microORMs

MicroORMs are lightweight data access libraries that map rows from relational databases to domain objects while avoiding the breadth of features offered by full ORMs. They provide a thin abstraction over SQL execution and result reading, emphasizing speed and precise control over queries.

Typically, a microORM requires the developer to supply SQL (or a small query builder) and uses the library to bind parameters and map the resulting rows to objects. Result mapping is usually simple, often supporting one-to-one mappings and, to a limited extent, lightweight handling of related data. Metadata and change-tracking features common in larger ORMs are minimized or absent, placing more responsibility on the developer for data integrity and query design.

Use cases include performance-critical applications, services with well-defined SQL, and teams that prefer explicit SQL craftsmanship or want minimal runtime overhead. MicroORMs are popular in ecosystems where developers seek object mapping without the overhead of a full ORM, or where precise SQL tuning is important.

Advantages include improved performance, lower memory usage, clearer SQL paths, and easier testing of SQL logic. Drawbacks can be increased boilerplate for complex data models, potential duplication of SQL across code, reduced automation for complex relationships or migrations, and greater responsibility for ensuring data integrity.

Examples widely cited in practice include Dapper in the .NET ecosystem, sqlx in Go, JDBI in Java, and ServiceStack’s OrmLite in C#. These libraries illustrate the core idea: mapping results to objects with minimal abstraction, while leaving SQL design largely in the developer’s hands.