PL/SQL (Procedural Language/SQL) extends SQL with procedural capabilities. Think of SQL as giving instructions ("get me all orders"), while PL/SQL is writing a program ("if order total > $1000, apply discount, else charge shipping"). SQL is declarative (what you want), PL/SQL is procedural (how to get it). This matters because complex business logic requires variables, loops, and conditional logic that SQL alone cannot provide. PL/SQL is compiled and cached, making it 3-5x faster than sending multiple SQL statements from application.

**1. PL/SQL Block Structure**

Every PL/SQL code follows this structure:

```sql

DECLARE

-- Variable declarations (optional)

v_customer_name VARCHAR2(100);

v_order_total NUMBER;

BEGIN

-- Executable statements (required)

SELECT name INTO v_customer_name

FROM customers

WHERE customer_id = 100;

DBMS_OUTPUT.PUT_LINE('Customer: ' || v_customer_name);

EXCEPTION

-- Exception handlers (optional)

WHEN NO_DATA_FOUND THEN

DBMS_OUTPUT.PUT_LINE('Customer not found');

WHEN OTHERS THEN

DBMS_OUTPUT.PUT_LINE('Error: ' || SQLERRM);

END;

/

```

**Why this matters**: The block structure provides clear separation: declarations, logic, error handling. The forward slash (/) executes the block. This structure is used in procedures, functions, triggers, and anonymous blocks.

**2. Variables and Data Types**

Variables store intermediate results. Use %TYPE to anchor to column type, %ROWTYPE for entire row.

```sql

DECLARE

-- Basic types

v_count NUMBER := 0; -- Initialize with value

v_name VARCHAR2(100);

v_price NUMBER(10,2);

v_active BOOLEAN := TRUE;

v_order_date DATE := SYSDATE;

-- Anchored types (recommended)

v_customer_name customers.name%TYPE; -- Same type as customers.name

v_order_amount orders.total_amount%TYPE;

-- Row type

v_customer_rec customers%ROWTYPE; -- Entire row structure

-- Constants