Mastering Oracle SQL

The SQL Inside Your PL/SQL

Now that we’ve explored calling PL/SQL from SQL, let’s turn the tables and explore the use of SQL inside your PL/SQL code. SQL is great at manipulating large sets of data, but there are situations where you need to work with data at the row level. PL/SQL, with its looping and cursor control capabilities, allows the flexibility to work at the set level using SQL or at the row level using cursors. However, many PL/SQL programmers forego the power of SQL and do everything at the row level, even when it is unnecessary and time-consuming to do so.

As an analogy, imagine that you are working at a warehouse, and a large shipment of parts arrives on the loading dock. Your job is to separate the shipment by part type and distribute the pieces to different areas of the warehouse. To make your job easier, the warehouse owner has procured the best forklift money can buy. There are two possible strategies to employ:

1. Pick up one box at a time, determine the type, and drive it to the appropriate destination.

2. Spend some time analyzing the situation, determine that every box on a pallet is of the same type, and drive entire pallets to the appropriate destination.

While this analogy might be overly simplistic, it does serve to illustrate the difference between set operations and row operations. Allowing the Oracle server to manipulate large sets in a single operation can often yield a performance improvement of several orders of magnitude over manipulating one row at a ...