MDEV-4297 mysql --binary-mode

backport mysql --binary-mode (bug#11747577, bug#33048)

5 files changed, 226 insertions, 72 deletions

diff --git a/client/client_priv.h b/client/client_priv.h
index 70a9d129433..e206804d057 100644
--- a/client/client_priv.h
+++ b/client/client_priv.h
@@ -1,5 +1,5 @@
 /*
-   Copyright (c) 2001, 2012, Oracle and/or its affiliates. All rights reserved.
+   Copyright (c) 2001, 2012, Oracle and/or its affiliates.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
diff --git a/client/my_readline.h b/client/my_readline.h
index 11ace987b44..57537308fed 100644
--- a/client/my_readline.h
+++ b/client/my_readline.h
@@ -36,7 +36,7 @@ typedef struct st_line_buffer
 
 extern LINE_BUFFER *batch_readline_init(ulong max_size,FILE *file);
 extern LINE_BUFFER *batch_readline_command(LINE_BUFFER *buffer, char * str);
-extern char *batch_readline(LINE_BUFFER *buffer);
+extern char *batch_readline(LINE_BUFFER *buffer, bool binary_mode);
 extern void batch_readline_end(LINE_BUFFER *buffer);
 
 #endif /* CLIENT_MY_READLINE_INCLUDED */
diff --git a/client/mysql.cc b/client/mysql.cc
index 3cd078ac1ce..2c8f7347840 100644
--- a/client/mysql.cc
+++ b/client/mysql.cc
@@ -1,6 +1,6 @@
 /*
    Copyright (c) 2000, 2012, Oracle and/or its affiliates.
-   Copyright (c) 2009, 2012, Monty Program Ab.
+   Copyright (c) 2009, 2013, Monty Program Ab.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
@@ -152,6 +152,7 @@ static uint verbose=0,opt_silent=0,opt_mysql_port=0, opt_local_infile=0;
 static uint my_end_arg;
 static char * opt_mysql_unix_port=0;
 static int connect_flag=CLIENT_INTERACTIVE;
+static my_bool opt_binary_mode= FALSE;
 static int interrupted_query= 0;
 static char *current_host,*current_db,*current_user=0,*opt_password=0,
             *current_prompt=0, *delimiter_str= 0,
@@ -1056,9 +1057,10 @@ static void initialize_readline (char *name);
 static void fix_history(String *final_command);
 #endif
 
-static COMMANDS *find_command(char *name,char cmd_name);
-static bool add_line(String &buffer,char *line,char *in_string,
-                     bool *ml_comment, bool truncated);
+static COMMANDS *find_command(char *name);
+static COMMANDS *find_command(char cmd_name);
+static bool add_line(String &buffer, char *line, ulong line_length,
+                     char *in_string, bool *ml_comment, bool truncated);
 static void remove_cntrl(String &buffer);
 static void print_table_data(MYSQL_RES *result);
 static void print_table_data_html(MYSQL_RES *result);
@@ -1077,6 +1079,45 @@ static sig_handler window_resize(int sig);
 #endif
 
 
+const char DELIMITER_NAME[]= "delimiter";
+const uint DELIMITER_NAME_LEN= sizeof(DELIMITER_NAME) - 1;
+inline bool is_delimiter_command(char *name, ulong len)
+{
+  /*
+    Delimiter command has a parameter, so the length of the whole command
+    is larger than DELIMITER_NAME_LEN.  We don't care the parameter, so
+    only name(first DELIMITER_NAME_LEN bytes) is checked.
+  */
+  return (len >= DELIMITER_NAME_LEN &&
+          !my_strnncoll(charset_info, (uchar*) name, DELIMITER_NAME_LEN,
+                        (uchar *) DELIMITER_NAME, DELIMITER_NAME_LEN));
+}
+
+/**
+   Get the index of a command in the commands array.
+
+   @param cmd_char    Short form command.
+
+   @return int
+     The index of the command is returned if it is found, else -1 is returned.
+*/
+inline int get_command_index(char cmd_char)
+{
+  /*
+    All client-specific commands are in the first part of commands array
+    and have a function to implement it.
+  */
+  for (uint i= 0; *commands[i].func; i++)
+    if (commands[i].cmd_char == cmd_char)
+      return i;
+  return -1;
+}
+
+static int delimiter_index= -1;
+static int charset_index= -1;
+static bool real_binary_mode= FALSE;
+
+
 int main(int argc,char *argv[])
 {
   char buff[80];
@@ -1085,6 +1126,8 @@ int main(int argc,char *argv[])
   DBUG_ENTER("main");
   DBUG_PROCESS(argv[0]);
   
+  charset_index= get_command_index('C');
+  delimiter_index= get_command_index('d');
   delimiter_str= delimiter;
   default_prompt = my_strdup(getenv("MYSQL_PS1") ? 
 			     getenv("MYSQL_PS1") : 
@@ -1598,6 +1641,13 @@ static struct my_option my_long_options[] =
     "Default authentication client-side plugin to use.",
     &opt_default_auth, &opt_default_auth, 0,
    GET_STR, REQUIRED_ARG, 0, 0, 0, 0, 0, 0},
+  {"binary-mode", 0,
+   "By default, ASCII '\\0' is disallowed and '\\r\\n' is translated to '\\n'. "
+   "This switch turns off both features, and also turns off parsing of all client"
+   "commands except \\C and DELIMITER, in non-interactive mode (for input "
+   "piped to mysql or loaded using the 'source' command). This is necessary "
+   "when processing output from mysqlbinlog that may contain blobs.",
+   &opt_binary_mode, &opt_binary_mode, 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0},
   { 0, 0, 0, 0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0}
 };
 
@@ -1878,24 +1928,59 @@ static int read_and_execute(bool interactive)
   ulong line_number=0;
   bool ml_comment= 0;  
   COMMANDS *com;
+  ulong line_length= 0;
   status.exit_status=1;
   
+  real_binary_mode= !interactive && opt_binary_mode;
   while (!aborted)
   {
     if (!interactive)
     {
-      line=batch_readline(status.line_buff);
       /*
-        Skip UTF8 Byte Order Marker (BOM) 0xEFBBBF.
-        Editors like "notepad" put this marker in
-        the very beginning of a text file when
-        you save the file using "Unicode UTF-8" format.
+        batch_readline can return 0 on EOF or error.
+        In that case, we need to double check that we have a valid
+        line before actually setting line_length to read_length.
       */
-      if (line && !line_number &&
-           (uchar) line[0] == 0xEF &&
-           (uchar) line[1] == 0xBB &&
-           (uchar) line[2] == 0xBF)
-        line+= 3;
+      line= batch_readline(status.line_buff, real_binary_mode);
+      if (line) 
+      {
+        line_length= status.line_buff->read_length;
+
+        /*
+          ASCII 0x00 is not allowed appearing in queries if it is not in binary
+          mode.
+        */
+        if (!real_binary_mode && strlen(line) != line_length)
+        {
+          status.exit_status= 1;
+          String msg;
+          msg.append("ASCII '\\0' appeared in the statement, but this is not "
+                     "allowed unless option --binary-mode is enabled and mysql is "
+                     "run in non-interactive mode. Set --binary-mode to 1 if ASCII "
+                     "'\\0' is expected. Query: '");
+          msg.append(glob_buffer);
+          msg.append(line);
+          msg.append("'.");
+          put_info(msg.c_ptr(), INFO_ERROR);
+          break;
+        }
+
+        /*
+          Skip UTF8 Byte Order Marker (BOM) 0xEFBBBF.
+          Editors like "notepad" put this marker in
+          the very beginning of a text file when
+          you save the file using "Unicode UTF-8" format.
+        */
+        if (!line_number &&
+             (uchar) line[0] == 0xEF &&
+             (uchar) line[1] == 0xBB &&
+             (uchar) line[2] == 0xBF)
+        {
+          line+= 3;
+          // decrease the line length accordingly to the 3 bytes chopped
+          line_length -=3;
+        }
+      }
       line_number++;
       if (!glob_buffer.length())
 	status.query_start_line=line_number;
@@ -1946,6 +2031,8 @@ static int read_and_execute(bool interactive)
       */
       if (opt_outfile && line)
 	fprintf(OUTFILE, "%s\n", line);
+
+      line_length= line ? strlen(line) : 0;
     }
     // End of file or system error
     if (!line)
@@ -1962,7 +2049,7 @@ static int read_and_execute(bool interactive)
       (We want to allow help, print and clear anywhere at line start
     */
     if ((named_cmds || glob_buffer.is_empty())
-	&& !ml_comment && !in_string && (com=find_command(line,0)))
+	&& !ml_comment && !in_string && (com= find_command(line)))
     {
       if ((*com->func)(&glob_buffer,line) > 0)
 	break;
@@ -1974,7 +2061,7 @@ static int read_and_execute(bool interactive)
 #endif
       continue;
     }
-    if (add_line(glob_buffer, line, &in_string, &ml_comment,
+    if (add_line(glob_buffer, line, line_length, &in_string, &ml_comment,
                  status.line_buff ? status.line_buff->truncated : 0))
       break;
   }
@@ -1996,70 +2083,131 @@ static int read_and_execute(bool interactive)
   tmpbuf.free();
 #endif
 
+  /*
+    If the function is called by 'source' command, it will return to interactive
+    mode, so real_binary_mode should be FALSE. Otherwise, it will exit the
+    program, it is safe to set real_binary_mode to FALSE.
+  */
+  real_binary_mode= FALSE;
   return status.exit_status;
 }
 
 
-static COMMANDS *find_command(char *name,char cmd_char)
+/**
+   It checks if the input is a short form command. It returns the command's
+   pointer if a command is found, else return NULL. Note that if binary-mode
+   is set, then only \C is searched for.
+
+   @param cmd_char    A character of one byte.
+
+   @return
+     the command's pointer or NULL.
+*/
+static COMMANDS *find_command(char cmd_char)
+{
+  DBUG_ENTER("find_command");
+  DBUG_PRINT("enter", ("cmd_char: %d", cmd_char));
+
+  int index= -1;
+
+  /*
+    In binary-mode, we disallow all mysql commands except '\C'
+    and DELIMITER.
+  */
+  if (real_binary_mode)
+  {
+    if (cmd_char == 'C')
+      index= charset_index;
+  }
+  else
+    index= get_command_index(cmd_char);
+
+  if (index >= 0)
+  {
+    DBUG_PRINT("exit",("found command: %s", commands[index].name));
+    DBUG_RETURN(&commands[index]);
+  }
+  else
+    DBUG_RETURN((COMMANDS *) 0);
+}
+
+/**
+   It checks if the input is a long form command. It returns the command's
+   pointer if a command is found, else return NULL. Note that if binary-mode 
+   is set, then only DELIMITER is searched for.
+
+   @param name    A string.
+   @return
+     the command's pointer or NULL.
+*/
+static COMMANDS *find_command(char *name)
 {
   uint len;
   char *end;
   DBUG_ENTER("find_command");
-  DBUG_PRINT("enter",("name: '%s'  char: %d", name ? name : "NULL", cmd_char));
 
-  if (!name)
+  DBUG_ASSERT(name != NULL);
+  DBUG_PRINT("enter", ("name: '%s'", name));
+
+  while (my_isspace(charset_info, *name))
+    name++;
+  /*
+    If there is an \\g in the row or if the row has a delimiter but
+    this is not a delimiter command, let add_line() take care of
+    parsing the row and calling find_command().
+  */
+  if ((!real_binary_mode && strstr(name, "\\g")) ||
+      (strstr(name, delimiter) &&
+       !is_delimiter_command(name, DELIMITER_NAME_LEN)))
+      DBUG_RETURN((COMMANDS *) 0);
+
+  if ((end=strcont(name, " \t")))
   {
-    len=0;
-    end=0;
+    len=(uint) (end - name);
+    while (my_isspace(charset_info, *end))
+      end++;
+    if (!*end)
+      end= 0;					// no arguments to function
+  }
+  else
+    len= (uint) strlen(name);
+
+  int index= -1;
+  if (real_binary_mode)
+  {
+    if (is_delimiter_command(name, len))
+      index= delimiter_index;
   }
   else
   {
-    while (my_isspace(charset_info,*name))
-      name++;
     /*
-      If there is an \\g in the row or if the row has a delimiter but
-      this is not a delimiter command, let add_line() take care of
-      parsing the row and calling find_command()
+      All commands are in the first part of commands array and have a function
+      to implement it.
     */
-    if (strstr(name, "\\g") || (strstr(name, delimiter) &&
-                                !(strlen(name) >= 9 &&
-                                  !my_strnncoll(&my_charset_latin1,
-                                                (uchar*) name, 9,
-                                                (const uchar*) "delimiter",
-                                                9))))
-      DBUG_RETURN((COMMANDS *) 0);
-    if ((end=strcont(name," \t")))
+    for (uint i= 0; commands[i].func; i++)
     {
-      len=(uint) (end - name);
-      while (my_isspace(charset_info,*end))
-	end++;
-      if (!*end)
-	end=0;					// no arguments to function
+      if (!my_strnncoll(&my_charset_latin1, (uchar*) name, len,
+                        (uchar*) commands[i].name, len) &&
+          (commands[i].name[len] == '\0') &&
+          (!end || commands[i].takes_params))
+      {
+        index= i;
+        break;
+      }
     }
-    else
-      len=(uint) strlen(name);
   }
 
-  for (uint i= 0; commands[i].name; i++)
+  if (index >= 0)
   {
-    if (commands[i].func &&
-        (((name &&
-           !my_strnncoll(&my_charset_latin1, (uchar*) name, len,
-                         (uchar*) commands[i].name, len) &&
-           !commands[i].name[len] &&
-           (!end || (end && commands[i].takes_params)))) ||
-         (!name && commands[i].cmd_char == cmd_char)))
-    {
-      DBUG_PRINT("exit",("found command: %s", commands[i].name));
-      DBUG_RETURN(&commands[i]);
-    }
+    DBUG_PRINT("exit", ("found command: %s", commands[index].name));
+    DBUG_RETURN(&commands[index]);
   }
   DBUG_RETURN((COMMANDS *) 0);
 }
 
 
-static bool add_line(String &buffer,char *line,char *in_string,
-                     bool *ml_comment, bool truncated)
+static bool add_line(String &buffer, char *line, ulong line_length,
+                     char *in_string, bool *ml_comment, bool truncated)
 {
   uchar inchar;
   char buff[80], *pos, *out;
@@ -2074,10 +2222,11 @@ static bool add_line(String &buffer,char *line,char *in_string,
   if (status.add_to_history && line[0] && not_in_history(line))
     add_history(line);
 #endif
-  char *end_of_line=line+(uint) strlen(line);
+  char *end_of_line= line + line_length;
 
-  for (pos=out=line ; (inchar= (uchar) *pos) ; pos++)
+  for (pos= out= line; pos < end_of_line; pos++)
   {
+    inchar= (uchar) *pos;
     if (!preserve_comments)
     {
       // Skip spaces at the beginning of a statement
@@ -2117,7 +2266,7 @@ static bool add_line(String &buffer,char *line,char *in_string,
 	*out++= (char) inchar;
 	continue;
       }
-      if ((com=find_command(NullS,(char) inchar)))
+      if ((com= find_command((char) inchar)))
       {
         // Flush previously accepted characters
         if (out != line)
@@ -2193,7 +2342,7 @@ static bool add_line(String &buffer,char *line,char *in_string,
 
       pos--;
 
-      if ((com= find_command(buffer.c_ptr(), 0)))
+      if ((com= find_command(buffer.c_ptr())))
       {
           
         if ((*com->func)(&buffer, buffer.c_ptr()) > 0)
@@ -2312,9 +2461,7 @@ static bool add_line(String &buffer,char *line,char *in_string,
   {
     uint length=(uint) (out-line);
 
-    if (!truncated && (length < 9 ||
-                       my_strnncoll (charset_info, (uchar *)line, 9,
-                                     (const uchar *) "delimiter", 9) ||
+    if (!truncated && (!is_delimiter_command(line, length) ||
                        (*in_string || *ml_comment)))
     {
       /* 
diff --git a/client/mysqlbinlog.cc b/client/mysqlbinlog.cc
index 387413cc5aa..d68b2a557e1 100644
--- a/client/mysqlbinlog.cc
+++ b/client/mysqlbinlog.cc
@@ -206,10 +206,8 @@ void print_annotate_event(PRINT_EVENT_INFO *print_event_info)
   }
 }
 
-static Exit_status dump_local_log_entries(PRINT_EVENT_INFO *print_event_info,
-                                          const char* logname);
-static Exit_status dump_remote_log_entries(PRINT_EVENT_INFO *print_event_info,
-                                           const char* logname);
+static Exit_status dump_local_log_entries(PRINT_EVENT_INFO *, const char*);
+static Exit_status dump_remote_log_entries(PRINT_EVENT_INFO *, const char*);
 static Exit_status dump_log_entries(const char* logname);
 static Exit_status safe_connect();
 
diff --git a/client/readline.cc b/client/readline.cc
index 791a044e0e1..b6643b86356 100644
--- a/client/readline.cc
+++ b/client/readline.cc
@@ -54,7 +54,7 @@ LINE_BUFFER *batch_readline_init(ulong max_size,FILE *file)
 }
 
 
-char *batch_readline(LINE_BUFFER *line_buff)
+char *batch_readline(LINE_BUFFER *line_buff, bool binary_mode)
 {
   char *pos;
   ulong out_length;
@@ -63,8 +63,17 @@ char *batch_readline(LINE_BUFFER *line_buff)
   if (!(pos=intern_read_line(line_buff, &out_length)))
     return 0;
   if (out_length && pos[out_length-1] == '\n')
-    if (--out_length && pos[out_length-1] == '\r')  /* Remove '\n' */
-      out_length--;                                 /* Remove '\r' */
+  {
+    /*
+      On Windows platforms we also need to remove '\r', unconditionally.  On
+      Unix-like platforms we only remove it if we are not on binary mode.
+     */
+
+    /* Remove '\n' */
+    if (--out_length && IF_WIN(1,!binary_mode) && pos[out_length-1] == '\r')
+      /* Remove '\r' */
+      out_length--;                                 
+  }
   line_buff->read_length=out_length;
   pos[out_length]=0;
   return pos;
@@ -226,7 +235,7 @@ char *intern_read_line(LINE_BUFFER *buffer, ulong *out_length)
   for (;;)
   {
     pos=buffer->end_of_line;
-    while (*pos != '\n' && *pos)
+    while (*pos != '\n' && pos != buffer->end)
       pos++;
     if (pos == buffer->end)
     {