19.2 Functions

my_ulonglong mysql_affected_rows(MYSQL *mysql)

Returns the number of rows affected by the most recent query. When used with a non- SELECT query, it can be used after the mysql_query call that sent the query. With SELECT, this function is identical to mysql_num_rows. This function returns 0, as expected, for queries that affect or return no rows. In the case of an error, the function returns -1.

When an UPDATE query causes no change in the value of a row, that row is not usually considered to be affected. However, if the CLIENT_FOUND_ROWS flag is set when connecting to the MySQL server (see the mysql_real_connect function), any rows that match the WHERE clause of the UPDATE query are considered affected.

Example

/* Insert a row into the people table */
mysql_query(&mysql, "INSERT INTO people VALUES ('', 'Illyana Rasputin', 16)";
num = mysql_affected_rows(&mysql);
/* num should be 1 if the INSERT (of a single row) was successful, and -1 if
   there was an error */
  
/* Make any of 'HR', 'hr', 'Hr', or 'hR' into 'HR'. This is an easy way to
   force a consistent capitalization in a field.
mysql_query(&mysql, "UPDATE people SET dept = 'HR' WHERE dept LIKE 'HR'");
affected = mysql_affected_rows(&mysql);
/* By default, 'affected' will contain the number of rows that were changed.
   That is, the number of rows that had a dept value of 'hr', 'Hr' or 'hR'.
   If the CLIENT_FOUND_ROWS flag was used, 'affected' will contain the number
   of rows that matched the where. */

my_bool mysql_change_user(MYSQL *mysql, char *username, char *password,
                          char *database)

Changes the currently authenticated user and database. This function reauthenticates the current connection using the given username and password. It also changes the default database to the given database (which can be NULL if no default is desired). If the password is incorrect for the given username, or if the new user does not have rights to access the given database, a false value is returned, and no action is taken. Otherwise, the rights of the new user take effect, the default database is selected, and a true value is returned.

Example

if (! mysql_change_user( &mysql, new_user, new_pass, new_db ) ) {
      printf("Change of User unsuccessful!");
      exit(1);
}
/* At this point, the connection is operating under the access rights of the
   new username, and the new database is the default. */

char *mysql_character_set_name(MYSQL *mysql)

Returns the name of the default character set used by the MySQL server. A generic installation of the MySQL source uses the ISO-8859-1 character set by default.

Example

printf("This server uses the %s character set by default\n",
         mysql_character_set_name(&mysql));

void mysql_close(MYSQL *mysql)

Ends a connection to the database server. If there is a problem when the connection is broken, the error can be retrieved from the mysql_err function.

Example

mysql_close(&mysql);
/* The connection should now be terminated */

MYSQL *mysql_connect(MYSQL *mysql, const char *host, const char *user,
                      const char *passwd) 

Creates a connection to a MySQL database server. The first parameter must be a predeclared MYSQL structure. The second parameter is the hostname or IP address of the MySQL server. If the host is an empty string or localhost, a connection is made to the MySQL server on the same machine. The final two parameters are the username and password used to make the connection. The password should be entered as plain text, not encrypted in any way. The return value is the MYSQL structure that was passed as the first argument, or NULL if the connection failed. (Because the structure is contained as an argument, the only use for the return value is to check if the connection succeeded.)

This function has been deprecated in the newer releases of MySQL, and the mysql_real_connect function should be used instead.

Example

/* Create a connection to the local MySQL server using the name "bob" and
   password "mypass" */
MYSQL mysql;
if(!mysql_connect(&mysql, "", "bob", "mypass")) {
                  printf("Connection error!\n");
                  exit(0);
}
/* If we've reached this point we have successfully connected to the database
   server. */

int mysql_create_db(MYSQL *mysql, const char *db)

Creates an entirely new database with the given name. The return value is 0 if the operation was successful and nonzero if there was an error.

This function has been deprecated in the newer releases of MySQL. MySQL now supports the CREATE DATABASE SQL statement. This should be used, via the mysql_query function, instead.

Example

/* Create the database 'new_database' */
result = mysql_create_db(&mysql, "new_database");

void mysql_data_seek(MYSQL_RES *res, unsigned int offset)

Moves to a specific row in a group of results. The first argument is the MYSQL_RES structure that contains the results. The second argument is the row number you wish to seek to, starting from 0. This function works only if the data was retrieved using mysql_store_result, because datasets retrieved with mysql_use_result are not guaranteed to be complete.

Example

/* Jump to the last row of the results */
mysql_data_seek(results, mysql_num_rows(results)-1);

mysql_debug(char *debug)

Manipulates the debugging functions if the client was compiled with debugging enabled. MySQL uses the Fred Fish debugging library, which has far too many features and options to detail here.

Example

/* This is a common use of the debugging library. It keeps a trace of the
   client program's activity in the file "debug.out" */
mysql_debug("d:t:O,debug.out");

int mysql_drop_db(MYSQL *mysql, const char *db)

Destroys the database with the given name. The return value is 0 if the operation was successful and nonzero if there was an error.

This function has been deprecated in the newer releases of MySQL. MySQL now supports the DROP DATABASE SQL statement. This should be used, via the mysql_query function, instead.

Example

/* Destroy the database 'old_database' */
result = mysql_drop_db(&mysql, "old_database");

int mysql_dump_debug_info(MYSQL *mysql)

This function causes the database server to enter debugging information about the current connection into its logs. You must have Process privilege in the current connection to use this function. The return value is 0 if the operation succeeded and nonzero if there was an error.

Example

result = mysql_dump_debug_info(&mysql);
/* The server's logs should now contain information about this connection. 
   If something went wrong so that this is not the case, 'result' will have
   a false value.*/

my_bool mysql_eof(MYSQL_RES *result)

Returns a nonzero value if there is no more data in the group of results being examined. If there is an error in the result set, 0 is returned. This function works only if the result set was retrieved with the mysql_use_result function (mysql_store_result retrieves the entire result set, making a call to mysql_eof unnecessary).

This function has been deprecated in the newer releases of MySQL. The mysql_errno and mysql_error functions return more information about any errors that occur, and they are more reliable.

Example

/* Read through the results until no more data comes out */
while((row = mysql_fetch_row(results))) {
       /* Do work */
}
   
if(!mysql_eof(results))
   printf("Error. End of results not reached.\n");

unsigned int mysql_errno(MYSQL *mysql) 

Returns the error number of the last error associated with the current connection. If there have been no errors in the connection, the function returns 0. The actual text of the error can be retrieved using the mysql_error function. The defined names for the client errors can be found in the errmsg.h header file. The defined names for the server error can be found in the mysqld_error.h header file.

Example

error = mysql_errno(&mysql);
printf("The last error was number %d\n", error);

char *mysql_error(MYSQL *mysql)

Returns the error message of the last error associated with the current connection. If there have been no errors in the connection, the function returns an empty string. Error messages originating on the server will always be in the language used by the server (chosen at startup time with the -- language option). The language of the client error messages can be chosen when compiling the client library. At the time of this writing, MySQL supports the following languages: Czech, Danish, Dutch, English, Estonian, French, German, Greek, Hungarian, Italian, Japanese, Korean, Norwegian (standard and ny), Polish, Portuguese, Romanian, Russian, Slovak, Spanish, and Swedish.

Example

printf("The last error was '%s'\n", mysql_error(&mysql));

unsigned int mysql_escape_string(char *to, const char *from, unsigned int length)

Encodes a string so that it is safe to insert into a MySQL table. The first argument is the receiving string, which must be at least one character greater than twice the length of the second argument: the original string. (That is, to must be greater then or equal to from*2+1.) The third argument indicates the number of bytes to be copied from the original string and encoded. The function returns the number of bytes in the encoded string, not including the terminating null character.

While not officially deprecated, this function is generally inferior to the mysql_real_escape_string function, which does everything this function does, but also takes into account the character set of the current connection, which may affect certain escape sequences.

Example

char name[15] = "Bob Marley's";
char enc_name[31];
mysql_escape_string(enc_name, name);
/* enc_name will now contain "Bob Marley\'s" (the single quote is escaped).

MYSQL_FIELD *mysql_fetch_field(MYSQL_RES *result)

Returns a MYSQL_FIELD structure describing the current field of the given result set. Repeated calls to this function return information about each field in the result set in turn until there are no more fields left. It then returns a null value.

Example

MYSQL_FIELD *field;
 
while((field = mysql_fetch_field(results)))
{
    /* You can examine the field information here */
}

MYSQL_FIELD * mysql_fetch_field_direct(MYSQL_RES * result, unsigned int fieldno)

This function is the same as mysql_fetch_field, except that you specify which field you wish to examine, instead of cycling through the fields. The first field in a result set is 0.

Example

MYSQL_FIELD *field;
 
/* Retrieve the third field in the result set for examination */
field = mysql_fetch_field_direct(results, 2);

MYSQL_FIELD *mysql_fetch_fields(MYSQL_RES * result)

This function is the same as mysql_fetch_field, except that it returns an array of MYSQL_FIELD structures containing the information for every field in the result set.

Example

MYSQL_FIELD *field; /* A pointer to a single field */
MYSQL_FIELD *fields; /* A pointer to an array of fields */
 
/* Retrieve all the field information for the results */
fields = mysql_fetch_fields(results);
/* Assign the third field to 'field' */
field = fields[2];

unsigned long *mysql_fetch_lengths(MYSQL_RES *result)

Returns an array of the lengths of each field in the current row. A null value is returned in the case of an error. You must have fetched at least one row (with mysql_fetch_row) before you can call this function. This function provides the only way to determine the lengths of variable-length fields, such as BLOB and VARCHAR, before you use the data.

This function is especially useful when reading binary data from a BLOB. Since all MySQL data is retrieved as strings (char *), it is common to use the strlen( ) function to determine the length of a data value. However, for binary data, strlen( ) returns inaccurate results, because it stops at the first null character. In these cases, you can use mysql_fetch_lengths to retrieve the accurate length for a data value.

Example

unsigned long *lengths;
 
row = mysql_fetch_row(results);
lengths = mysql_fetch_lengths(results);
printf("The third field is %d bytes long\n", lengths[2]);

MYSQL_ROW mysql_fetch_row(MYSQL_RES *result)

Retrieves the next row of the result and returns it as a MYSQL_ROW structure. A null value is returned if there are no more rows or if there is an error. In the current implementation, the MYSQL_ROW structure is an array of character strings that can be used to represent any data. If a data element is NULL within the database, the MYSQL_ROW array element for that data element is a null pointer. This is necessary to distinguish between a value that is NULL and a value that is simply an empty string (which is returned as a non-null pointer to a null value).

Example

MYSQL_ROW row;
 
row = mysql_fetch_row(results);
printf("The data in the third field of this row is: %s\n", row[2]);

unsigned int mysql_field_count(MYSQL *mysql)

Returns the number of columns contained in a result set. This function is most useful when checking the type of query last executed. If a call to mysql_store_result returns a null pointer for a result set, either the query was a non-SELECT query (such as UPDATE, INSERT, etc.) or there was an error. By calling mysql_field_count, you can determine which was the case, because a non-SELECT query always returns zero fields, and a SELECT query always returns at least one field. Even if the returned query has no rows, the fields of the SELECT query will be reflected here.

Example

MYSQL_FIELD field;
MYSQL_RES *result;
 
// A query has been executed and returned success
result = mysql_store_result(  );
if (! result ) {
     // Ooops, the result pointer is null, either the query was a non-SELECT
     // query or something bad happened!
     if ( mysql_field_count(&mysql) ) {
          // The number of columns queried is greater than zero, it must have
          // been a SELECT query and an error must have occurred.
     } else {
          // Since the number of columns queried is zero, it must have been
          // a non-SELECT query, so all is well...
     }
}

MYSQL_FIELD_OFFSET mysql_field_seek(MYSQL_RES *result, MYSQL_ FIELD_OFFSET offset)

Moves the internal pointer of a result set to the given field of the current row. The position set by this function is used when mysql_fetch_field is called. The MYSQL_FIELD_OFFSET value passed should be the return value of a mysql_field_tell call (or another mysql_field_seek). If the value passed is 0, the function seeks to the beginning of the row. The return value is the position of the cursor before the function was called.

Example

MYSQL_FIELD field;
 
/* result is a MYSQL_RES structure containing a result set */
/* ... do some stuff */
/* Seek back to the beginning of the row */
old_pos = mysql_field_seek(results, 0);
/* Fetch the first field of the row */
field = mysql_fetch_field(results);
/* Go back to where you were */
mysql_field_seek(results, old_pos);

MYSQL_FIELD_OFFSET mysql_field_tell(MYSQL_RES *result)

Returns the value of the current field position within the current row of the result set. This value is used with mysql_field_seek.

Example

MYSQL_FIELD field1, field2, field3;
/* results is a MYSQL_RES structure containing a result set */
 
/* Record my current position */
old_pos = mysql_field_tell(results);
/* Fetch three more fields */
field1 = mysql_fetch_field(results);
field2 = mysql_fetch_field(results);
field3 = mysql_fetch_field(results);
/* Go back to where you where */
mysql_field_seek(results, old_pos);

void mysql_free_result(MYSQL_RES *result) 

Frees the memory associated with a MYSQL_RES structure. This must be called whenever you are finished using this type of structure, or else memory problems will occur. This should be used only on a pointer to an actual MYSQL_RES structure. For example, if a call to mysql_store_result returns a null pointer, this function should not be used.

Example

MYSQL_RES *results;
/* Do work with results */
/* free results... we know it's not null since we just did work with
   it, but we'll check just to be safe. */
if (results)
    mysql_free_result(results);

char *mysql_get_client_info(void)

Returns a string containing the MySQL library version used by the client program.

Example

printf("This program uses MySQL client library version %s\n", 
       mysql_get_client_info(  )));

char *mysql_get_host_info(MYSQL *mysql) 

Returns a string containing the hostname of the MySQL database server and the type of connection used (e.g., Unix socket or TCP).

Example

printf("Connection info: %s", mysql_get_host_info(&mysql));

unsigned int mysql_get_proto_info(MYSQL *mysql)

Returns the MySQL protocol version used in the current connection as an integer. As a general rule, the MySQL network protocol changes only between minor releases of MySQL. That is, all releases of MySQL 3.23.x should have the same protocol version number.

Example

printf("This connection is using MySQL connection protocol ver. %d\n",
        mysql_get_proto_info(  ));

char *mysql_get_server_info(MYSQL *mysql)

Returns a string containing the version number of the MySQL database server used by the current connection.

Example

printf("You are currently connected to MySQL server version %s\n",
        mysql_get_server_info(&mysql);

char *mysql_info(MYSQL *mysql)

Returns a string containing information about the most recent query, if the query was of a certain type. Currently, the following SQL queries supply extra information via this function: INSERT INTO (when used with a SELECT clause or a VALUES clause with more than one record), LOAD DATA INFILE, ALTER TABLE, and UPDATE. If the most recent query provided no additional information (e.g., it was not one of the above queries), this function returns a null value.

The format of the returned string depends on the query. If the query is INSERT INTO or ALTER TABLE, the string is:

Records: n Duplicates: n Warnings: n

If the query is LOAD DATA INFILE, the string is:

Records: n Deleted: n Skipped: n Warnings: n

If the query is UPDATE, the string is:

Rows matched: n Changed: n Warnings: n

Example

/* We just sent LOAD DATA INFILE query reading a set of record from a file into
   an existing table */
printf("Results of data load: %s\n", mysql_info(&mysql));
/* The printed string looks like this:
Records: 30 Deleted: 0 Skipped: 0 Warnings: 0
*/

MYSQL *mysql_init(MYSQL *mysql)

Initializes a MYSQL structure for creating a connection to a MySQL database server. A call to this function, followed by one to mysql_real_connect, is currently the approved method of initializing a server connection. You can pass mysql_init either a pointer to a MYSQL structure that you declared or a null pointer. If you pass your own pointer, you are responsible for freeing its data when the time comes (which must be after the connection is closed).

If you pass a null pointer, mysql_init creates a MYSQL structure, initializes it, and returns a pointer to the structure. Structures created by this function are freed automatically when mysql_close is called. A null value is returned if there is not enough memory available to initialize the structure.

As of the current release of MySQL, MySQL clients will crash on certain platforms (such as SCO Unix) when you pass in a pointer to a MYSQL structure that you allocated yourself. If this is happening to you, just pass in NULL and use the pointer created by the MySQL library. As a bonus, you don't have to worry about freeing the structure if you do this.

Example

MYSQL mysql;
 
if (!mysql_init(&mysql)) {
                printf("Error initializing MySQL client\n");
                exit(1);
}
/* Now you can call mysql_real_connect(  ) to connect to a server... */
/* Alternative method: */
MYSQL *mysql;

mysql = mysql_init(NULL);
if (!mysql) {
     printf("Error initializing MySQL client\n");
     exit(1);
}

my_ulonglong mysql_insert_id(MYSQL *mysql)

Returns the generated value for an AUTO_INCREMENT field if the last query created a new row. This function is usually called immediately after a value has been inserted into an AUTO_INCREMENT field to determine the value that was inserted. This value is reset to 0 after any query that does not insert a new auto-increment row.

The MySQL-specific SQL function LAST_INSERT_ID( ) also returns the value of the most recent auto-increment. In addition, it is not reset after each query, so it can be called at any time to retrieve that value of the last auto-increment INSERT executed during the current session.

Example

/* We just inserted an employee record with automatically generated ID into
   a table */
id = mysql_insert_id(&mysql);
printf("The new employee has ID %d\n", id);
/* As soon as we run another query, mysql_insert_id will return 0 */

int mysql_kill(MYSQL *mysql, unsigned long pid) 

Attempts to kill the MySQL server thread with the specified process ID. This function returns 0 if the operation was successful and nonzero on failure. You must have Process privilege in the current connection to use this function.

The process IDs are part of the process information returned by the mysql_list_processes function.

Example

/* Kill thread 4 */
result = mysql_kill(&mysql, 4);

MYSQL_RES *mysql_list_dbs(MYSQL *mysql, const char *wild)

Returns a MYSQL_RES structure containing the names of all existing databases that match the pattern given by the second argument. This argument may be any standard SQL regular expression. If a null pointer is passed instead, all databases are listed. Like all MYSQL_RES structures, the return value of this function must be freed with mysql_free_result. This function returns a null value in the case of an error.

The information obtained from this function can also be obtained through an SQL query using the statement SHOW databases.

Example

MYSQL_RES databases;
databases = mysql_list_dbs(&mysql, (char *)NULL);
/* 'databases' now contains the names of all of the databases in the
   MySQL server */
/* ... */
mysql_free_result( databases );
/* Find all databases that start with 'projectName' */
databases = mysql_list_dbs(&mysql, "projectName%");

MYSQL_RES *mysql_list_fields(MYSQL *mysql, const char *table, const char *wild)

Returns a MYSQL_RES structure containing the names of all existing fields in the given table that match the pattern given by the third argument. This argument may be any standard SQL regular expression. If a null pointer is passed instead, all fields are listed. Like all MYSQL_RES structures, the return value of this function must be freed with mysql_free_result. This function returns a null value in the case of an error.

The information obtained from this function can also be obtained through an SQL query using the statement SHOW COLUMNS FROM table.

Example

MYSQL_RES fields;
fields = mysql_list_fields(&mysql, "people", "address%");
/* 'fields' now contains the names of all fields in the 'people' table
    that start with 'address' */
/* ... */
mysql_free_result( fields );

MYSQL_RES *mysql_list_processes(MYSQL *mysql)

Returns a MYSQL_RES structure containing information on all the threads currently running on the MySQL database server. (The term processes in the function name refers to the internal MySQL processes, or threads, not to any operating system-level process.) The information contained in the structure can be used with mysql_kill to remove faulty threads. Like all MYSQL_RES structures, the return value of this function must be freed with mysql_free_result. This function returns a null value in the case of an error.

The returned result set contains the information in the following order:

The MySQL process ID

This is the ID used with mysql_kill to kill a thread.

Username

The MySQL username of the user executing a thread.

Hostname

The location of the client running a thread.

Database

The current database for the client running a thread.

Action

The type of action last run in a thread. All SQL queries of any type show up as Query, so this will be the most common value.

Time

The amount of time taken (in seconds) to execute the last action in a thread.

State

The state of the current thread. This indicates whether the thread is active (currently executing a command) or idle.

Info

Any extra information about the thread. For SQL queries, this will contain the text of the query.

The information obtained from this function can also be obtained through an SQL query using the statement SHOW PROCESSLIST.

Example

MYSQL_RES *threads;
MYSQL_ROW row
threads = mysql_list_processes(&mysql);
 
row = mysql_fetch_row( threads );
printf("The ID of the first active thread is %d\n", row[0]);

MYSQL_RES *mysql_list_tables(MYSQL *mysql, const char *wild)

Returns a MYSQL_RES structure containing the names of all existing tables in the current database that match the pattern given by the second argument. This argument may be any standard SQL regular expression. If a null pointer is passed, all tables are listed. Like all MYSQL_RES structures, the return value of this function must be freed with mysql_free_result. This function returns a null value in the case of an error.

The information obtained from this function can also be obtained through an SQL query using the statement SHOW TABLES.

Example

MYSQL_RES tables;
tables = mysql_list_tables(&mysql, "p%");
/* 'tables' now contains the names of all tables in the current database
    that start with 'p' */

unsigned int mysql_num_fields(MYSQL_RES *result)

Returns the number of fields contained in each row of the given result set. This is different from the mysql_field_count function, in that it operates on an actual result set, which is known to contain data, whereas mysql_field_count checks the last executed query (usually to determine if an error occurred).

Example

/* 'results' is a MYSQL_RES result set structure */
num_fields = mysql_num_fields(results);
printf("There are %d fields in each row\n", num_fields);

int mysql_num_rows(MYSQL_RES *result)

Returns the number of rows of data in the result set. This function is accurate only if the result set was retrieved with mysql_store_result. If mysql_use_result was used, the value returned by this function is the number of rows accessed so far.

Example

/* 'results' is a MYSQL_RES result set structure */
num_rows = mysql_num_rows(results);
printf("There were %d rows returned, that I know about\n", num_rows);

char *mysql_odbc_escape_string(MYSQL *mysql, char *result_string, unsigned long 
result_string_length, char *original_string, unsigned long original_string_length, 
void *parameters, char *(*extend_buffer))

Creates a properly escaped SQL query string from a given string. This function is intended for use with ODBC clients, and mysql_real_escape_string provides the same functionality with a simpler interface. This function takes the string given in the fourth argument (with the length given in the fifth argument, not including the terminating null character) and escapes it so that the resulting string (which is put into the address given in the second argument, with a maximum length given in the third argument) is safe to use as a MySQL SQL statement. This function returns a copy of the result string (the second argument). The seventh argument must be a pointer to a function that can be used to allocate memory for the result string. The function must take three arguments: a pointer to a set of parameters that control how the memory is allocated (these parameters are passed in as the sixth argument to the original function), a pointer to the result string, and a pointer to the maximum length of the result string.

Example

char *data = "\000\002\001";
int data_length = 3;
char *result;
int result_length = 5; /* We don't want the final string to be longer than 5.
 extend_buffer(  ) is a function that meets the criteria given above. */
mysql_odbc_escape_string( &mysql, result, result_length, data, data_length,
   NULL, extend_buffer );
/* 'result' now contains the string '\\\000\002\001'
 (that is, a backslash, followed by ASCII 0, then ASCII 2 then ASCII 1. */

void mysql_odbc_remove_escape(MYSQL *mysql, char *string )

Removes escape characters from a string. This function is intended for ODBC drivers themselves, and not for general use. Given a string, this function removes the escape character (\) preceding any other escape characters. This modifies (and thus shortens) the original string that was passed in.

Example

char *escaped = "\\'an escaped quoted string.\\'";
/* escaped contains the string: \' and escaped quoted string.\' */
mysql_odbc_remove_escape(&mysql, escaped);
/* escaped now contains the string: 'an escaped quoted string.' */

int mysql_options(MYSQL *mysql, enum mysql_option option, void *value)

Sets a connect option for an upcoming MySQL connection. This function must be called after a MYSQL structure has been initialized using mysql_init and before a connection has actually been established using mysql_real_connect. This function can be called multiple times to set more than one option. The options affect the upcoming connection. The value of the third argument depends on the type of option. Some options require a character string as an argument, while others take a pointer to an integer, or nothing at all. The options are as follows (the type of the third argument is given in parentheses after the option name):

MYSQL_INIT_COMMAND (char *)

Specifies an SQL query to execute as soon as the connection is established. This query is reexecuted if the connection is lost and automatically reconnected.

MYSQL_OPT_COMPRESS ()

Causes the connection to use a compressed protocol with the server to increase speed.

MYSQL_OPT_CONNECT_TIMEOUT (unsigned int *)

Specifies the number of seconds to wait before giving up on connecting to the server.

MYSQL_OPT_NAMED_PIPE ()

Causes the connection to use named pipes, as opposed to TCP, to connect to a local MySQL server running on Windows NT.

MYSQL_READ_DEFAULT_FILE (char *)

Specifies the name of the file to read for default options, in place of the default file my.cnf.

MYSQL_READ_DEFAULT_GROUP (char *)

Specifies the name of a section within the configuration file to read for the connection options, in place of the default client. See Chapter 4 for information about the configuration file options.

Example

MYSQL mysql;
mysql_init( &mysql );
/* Prepare this connection to use the compressed protocol, execute the
   query "SHOW tables" upon connection, and read addition options from the
   'startup' stanze in the file .mysqlrc */
mysql_options(&mysql, MYSQL_OPT_COMPRESS, 0 );
mysql_options(&mysql, MYSQL_INIT_COMMAND, "SHOW tables" );
mysql_options(&mysql, MYSQL_READ_DEFAULT_FILE, ".mysqlrc" );
mysql_options(&mysql, MYSQL_READ_DEFAULT_GROUP, "startup" );
/* Now it is time to call mysql_real_connect(  ) to make the connection using
   these options */

int mysql_ping(MYSQL *mysql)

Checks to see if the connection to the MySQL server is still alive. If it is not, the client attempts to reconnect automatically. This function returns if the connection is alive and nonzero if it cannot successfully contact the server.

Example

while(mysql_ping(&mysql)) printf("Error, attempting reconnection...\n");

int mysql_query(MYSQL *mysql, const char *query)

Executes the SQL query given in the second argument. If the query contains any binary data (particularly the null character), this function cannot be used, and mysql_real_query should be used instead. The function returns 0 if the query was successful and nonzero in the case of an error.

Once a query has been executed using this function, the result set can be retrieved using the mysql_store_result or mysql_use_result function.

Example

error = mysql_query(&mysql, "SELECT * FROM people WHERE name like 'Bill%'");
if (error) { 
     printf("Error with query!\n"); 
     exit(1);
}

int mysql_read_query_result(MYSQL *mysql)

Processes the result of a query executed with the mysql_send_query command. Any data processed this way is not returned. Therefore, this function is useful only when running non-SELECT queries or for debugging (because the return value still accurately reports if there was an error). The function returns 0 on success and a nonzero number if an error occurred.

Example

mysql_send_query(&mysql, "SELECT * INTO OUTFILE results.out FROM mytable");
/* This executes the query, but does not process the results, which is necessary
   in order to write the values into the outfile */
mysql_read_query_result(&mysql);
/* Now the results have been processed and the data written to the outfile */

MYSQL *mysql_real_connect(MYSQL *mysql, const char *host, const char *user,
                          const char *passwd, const char *db, uint port,
                          const char *unix_socket, uint client_flag)

Creates a connection with a MySQL database server. There are eight arguments to this function:

• An initialized MYSQL structure, created with mysql_init.

• The hostname or IP address of the MySQL database server (use an empty string or localhost to connect to the local MySQL server over a Unix socket).

• The username used to connect to the database server (an empty string indicates that the Unix login name of the person running the client should be used).

• The password used to authenticate the given user. If an empty string is used, only users with no passwords are checked for authentication. That is, if a user happened to have a password set to an empty string, he would never be authenticated.

• The initial database selected when you connect (an empty string indicates that no database should be selected).

• The port used to remotely connect to a MySQL database server over TCP. (0 may be used to accept the default port).

• The filename of the Unix socket used to connect to a MySQL server on the local machine (an empty string may be used to accept the default socket).

• Zero or more of a set of flags used under special circumstances:

  CLIENT_FOUND_ROWS

  When using queries that change tables, returns the number of rows found in the table, not the number of rows affected.

  CLIENT_IGNORE_SPACE

  Allows spaces after built-in MySQL functions in SQL queries. Traditionally, functions must be followed immediately by their arguments in parentheses. If this option is used, the function names become reserved words and cannot be used for names of tables, columns, or databases.

  CLIENT_INTERACTIVE

  Causes the server to wait for a longer amount of time (specified by the interactive_timeout server variable) before automatically breaking a connection. This is useful for interactive clients that do not allow the user to enter any data for significant periods of time.

  CLIENT_NO_SCHEMA

  Prevents the client from using the full database.table.column form to specify a column from any database.

  CLIENT_COMPRESS

  Causes the client to use compression when communicating with the server.

  CLIENT_ODBC

  Tells the server the client is an ODBC connection.

  CLIENT_SSL

  Tells the client to use SSL encryption to secure the connection. The server must have been compiled to support SSL.

Example

MYSQL *mysql;
mysql = mysql_init( NULL );
/* Connect to the server on the local host with standard options. */
if (! mysql_real_connect(&mysql, "localhost", "bob", "mypass", "", 0, "", 0))
{ print "Error connecting!\n";
  exit(1);
}
 
/* or... */
/* Connect to the server at my.server.com using a compressed, secure protocol */
if (! mysql_real_connect(&mysql, "my.server.com", "bob", "mypass",
                          "", 0, "", CLIENT_COMPRESS|CLIENT_SSL)) {
          print "Error connecting!\n";
          exit(1);
}

unsigned long mysql_real_escape_string(MYSQL *mysql, char *result_string,
                                       char *original_string,
                                       unsigned long orginal_string_length)

Creates a properly escaped SQL query string from a given string. This function takes the string given in the third argument (with the length given in the fourth argument, not including the terminating null character) and escapes it so that the resulting string (which is put into the address given in the second argument) is safe to use as a MySQL SQL statement. This function returns the new length of the resulting string (not including the terminating null character). To be completely safe, the space allocated for the result string should be at least twice as big as the original string (in case each character has to be escaped) plus one (for the terminating null character).

This function is safe to use with binary data. The string can contain null characters or any other binary data. This is why it is necessary to include the length of the string. Otherwise, the MySQL library could not determine how long the string was if any null characters were present.

Example

# Properly escape a query that contains binary data.
char *data = "\002\001\000";
int original_length = 4 # 3 characters plus one for the null. 
char real_data[7]; # Twice as big as the original string (3) 
                   # plus one for the null.
int new_length;
 
new_length = mysql_real_escape_string(&mysql, data, real_data, original_length);
/* real_query can now be safely used in as a SQL query. */
/* The returned length is '4' since the only character that needed escaping
   was \000 (the null character) */

int mysql_real_query(MYSQL *mysql, const char *query, unsigned int length)

Executes the SQL query given in the second argument. The length of the query (not including any terminating null character) must be given in the third argument. By supplying the length, you can use binary data, including null characters, in the query. This function is also faster than mysql_query. The function returns 0 if the query was successful and nonzero in the case of an error.

Once a query has been executed using this function, the result set can be retrieved using the mysql_store_result or mysql_use_result function.

Example

error = mysql_real_query(&mysql, "SELECT * FROM people WHERE name like 'Bill%'",
        44);
if (error) { 
     printf("Error with query!\n"); 
     exit(1);
}

int mysql_refresh(MYSQL *mysql, unsigned int options)

Instructs the server to refresh various system operations. Which operations are refreshed depends on the second argument, which is a bitwise combination of any of the following options:

REFRESH_GRANT

Reloads the permissions tables. This is the same as mysql_reload( ) or the FLUSH PRIVILEGES SQL command.

REFRESH_LOG

Flushes the log files to disk and then closes and reopens them.

REFRESH_TABLES

Flushes any open tables to disk.

REFRESH_READ_LOCK

Reinstates the read lock on the database files stored on disk.

REFRESH_HOSTS

Reloads the internal cache MySQL keeps of known hostnames to limit DNS lookup calls.

REFRESH_STATUS

Refreshes the status of the server by checking the status of various internal operations.

REFRESH_THREADS

Clears out any dead or inactive threads from the internal cache of threads.

REFRESH_MASTER

Flushes, closes, and reopens the binary log that tracks all changes to MySQL tables. This log is sent to all slave servers for replication.

REFRESH_SLAVE

Resets the connection with any slave servers that are replicating the master MySQL server.

Example

/* Flush the log files and the table data to disk */
if (!mysql_refresh( &mysql, REFRESH_LOG|REFRESH_TABLES )) {
     printf("Error sending refresh command...\n");
}

int mysql_reload(MYSQL *mysql)

Reloads the permission tables on the MySQL database server. You must have Reload permission on the current connection to use this function. If the operation is successful, 0 is returned; otherwise, a nonzero value is returned.

This function is deprecated and will be removed in a future version of the API. The same functionality can be obtained by using the SQL query FLUSH PRIVILEGES.

Example

/* Make some changes to the grant tables... */
result = mysql_reload(&mysql);
/* The changes now take effect... */

MYSQL_ROW_OFFSET mysql_row_seek(MYSQL_RES *result, MYSQL_ROW_OFFSET offset)

Moves the pointer for a result set (MYSQL_RES structure) to a specific row and returns the original row number. This function requires that the offset be an actual MYSQL_ROW_OFFSET structure, not a simple row number. If you have only a row number, use the mysql_data_seek function. You can obtain a MYSQL_ROW_OFFSET structure from a call to either mysql_row_tell or mysql_row_seek.

This function is useful only if the result set contains all the data from the query. Therefore, it should be used in conjunction with mysql_store_result and not used with mysql_use_result.

Example

/* result is a result set pointer created with mysql_store_result(  ) */
MYSQL_ROW_OFFSET where, other_place;
where = mysql_row_tell( result );
/* Do some more work with the result set... */
/* Go back to where you were before, but remember where you are now: */
other_place = mysql_row_seek( result, where );
/* Do some more work... */
/* Go back to the second marker: */
mysql_row_seek( result, other_place );

MYSQL_ROW_OFFSET mysql_row_tell(MYSQL_RES *result)

Returns the value of the cursor as mysql_fetch_row reads the rows of a result set. The return value of this function can be used with mysql_row_seek to jump to a specific row in the result set. Values from the function are useful only if the result set was created with mysql_store_result (which contains all the data) and not mysql_use_result (which does not).

Example

/* results is a result set pointer created with mysql_store_result(  ) */
MYSQL_ROW_OFFSET saved_pos = mysql_row_tell(results);
/* I can now jump back to this row at any time using mysql_row_seek(  ) */

int mysql_select_db(MYSQL *mysql, const char *db)

Changes the current database. The user must have permission to access the new database. The function returns 0 if the operation was successful and nonzero in the case of an error.

Example

if ( mysql_select_db(&mysql, "newdb") ) {
     printf("Error changing database, you probably don't have permission.\n");
}

int mysql_send_query(MYSQL *mysql, char *query, unsigned int query_length)

Executes a single MySQL query, without providing any results to the user. This function is useful for quickly executing a non-SELECT statement that does not return any data to the user. This function is also useful for debugging, since the return value still accurately reports whether an error occurred. The function returns 0 on success and nonzero if an error occurred.

Example

/* Quickly insert a row into a table */
mysql_send_query(&mysql, "INSERT INTO mytable VALUES ('blah', 'fnor')");

int mysql_shutdown(MYSQL *mysql)

Shuts down the MySQL database server. The user must have Shutdown privileges on the current connection to use this function. The function returns 0 if the operation was successful and nonzero in the case of an error.

Example

if ( mysql_shutdown(&mysql) ) {
       printf("Server not shut down... Check your permissions...\n");
} else {
       printf("Server successfully shut down!\n");
}

char *mysql_ssl_cipher(MYSQL *mysql)

Returns the name of the SSL cipher that is used (or will be used) with the current connection. This could be RSA, blowfish, or any other cipher supported by the server's SSL library (MySQL uses OpenSSL by default, if SSL is enabled). The MySQL server and client must both have been compiled with SSL support for this function to work properly.

Example

printf("This connection is using the %s cipher for security.\n",
             Mysql_ssl_cipher(&mysql));

int mysql_ssl_clear(MYSQL *mysql)

Clears any SSL information associated with the current connection. If mysql_ssl_clear is called before the connection has been made, the connection will be made without SSL. The function returns 0 on success and nonzero if an error occurs. It must be called before mysql_real_connect to have any effect. The MySQL server and client must both have been compiled with SSL support for this function to work properly.

Example

/* init a MYSQL structure and set SSL options...*/
/* Changed my mind, I don't want this connection to use SSL: */
mysql_ssl_clear(&mysql);

int mysql_ssl_set(MYSQL *mysql, char *key, char *certificate, char *authority,
                   char *authority_path)

Sets SSL information for the current connection and causes the connection to be made using SSL for encryption. This function must be called before mysql_real_connect to have any effect. The arguments are (beyond the pointer to the MYSQL structure) the text of the SSL public key used for the connection, the filename of the certificate used, the name of the authority that issued the certificate, and the directory that contains the authority's certificates.

The function returns 0 on success and nonzero if an error occurs. The MySQL server and client both must have been compiled with SSL support for this function to work properly.

Example

/* 'key' contains an SSL public key.
   'cert' contains the filename of a certificate
   'ca' contains the name of the certificate authority
   'capath' contains the directory containing the certificate
*/
/* Create an initialized MYSQL structure using mysql_init */
mysql_ssl_set(&mysql, key, cert, ca, capath);
/* Now, when mysql_real_connect is called, the connection will use SSL for
   encryption. */

char *mysql_stat(MYSQL *mysql)

Returns information about the current operating status of the database server. This includes the uptime, the number of running threads, and the number of queries being processed, among other information.

Example

printf("Server info\n-----------\n%s\n", mysql_stat(&mysql));
/* Output may look like this:
Server info
-----------
Uptime: 259044  Threads: 1  Questions: 24  Slow queries: 0  Opens: 6  
Flush tables: 1 Open tables: 0 Queries per second avg: 0.000
Everything below the row of hyphens is all on one line
*/

MYSQL_RES *mysql_store_result(MYSQL *mysql)

Reads the entire result of a query and stores it in a MYSQL_RES structure. Either this function or mysql_use_result must be called to access return information from a query. You must call mysql_free_result to free the MYSQL_RES structure when you are done with it.

The function returns a null value in the case of an error. The function also returns a null value if the query was not of a type that returns data (such as an INSERT or UPDATE query). If you receive a null pointer and are not sure if the query was supposed to return data or not, you can call mysql_field_count to find the number of fields the query was supposed to return. If zero, then it was a non-SELECT statement, and the pointer should be null. Otherwise, an error has occurred.

If the query was a SELECT-type statement, but happens to contain no data, this function will still return a valid (but empty) MYSQL_RES structure (it will not be a null pointer).

Example

MYSQL_RES results;
mysql_query(&mysql, "SELECT * FROM people");
results = mysql_store_result(&mysql);
/* 'results' should now contain all of the information from the 'people' table */
if (!results) { printf("An error has occurred!\n"); }

/* 'query' is some query string we obtained elsewhere, 
    we're not sure what it is... */
mysql_query(&mysql, query);
results = mysql_store_result(&mysql);
if (!results) { /* An error might have occurred, 
                   or maybe this is just a non-SELECT statement */
     if (! mysql_field_count(&mysql) ) { /* Aha! This is zero so it was 
                                            a non-SELECT statement */
               printf("No error here, just a non-SELECT statement...\n");
     } else {
               printf("An error has occurred!\n");
     }
}

unsigned long mysql_thread_id(MYSQL * mysql)

Returns the thread ID of the current connection. This value can be used with mysql_kill to terminate the thread in case of an error. The thread ID will change if you disconnect from the server and reconnect (which may happen automatically, without warning, if you use mysql_ping). Therefore, you should call this function immediately before you use the value.

Example

thread_id = mysql_thread_id(&mysql);
/* This number can be used with mysql_kill(  ) to terminate the current thread. */

unsigned int mysql_thread_safe(void)

Indicates whether the MySQL client library is safe to use in a threaded environment. The function returns a true value if the library is thread safe and 0 (false) if it is not.

Example

if (mysql_thread_safe(  )) {
     printf("This library is thread safe... thread away!\n");
} else {
     printf("This library is *not* thread safe, be careful!\n");
}

MYSQL_RES *mysql_use_result(MYSQL *mysql)

Reads the result of a query row by row and allows access to the data through a MYSQL_RES structure. Either this function or mysql_store_result must be called to access return information from a query. Because this function does not read the entire data set all at once, it is faster and more memory efficient than mysql_store_result. However, when using this function, you must read all the rows of the data set from the server or else the next query will receive the leftover data. Also, you cannot run any other queries until you are done with the data in this query. Even worse, no other threads running on the server can access the tables used by the query until you are finished. For this reason, you should use this function only when you are certain you can read the data in a timely manner and release it. You must call mysql_free_result to free the MYSQL_RES structure when you are done with it.

The function returns a null value in the case of an error. The function also returns a null value if the query was not of a type that returns data (such as an INSERT or UPDATE query). If you receive a null pointer and are not sure if the query was supposed to return data or not, you can call mysql_field_count to find the number of fields the query was supposed to return. If zero, then it was a non-SELECT statement, and the pointer should be null. Otherwise, an error has occurred.

If the query was a SELECT-type statement, but happens to contain no data, this function still returns a valid (but empty) MYSQL_RES structure; i.e., it does not return a null pointer.

Example

MYSQL_RES results;
mysql_query(&mysql, "SELECT * FROM people");
results = mysql_store_result(&mysql);
/* 'results' will now allow access (using mysql_fetch_row) to the table
   data, one row at a time */