Programming the Perl DBI

Handling Statements

In general, the most common reason for using dbish is to issue ad-hoc SQL statements to a database, either to check that the statement works before including it in a Perl program, or just to get some quick answers. This task is exactly what dbish was designed for.

dbish commands are entered as a forward slash (/) followed by a command name and optionally some extra arguments. For example:

/help

Anything entered that doesn’t start with a forward slash is considered to be part of an SQL statement and is appended to a ``statement buffer.'' Once the SQL statement is complete, you can execute it, and the results, if any, will be returned to your screen.

For example, to query the names of all sites in the megalithic database, type:

stones@dbi:Oracle:archaeo> SELECT name FROM megaliths
Current statement buffer (enter '/' to execute or '/help' for help):
SELECT name FROM megaliths

stones@dbi:Oracle:archaeo> /
'Avebury'
'Stonehenge'
'Lundin Links'
...
[132 rows of 1 fields returned]
stones@dbi:Oracle:archaeo>

Note that a forward slash by itself can be used to execute statements. After executing a statement, the statement buffer is cleared. But suppose we start typing in a new query and then change our minds about what we want to return:

stones@dbi:Oracle:archaeo> SELECT name FROM megaliths Current statement buffer (enter '/' to execute or '/help' for help): SELECT name FROM megaliths stones@dbi:Oracle:archaeo> SELECT name, mapref FROM megaliths Current statement buffer (enter ...