Programming the Perl DBI

Atomic Fetching

When you want to fetch only one row, atomic fetching allows you to compress the four-stage data fetching cycle (as described earlier) into a single method. The two methods you can use for atomic fetching are selectrow_array( ) and selectrow_arrayref( ) . They behave in a similar fashion to their row-oriented cousins, fetchrow_array( ) and fetchrow_arrayref( ), the major differences being that the two atomic methods do not require a prepared and executed statement handle to work, and, more importantly, that they will return only one row of data.

Because neither method requires a statement handle to be used, they are actually invoked via a database handle. For example, to select the name and type fields from any arbitrary row in our megaliths database, we can write the following code:

### Assuming a valid $dbh exists...
( $name, $mapref ) = 
    $dbh->selectrow_array( "SELECT name, mapref
                            FROM megaliths" );
print "Megalith $name is located at $mapref\n";

This is far more convenient than using the prepare() and execute() then the fetchrow_array( ) or fetchrow_arrayref( ) methods for single rows.

Finally, bind values can be supplied, which again helps with the reuse of database resources.