Database and SQL

COUNT() counts all rows = 4. COUNT(c1+c2) counts non-null results - NULL+anything=NULL, and 2+0=2, 1+1=2, so only 2 rows have non-null sums. COUNT(DISTINCT(c1+c2)) counts distinct non-null values - both sums equal 2, so only 1 distinct value. Note: the row with NULL NULL is counted in COUNT() but not in the expression counts.

NOT EXISTS returns true when the subquery returns no rows. This query selects employees where their department number doesn't exist in the DEPT table - meaning they're orphaned records with invalid department references. This is a common pattern for finding referential integrity violations. Options B, C, and D describe specific department conditions that don't match the NOT EXISTS logic.

SELECT is the fundamental SQL statement used to query and retrieve data from database tables. It's the most commonly used SQL command and forms the basis of data extraction operations. Options like FETCH, EXTRACT, and GET are not standard SQL keywords for data retrieval.

DELETE is the DML statement used to remove specific rows from a database table based on conditions. DROP removes the entire table structure, TRUNCATE removes all rows but keeps the table structure, and REMOVE is not a standard SQL command. DELETE is the correct choice for removing data while preserving the table.

A trigger is procedural code (PL/SQL in Oracle) that automatically executes in response to certain events on a table or view. The answer captures both aspects: it's written in PL/SQL (option A) and is executed by events (option B). Triggers are used to enforce business rules, maintain data integrity, and automate database tasks.

DML (Data Manipulation Language) statements are used to manipulate data within existing table structures. Modifying rows (using UPDATE, INSERT, DELETE) is DML operation. Adding a column is DDL (Data Definition Language) because it changes the table structure, not just the data. The question tests understanding of DML vs DDL distinction.

The ORDER BY clause is used in SQL to sort the result set in ascending or descending order based on one or more columns. ORDER alone is not a valid clause - it must be followed by BY. SORT is not a standard SQL keyword for sorting results. GROUP BY is used for aggregation, not sorting.

The standard execution and syntax order of clauses in a SQL SELECT statement is WHERE (I), followed by GROUP BY (II), then HAVING (III), and finally ORDER BY (IV). This ensures rows are filtered, grouped, groups are filtered, and then results are sorted.

Views containing columns defined by expressions (calculated/computed columns) cannot be used directly for INSERT operations because the database cannot determine how to map values back to base table columns. The other options (FROM, WHERE) are standard SQL clauses that do not prevent data insertion through views.

Teradata Fast Load supports loading into multiset tables, which by definition allow duplicate rows. Unlike set tables that enforce uniqueness, multiset tables accept duplicates, and Fast Load can load them without error.

The %TYPE attribute in PL/SQL allows you to declare a variable that automatically inherits the datatype of a specific column in a database table. This provides maintainability - if the column's datatype changes, variables using %TYPE automatically adapt without code modification.

Standard SQL does not have built-in procedural error handling mechanisms like try-catch blocks. While modern SQL implementations (PL/SQL, T-SQL) extend SQL with programming constructs for exception handling, pure SQL itself is a declarative language without explicit error handling facilities.

A PL/SQL block consists of optional DECLARE and EXCEPTION sections, and mandatory BEGIN and END sections. The options contain typos ('DELCARE' or 'DECALRE'), but the concept identifies the DECLARE (spelled DECALRE here) and EXCEPTION clauses as the optional parts of the block structure.

A ROLLBACK statement explicitly ends a transaction by undoing any uncommitted changes. Therefore, it is indeed used to close a transaction.