Question

I am using c++ with ADO to connect to a mySql database, and I am using the standard ADO/C++ method to create a connection to the mySql database, and recordset is the pointer to the retrieved first record

_RecordsetPtr recordset;
recordset->Open("Select * from table",p_connection_.GetInterfacePtr(),adOpenForwardOnly,adLockReadOnly,adCmdText);  

My concern is if the table contains too many records, and if I query all records, it will consume alot of memory?

I want to only retrieve, maybe 100 records each time and process them. Is it possible? The table does not contain id or index as its attribute, so “Select * from table where id >= 1 and id <= 100” does not work.

Answer 1

You will want to use limits on the query and cycle through them.

//SELECT * FROM table LIMIT 0 OFFSET 100

int tlimit, blimit;
std::string query;
std::stringstream sstm;
_RecordsetPtr recordset, count;

count->Open("SELECT COUNT(*) FROM table",p_connection_.GetInterfacePtr(),adOpenForwardOnly,adLockReadOnly,adCmdText);

for(int i = 0; i < count/100 + 1; i++)
{
   tlimit = 100 * i + 100;
   blimit = 100 * i;

   sstm << "SELECT * FROM table LIMIT " << blimit << " OFFSET " << tlimit;
   query = sstm.str();

   recordset->Open(query,p_connection_.GetInterfacePtr(),adOpenForwardOnly,adLockReadOnly,adCmdText);

   //suggest passing the recordset to a function to do what ever you want with it here
}

Note that if you are not using a database that starts its records off at 1 you will have to modify that algorithm a bit.