Fetching Result Data

THIS TOPIC APPLIES TO: yesSQL ServeryesAzure SQL DatabaseyesAzure SQL Data Warehouse yesParallel Data Warehouse

An ODBC application has three options for fetching result data.

The first option is based on SQLBindCol. Before fetching the result set, the application uses SQLBindCol to bind each column in the result set to a program variable. After the columns have been bound, the driver transfers the data of the current row into the variables bound to the result set columns each time the application calls SQLFetch or SQLFetchScroll. The driver handles data conversions if the result set column and program variable have different data types. If the application has SQL_ATTR_ROW_ARRAY_SIZE set greater than 1, it can bind result columns to arrays of variables, which will all be filled on each call to SQLFetchScroll.

The second option is based on SQLGetData. The application does not use SQLBindCol to bind result set columns to program variables. After each call to SQLFetch, the application calls SQLGetData once for each column in the result set. SQLGetData instructs the driver to transfer data from a specific result set column to a specific program variable and specifies the data types of the column and variable. This allows the driver to convert data if the result column and program variable have different data types. Text, ntext, and image columns are typically too large to fit into a program variable but can still be retrieved using SQLGetData. If the text, ntext, or image data in the result column is larger than the program variable, SQLGetData returns SQL_SUCCESS_WITH_INFO and SQLSTATE 01004 (string data, right truncated). Successive calls to SQLGetData return successive chunks of the text or image data. When the end of the data is reached, SQLGetData returns SQL_SUCCESS. Each fetch returns a set of rows, or rowset, if SQL_ATTR_ROW_ARRAY_SIZE is greater than 1. Before using SQLGetData, you must first use SQLSetPos to specify a specific row within the rowset as the current row.

The third option is to use a mix of SQLBindCol and SQLGetData. An application could, for example, bind the first ten columns of a result set and then, on each fetch, call SQLGetData three times to retrieve the data from three unbound columns. This would typically be used when a result set contains one or more text or image columns.

Depending on the cursor options set for the result set, an application can also use the scrolling options of SQLFetchScroll to scroll around the result set.

Excess use of SQLBindCol to bind a result set column to a program variable is expensive because SQLBindCol causes an ODBC driver to allocate memory. When you bind a result column to a variable, that binding remains in effect until you either call SQLFreeHandle to free the statement handle or call SQLFreeStmt with fOption set to SQL_UNBIND. The bindings are not automatically undone when the statement completes.

This logic allows you to effectively deal with executing the same SELECT statement several times with different parameters. Because the result set keeps the same structure, you can bind the result set once, process all the SELECT statements, then call SQLFreeStmt with fOption set to SQL_UNBIND after the last execution. You should not call SQLBindCol to bind the columns in a result set without first calling SQLFreeStmt with fOption set to SQL_UNBIND to free any previous bindings.

When using SQLBindCol, you can either do row-wise or column-wise binding. Row-wise binding is somewhat faster than column-wise binding.

You can use SQLGetData to retrieve data on a column-by-column basis instead of binding result set columns using SQLBindCol. If a result set contains only a few rows, using SQLGetData instead of SQLBindCol is faster; otherwise, SQLBindCol gives the best performance. If you do not always put the data in the same set of variables, you should use SQLGetData instead of constantly rebinding. You can only use SQLGetData on columns that are in the select list after all columns are bound with SQLBindCol. The column must also appear after any columns on which you have already used SQLGetData.

The ODBC functions that deal with moving data into or out of program variables, such as SQLGetData, SQLBindCol, and SQLBindParameter, support implicit data type conversion. For example, if an application binds an integer column to a character string program variable, the driver automatically converts the data from integer to character before placing it into the program variable.

Data conversion in applications should be minimized. Unless data conversion is required for the processing done by the application, applications should bind columns and parameters to program variables of the same data type. If the data must be converted from one type to another, however, it is more efficient to have the driver do the conversion than doing it in the application. The SQL Server Native Client ODBC driver normally just transfers data directly from the network buffers to the variables of the application. Requesting the driver to do data conversion forces the driver to buffer the data and use CPU cycles to convert the data.

Program variables should be large enough to hold data transferred in from a column, except for text, ntext, and image data. If an application attempts to retrieve result set data and place it into a variable that is too small to hold it, the driver generates a warning. This forces the driver to allocate memory for the message, and the driver and application both have to spend CPU cycles processing the message and doing error handling. The application should either allocate a variable large enough to hold the data being retrieved or use the SUBSTRING function in the select list to reduce the size of the column in the result set.

Care must be taken when using SQL_C_DEFAULT to specify the type of the C variable. SQL_C_DEFAULT specifies that the type of the C variable matches the SQL data type of the column or parameter. If SQL_C_DEFAULT is specified for an ntext, nchar, or nvarchar column, Unicode data is returned to the application. This can cause various problems if the application has not been coded to handle Unicode data. The same types of problems can occur with the uniqueidentifier (SQL_GUID) data type.

text, ntext, and image data is typically too large to fit into a single program variable, and is usually processed with SQLGetData instead of SQLBindCol. When using server cursors, the SQL Server Native Client ODBC driver is optimized to not transmit the data for unbound text, ntext, or image columns at the time the row is fetched. The text, ntext, or image data is not actually retrieved from the server until the application issues SQLGetData for the column.

This optimization can be applied to applications so that no text, ntext, or image data is displayed while a user is scrolling up and down a cursor. After the user selects a row, the application can call SQLGetData to retrieve the text, ntext, or image data. This saves transmitting the text, ntext, or image data for any of the rows the user does not select and can save the transmission of very large amounts of data.

See Also

Processing Results (ODBC)