memory while loading time zone description");
     return 0;
   }
   tz_info= (TIME_ZONE_INFO *)alloc_buff;
@@ -1757,7 +1756,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
     Let us find out time zone id by its name (there is only one index
     and it is specifically for this purpose).
   */
-  table= tz_tables->table; 
+  table= tz_tables->table;
   tz_tables= tz_tables->next;
   table->field[0]->store(tz_name->ptr(), tz_name->length(), &my_charset_latin1);
   /*
@@ -1770,7 +1769,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
   if (table->file->index_read(table->record[0], (byte*)table->field[0]->ptr,
                               0, HA_READ_KEY_EXACT))
   {
-    sql_print_error("Error: Can't find description of time zone.");
+    sql_print_error("Can't find description of time zone.");
     goto end;
   }
 
@@ -1783,7 +1782,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
     understand whenever this timezone uses leap seconds (again we are
     using the only index in this table).
   */
-  table= tz_tables->table; 
+  table= tz_tables->table;
   tz_tables= tz_tables->next;
   table->field[0]->store((longlong)tzid);
   (void)table->file->ha_index_init(0);
@@ -1791,7 +1790,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
   if (table->file->index_read(table->record[0], (byte*)table->field[0]->ptr,
                               0, HA_READ_KEY_EXACT))
   {
-    sql_print_error("Error: Can't find description of time zone.");
+    sql_print_error("Can't find description of time zone.");
     goto end;
   }
 
@@ -1810,7 +1809,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
     only for our time zone guess what are we doing?
     Right - using special index.
   */
-  table= tz_tables->table; 
+  table= tz_tables->table;
   tz_tables= tz_tables->next;
   table->field[0]->store((longlong)tzid);
   (void)table->file->ha_index_init(0);
@@ -1948,8 +1947,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
 #endif
                                sizeof(TRAN_TYPE_INFO) * tz_info->typecnt)))
   {
-    sql_print_error("Error: Out of memory while loading time zone "
-                    "description");
+    sql_print_error("Out of memory while loading time zone description");
     goto end;
   }
 
@@ -1974,12 +1972,12 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
   */
   if (tz_info->typecnt < 1)
   {
-    sql_print_error("Error: loading time zone without transition types");
+    sql_print_error("loading time zone without transition types");
     goto end;
   }
   if (prepare_tz_info(tz_info, &tz_storage))
   {
-    sql_print_error("Error: Unable to build mktime map for time zone");
+    sql_print_error("Unable to build mktime map for time zone");
     goto end;
   }
 
@@ -1991,7 +1989,7 @@ tz_load_from_open_tables(const String *tz_name, TABLE_LIST *tz_tables)
                             &my_charset_latin1),
        my_hash_insert(&tz_names, (const byte *)tmp_tzname)))
   {
-    sql_print_error("Error: Out of memory while loading time zone");
+    sql_print_error("Out of memory while loading time zone");
     goto end;
   }
 

strings/ctype-utf8.c

 /* Copyright (C) 2000 MySQL AB
-   
+
    This library is free software; you can redistribute it and/or
    modify it under the terms of the GNU Library General Public
    License as published by the Free Software Foundation; either
    version 2 of the License, or (at your option) any later version.
-   
+
    This library is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    Library General Public License for more details.
-   
+
    You should have received a copy of the GNU Library General Public
    License along with this library; if not, write to the Free
    Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
@@ -1524,7 +1524,7 @@ MY_UNICASE_INFO *uni_plane[256]={
 
 #ifdef HAVE_CHARSET_utf8
 
-/* 
+/*
   We consider bytes with code more than 127 as a letter.
   This garantees that word boundaries work fine with regular
   expressions. Note, there is no need to mark byte 255  as a
@@ -1590,99 +1590,108 @@ static uchar to_upper_utf8[] = {
   240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255
 };
 
+static inline int bincmp(const uchar *s, const uchar *se,
+                         const uchar *t, const uchar *te)
+{
+  int slen=se-s, tlen=te-t;
+  int len=min(slen,tlen);
+  int cmp= memcmp(s,t,len);
+  return cmp ? cmp : slen-tlen;
+}
+
 
 static int my_utf8_uni(CHARSET_INFO *cs __attribute__((unused)),
-		       my_wc_t * pwc, const uchar *s, const uchar *e)
+                       my_wc_t * pwc, const uchar *s, const uchar *e)
 {
   unsigned char c;
-  
+
   if (s >= e)
     return MY_CS_TOOFEW(0);
 
   c= s[0];
-  if (c < 0x80) 
+  if (c < 0x80)
   {
     *pwc = c;
     return 1;
-  } 
-  else if (c < 0xc2) 
+  }
+  else if (c < 0xc2)
     return MY_CS_ILSEQ;
-  else if (c < 0xe0) 
+  else if (c < 0xe0)
   {
-    if (s+2 > e) /* We need 2 characters */ 
+    if (s+2 > e) /* We need 2 characters */
       return MY_CS_TOOFEW(0);
-    
+
     if (!((s[1] ^ 0x80) < 0x40))
       return MY_CS_ILSEQ;
-    
+
     *pwc = ((my_wc_t) (c & 0x1f) << 6) | (my_wc_t) (s[1] ^ 0x80);
     return 2;
-  } 
-  else if (c < 0xf0) 
+  }
+  else if (c < 0xf0)
   {
     if (s+3 > e) /* We need 3 characters */
       return MY_CS_TOOFEW(0);
-    
+
     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40 && (c >= 0xe1 || s[1] >= 0xa0)))
       return MY_CS_ILSEQ;
-    
-    *pwc = ((my_wc_t) (c & 0x0f) << 12)   | 
-           ((my_wc_t) (s[1] ^ 0x80) << 6) | 
+
+    *pwc = ((my_wc_t) (c & 0x0f) << 12)   |
+           ((my_wc_t) (s[1] ^ 0x80) << 6) |
             (my_wc_t) (s[2] ^ 0x80);
-    
+
     return 3;
-  } 
+  }
 #ifdef UNICODE_32BIT
-  else if (c < 0xf8 && sizeof(my_wc_t)*8 >= 32) 
+  else if (c < 0xf8 && sizeof(my_wc_t)*8 >= 32)
   {
     if (s+4 > e) /* We need 4 characters */
       return MY_CS_TOOFEW(0);
-    
-    if (!((s[1] ^ 0x80) < 0x40 && 
-          (s[2] ^ 0x80) < 0x40 && 
-          (s[3] ^ 0x80) < 0x40 && 
+
+    if (!((s[1] ^ 0x80) < 0x40 &&
+          (s[2] ^ 0x80) < 0x40 &&
+          (s[3] ^ 0x80) < 0x40 &&
           (c >= 0xf1 || s[1] >= 0x90)))
       return MY_CS_ILSEQ;
-    
-    *pwc = ((my_wc_t) (c & 0x07) << 18)    | 
-           ((my_wc_t) (s[1] ^ 0x80) << 12) | 
-           ((my_wc_t) (s[2] ^ 0x80) << 6)  | 
+
+    *pwc = ((my_wc_t) (c & 0x07) << 18)    |
+           ((my_wc_t) (s[1] ^ 0x80) << 12) |
+           ((my_wc_t) (s[2] ^ 0x80) << 6)  |
             (my_wc_t) (s[3] ^ 0x80);
-    
+
     return 4;
   }
-   else if (c < 0xfc && sizeof(my_wc_t)*8 >= 32) 
+   else if (c < 0xfc && sizeof(my_wc_t)*8 >= 32)
   {
     if (s+5 >e) /* We need 5 characters */
       return MY_CS_TOOFEW(0);
-    
-    if (!((s[1] ^ 0x80) < 0x40 && 
-          (s[2] ^ 0x80) < 0x40 && 
-          (s[3] ^ 0x80) < 0x40 && 
-          (s[4] ^ 0x80) < 0x40 && 
+
+    if (!((s[1] ^ 0x80) < 0x40 &&
+          (s[2] ^ 0x80) < 0x40 &&
+          (s[3] ^ 0x80) < 0x40 &&
+          (s[4] ^ 0x80) < 0x40 &&
           (c >= 0xf9 || s[1] >= 0x88)))
       return MY_CS_ILSEQ;
-    
-    *pwc = ((my_wc_t) (c & 0x03) << 24) | 
+
+    *pwc = ((my_wc_t) (c & 0x03) << 24) |
            ((my_wc_t) (s[1] ^ 0x80) << 18) |
            ((my_wc_t) (s[2] ^ 0x80) << 12) |
            ((my_wc_t) (s[3] ^ 0x80) << 6) |
             (my_wc_t) (s[4] ^ 0x80);
     return 5;
-  } 
-  else if (c < 0xfe && sizeof(my_wc_t)*8 >= 32) 
+  }
+  else if (c < 0xfe && sizeof(my_wc_t)*8 >= 32)
   {
     if ( s+6 >e ) /* We need 6 characters */
       return MY_CS_TOOFEW(0);
-    
-    if (!((s[1] ^ 0x80) < 0x40   && 
-          (s[2] ^ 0x80) < 0x40   && 
-          (s[3] ^ 0x80) < 0x40   && 
-          (s[4] ^ 0x80) < 0x40   && 
-          (s[5] ^ 0x80) < 0x40   && 
+
+    if (!((s[1] ^ 0x80) < 0x40   &&
+          (s[2] ^ 0x80) < 0x40   &&
+          (s[3] ^ 0x80) < 0x40   &&
+          (s[4] ^ 0x80) < 0x40   &&
+          (s[5] ^ 0x80) < 0x40   &&
           (c >= 0xfd || s[1] >= 0x84)))
       return MY_CS_ILSEQ;
-    
+
     *pwc = ((my_wc_t) (c & 0x01) << 30)
       | ((my_wc_t) (s[1] ^ 0x80) << 24)
       | ((my_wc_t) (s[2] ^ 0x80) << 18)
@@ -1702,12 +1711,12 @@ static int my_uni_utf8 (CHARSET_INFO *cs __attribute__((unused)) ,
 
   if (r >= e)
     return MY_CS_TOOSMALL;
-  
-  if (wc < 0x80) 
+
+  if (wc < 0x80)
     count = 1;
-  else if (wc < 0x800) 
+  else if (wc < 0x800)
     count = 2;
-  else if (wc < 0x10000) 
+  else if (wc < 0x10000)
     count = 3;
 #ifdef UNICODE_32BIT
   else if (wc < 0x200000)
@@ -1718,15 +1727,15 @@ static int my_uni_utf8 (CHARSET_INFO *cs __attribute__((unused)) ,
     count = 6;
 #endif
     else return MY_CS_ILUNI;
-  
-  /* 
-    e is a character after the string r, not the last character of it. 
+
+  /*
+    e is a character after the string r, not the last character of it.
     Because of it (r+count > e), not (r+count-1 >e )
    */
-  if ( r+count > e ) 
+  if ( r+count > e )
     return MY_CS_TOOSMALL;
-  
-  switch (count) { 
+
+  switch (count) {
     /* Fall through all cases!!! */
 #ifdef UNICODE_32BIT
     case 6: r[5] = (uchar) (0x80 | (wc & 0x3f)); wc = wc >> 6; wc |= 0x4000000;
@@ -1806,8 +1815,8 @@ static void my_casedn_str_utf8(CHARSET_INFO *cs, char * s)
 }
 
 
-static int my_strnncoll_utf8(CHARSET_INFO *cs, 
-			     const uchar *s, uint slen,
+static int my_strnncoll_utf8(CHARSET_INFO *cs,
+                             const uchar *s, uint slen,
                              const uchar *t, uint tlen,
                              my_bool t_is_prefix)
 {
@@ -1821,13 +1830,13 @@ static int my_strnncoll_utf8(CHARSET_INFO *cs,
     int plane;
     s_res=my_utf8_uni(cs,&s_wc, s, se);
     t_res=my_utf8_uni(cs,&t_wc, t, te);
-    
+
     if ( s_res <= 0 || t_res <= 0 )
     {
-      /* Incorrect string, compare by char value */
-      return ((int)s[0]-(int)t[0]); 
+      /* Incorrect string, compare byte by byte value */
+      return bincmp(s, se, t, te);
     }
-    
+
     plane=(s_wc>>8) & 0xFF;
     s_wc = uni_plane[plane] ? uni_plane[plane][s_wc & 0xFF].sort : s_wc;
     plane=(t_wc>>8) & 0xFF;
@@ -1836,7 +1845,7 @@ static int my_strnncoll_utf8(CHARSET_INFO *cs,
     {
       return  ((int) s_wc) - ((int) t_wc);
     }
-    
+
     s+=s_res;
     t+=t_res;
   }
@@ -1850,11 +1859,11 @@ static int my_strnncoll_utf8(CHARSET_INFO *cs,
 
   SYNOPSIS
     my_strnncollsp_utf8()
-    cs			character set handler
-    a			First string to compare
-    a_length		Length of 'a'
-    b			Second string to compare
-    b_length		Length of 'b'
+    cs                  character set handler
+    a                   First string to compare
+    a_length            Length of 'a'
+    b                   Second string to compare
+    b_length            Length of 'b'
 
   IMPLEMENTATION
     If one string is shorter as the other, then we space extend the other
@@ -1867,32 +1876,32 @@ static int my_strnncoll_utf8(CHARSET_INFO *cs,
     "a\0" < "a "
 
   RETURN
-    < 0	 a <  b
-    = 0	 a == b
-    > 0	 a > b
+    < 0  a <  b
+    = 0  a == b
+    > 0  a > b
 */
 
-static int my_strnncollsp_utf8(CHARSET_INFO *cs, 
-			     const uchar *s, uint slen,
-			     const uchar *t, uint tlen)
+static int my_strnncollsp_utf8(CHARSET_INFO *cs,
+                             const uchar *s, uint slen,
+                             const uchar *t, uint tlen)
 {
   int s_res,t_res;
   my_wc_t s_wc,t_wc;
   const uchar *se= s+slen;
   const uchar *te= t+tlen;
-  
+
   while ( s < se && t < te )
   {
     int plane;
     s_res=my_utf8_uni(cs,&s_wc, s, se);
     t_res=my_utf8_uni(cs,&t_wc, t, te);
-    
+
     if ( s_res <= 0 || t_res <= 0 )
     {
-      /* Incorrect string, compare by char value */
-      return ((int)s[0]-(int)t[0]); 
+      /* Incorrect string, compare byte by byte value */
+      return bincmp(s, se, t, te);
     }
-    
+
     plane=(s_wc>>8) & 0xFF;
     s_wc = uni_plane[plane] ? uni_plane[plane][s_wc & 0xFF].sort : s_wc;
     plane=(t_wc>>8) & 0xFF;
@@ -1901,14 +1910,14 @@ static int my_strnncollsp_utf8(CHARSET_INFO *cs,
     {
       return  ((int) s_wc) - ((int) t_wc);
     }
-    
+
     s+=s_res;
     t+=t_res;
   }
-  
+
   slen= se-s;
   tlen= te-t;
-  
+
   if (slen != tlen)
   {
     int swap= 0;
@@ -1940,35 +1949,35 @@ static int my_strnncollsp_utf8(CHARSET_INFO *cs,
 
 
 static int my_strncasecmp_utf8(CHARSET_INFO *cs,
-		const char *s, const char *t,  uint len)
+                const char *s, const char *t,  uint len)
 {
   int s_res,t_res;
   my_wc_t s_wc,t_wc;
   const char *se=s+len;
   const char *te=t+len;
-  
+
   while ( s < se && t < te )
   {
     int plane;
-    
+
     s_res=my_utf8_uni(cs,&s_wc, (const uchar*)s, (const uchar*)se);
     t_res=my_utf8_uni(cs,&t_wc, (const uchar*)t, (const uchar*)te);
-    
+
     if ( s_res <= 0 || t_res <= 0 )
     {
-      /* Incorrect string, compare by char value */
-      return ((int)s[0]-(int)t[0]); 
+      /* Incorrect string, compare byte by byte value */
+      return bincmp(s, se, t, te);
     }
-    
+
     plane=(s_wc>>8) & 0xFF;
     s_wc = uni_plane[plane] ? uni_plane[plane][s_wc & 0xFF].tolower : s_wc;
 
     plane=(t_wc>>8) & 0xFF;
     t_wc = uni_plane[plane] ? uni_plane[plane][t_wc & 0xFF].tolower : t_wc;
-    
+
     if ( s_wc != t_wc )
       return  ((int) s_wc) - ((int) t_wc);
-    
+
     s+=s_res;
     t+=t_res;
   }
@@ -1983,9 +1992,9 @@ static int my_strcasecmp_utf8(CHARSET_INFO *cs, const char *s, const char *t)
   return  my_strncasecmp_utf8(cs, s, t, len);
 }
 
-static int my_strnxfrm_utf8(CHARSET_INFO *cs, 
-			    uchar *dst, uint dstlen, 
-			    const uchar *src, uint srclen)
+static int my_strnxfrm_utf8(CHARSET_INFO *cs,
+                            uchar *dst, uint dstlen,
+                            const uchar *src, uint srclen)
 {
   my_wc_t wc;
   int res;
@@ -2002,10 +2011,10 @@ static int my_strnxfrm_utf8(CHARSET_INFO *cs,
     }
     src+=res;
     srclen-=res;
-    
+
     plane=(wc>>8) & 0xFF;
     wc = uni_plane[plane] ? uni_plane[plane][wc & 0xFF].sort : wc;
-    
+
     if ((res=my_uni_utf8(cs,wc,dst,de)) <0)
     {
       break;
@@ -2026,18 +2035,18 @@ static int my_mbcharlen_utf8(CHARSET_INFO *cs  __attribute__((unused)) , uint c)
 {
   if (c < 0x80)
     return 1;
-  else if (c < 0xc2) 
+  else if (c < 0xc2)
     return 0; /* Illegal mb head */
-  else if (c < 0xe0) 
+  else if (c < 0xe0)
     return 2;
-  else if (c < 0xf0) 
+  else if (c < 0xf0)
     return 3;
 #ifdef UNICODE_32BIT
-  else if (c < 0xf8) 
+  else if (c < 0xf8)
     return 4;
-  else if (c < 0xfc) 
+  else if (c < 0xfc)
     return 5;
-  else if (c < 0xfe) 
+  else if (c < 0xfe)
     return 6;
 #endif
   return 0; /* Illegal mb head */;
@@ -2046,7 +2055,7 @@ static int my_mbcharlen_utf8(CHARSET_INFO *cs  __attribute__((unused)) , uint c)
 
 static MY_COLLATION_HANDLER my_collation_ci_handler =
 {
-    NULL,		/* init */
+    NULL,               /* init */
     my_strnncoll_utf8,
     my_strnncollsp_utf8,
     my_strnxfrm_utf8,
@@ -2059,7 +2068,7 @@ static MY_COLLATION_HANDLER my_collation_ci_handler =
 
 MY_CHARSET_HANDLER my_charset_utf8_handler=
 {
-    NULL,		/* init */
+    NULL,               /* init */
     my_ismbchar_utf8,
     my_mbcharlen_utf8,
     my_numchars_mb,
@@ -2089,27 +2098,27 @@ MY_CHARSET_HANDLER my_charset_utf8_handler=
 
 CHARSET_INFO my_charset_utf8_general_ci=
 {
-    33,0,0,		/* number       */
-    MY_CS_COMPILED|MY_CS_PRIMARY|MY_CS_STRNXFRM|MY_CS_UNICODE,	/* state  */
-    "utf8",		/* cs name      */
-    "utf8_general_ci",	/* name         */
-    "",			/* comment      */
-    NULL,		/* tailoring    */
-    ctype_utf8,		/* ctype        */
-    to_lower_utf8,	/* to_lower     */
-    to_upper_utf8,	/* to_upper     */
-    to_upper_utf8,	/* sort_order   */
-    NULL,		/* contractions */
-    NULL,		/* sort_order_big*/
-    NULL,		/* tab_to_uni   */
-    NULL,		/* tab_from_uni */
-    NULL,		/* state_map    */
-    NULL,		/* ident_map    */
-    1,			/* strxfrm_multiply */
-    1,			/* mbminlen     */
-    3,			/* mbmaxlen     */
-    0,			/* min_sort_char */
-    255,		/* max_sort_char */
+    33,0,0,             /* number       */
+    MY_CS_COMPILED|MY_CS_PRIMARY|MY_CS_STRNXFRM|MY_CS_UNICODE,  /* state  */
+    "utf8",             /* cs name      */
+    "utf8_general_ci",  /* name         */
+    "",                 /* comment      */
+    NULL,               /* tailoring    */
+    ctype_utf8,         /* ctype        */
+    to_lower_utf8,      /* to_lower     */
+    to_upper_utf8,      /* to_upper     */
+    to_upper_utf8,      /* sort_order   */
+    NULL,               /* contractions */
+    NULL,               /* sort_order_big*/
+    NULL,               /* tab_to_uni   */
+    NULL,               /* tab_from_uni */
+    NULL,               /* state_map    */
+    NULL,               /* ident_map    */
+    1,                  /* strxfrm_multiply */
+    1,                  /* mbminlen     */
+    3,                  /* mbmaxlen     */
+    0,                  /* min_sort_char */
+    255,                /* max_sort_char */
     &my_charset_utf8_handler,
     &my_collation_ci_handler
 };
@@ -2117,27 +2126,27 @@ CHARSET_INFO my_charset_utf8_general_ci=
 
 CHARSET_INFO my_charset_utf8_bin=
 {
-    83,0,0,		/* number       */
-    MY_CS_COMPILED|MY_CS_BINSORT|MY_CS_UNICODE,	/* state  */
-    "utf8",		/* cs name      */
-    "utf8_bin",		/* name         */
-    "",			/* comment      */
-    NULL,		/* tailoring    */
-    ctype_utf8,		/* ctype        */
-    to_lower_utf8,	/* to_lower     */
-    to_upper_utf8,	/* to_upper     */
-    NULL,		/* sort_order   */
-    NULL,		/* contractions */
-    NULL,		/* sort_order_big*/
-    NULL,		/* tab_to_uni   */
-    NULL,		/* tab_from_uni */
-    NULL,		/* state_map    */
-    NULL,		/* ident_map    */
-    1,			/* strxfrm_multiply */
-    1,			/* mbminlen     */
-    3,			/* mbmaxlen     */
-    0,			/* min_sort_char */
-    255,		/* max_sort_char */
+    83,0,0,             /* number       */
+    MY_CS_COMPILED|MY_CS_BINSORT|MY_CS_UNICODE, /* state  */
+    "utf8",             /* cs name      */
+    "utf8_bin",         /* name         */
+    "",                 /* comment      */
+    NULL,               /* tailoring    */
+    ctype_utf8,         /* ctype        */
+    to_lower_utf8,      /* to_lower     */
+    to_upper_utf8,      /* to_upper     */
+    NULL,               /* sort_order   */
+    NULL,               /* contractions */
+    NULL,               /* sort_order_big*/
+    NULL,               /* tab_to_uni   */
+    NULL,               /* tab_from_uni */
+    NULL,               /* state_map    */
+    NULL,               /* ident_map    */
+    1,                  /* strxfrm_multiply */
+    1,                  /* mbminlen     */
+    3,                  /* mbmaxlen     */
+    0,                  /* min_sort_char */
+    255,                /* max_sort_char */
     &my_charset_utf8_handler,
     &my_collation_mb_bin_handler
 };
@@ -2155,8 +2164,8 @@ static void test_mb(CHARSET_INFO *cs, uchar *s)
       int len=my_mbcharlen_utf8(cs,*s);
       while(len--)
       {
-      	printf("%c",*s);
-      	s++;
+        printf("%c",*s);
+        s++;
       }
       printf("\n");
     }
@@ -2172,23 +2181,23 @@ int main()
 {
   char str[1024]=" utf8 test проба ПЕРА по-РУССКИ";
   CHARSET_INFO *cs;
-  
+
   test_mb(cs,(uchar*)str);
-  
+
   printf("orig      :'%s'\n",str);
-  
+
   my_caseup_utf8(cs,str,15);
   printf("caseup    :'%s'\n",str);
-  
+
   my_caseup_str_utf8(cs,str);
   printf("caseup_str:'%s'\n",str);
-  
+
   my_casedn_utf8(cs,str,15);
   printf("casedn    :'%s'\n",str);
-  
+
   my_casedn_str_utf8(cs,str);
   printf("casedn_str:'%s'\n",str);
-  
+
   return 0;
 }