Programming the Perl DBI

Database Handles

Database handles are the first step towards actually doing work with the database, in that they encapsulate a single connection to a particular database. Prior to executing SQL statements within a database, we must actually connect to the database. This is usually achieved through the DBI’s connect() method:

$dbh = DBI->connect( $data_source, ... );

The majority of databases nowadays tend to operate in a multiuser mode, allowing many simultaneous connections, and database handles are designed accordingly. An example might be if you wanted to write a stock-monitoring program that simultaneously monitored data in tables within different user accounts in the database. A DBI script could make multiple connections to the database, one for each user account, and execute SQL statements on each. Database handles are completely encapsulated objects, meaning that transactions from one database handle cannot ``cross-over'' or ``leak'' into another.

Database handles are children of their corresponding driver handle, which supports the notion that we could also make multiple simultaneous connections to multiple database types, as well as multiple simultaneous connections to databases of the same type. For example, a more complicated DBI script could make two connections to each of an Oracle and an Informix database to perform the above-mentioned monitoring. Figure 4.3, shown earlier, illustrates the capability of having multiple database handles connecting through a driver handle ...