DML

Data Manipulation Language (DML) is a subset of SQL used to access and modify data stored in relational databases. DML operations enable applications to insert new rows, update existing data, delete rows, and retrieve data through queries. Although many references place data retrieval under a separate data query language (DQL), most discussions of DML cover the core data-changing statements and their use in data-driven applications.

The primary DML statements are INSERT, UPDATE, and DELETE, which respectively add, modify, and remove rows. Some databases treat SELECT as part of DML as well; others classify it as DQL. DML does not define or alter database schemas; that role belongs to Data Definition Language (DDL). DCL (data control language) and TCL (transaction control language) sit alongside DML in the SQL family to manage permissions and transactions.

DML operations are typically executed within transactions and can be rolled back if errors or constraints are violated. Many systems support features such as bulk or batched operations, parameterized queries, and prepared statements to improve performance and security. Some dialects provide extensions, such as MERGE for upsert functionality or RETURNING clauses to retrieve affected rows immediately after execution.

DML is a foundational element of working with relational databases, appearing in most database management systems and programming environments that interact with SQL databases. Understanding DML semantics is essential for data integrity, performance tuning, and correct application behavior when manipulating stored data.