Executing database commands | LoopBack Documentation

Page Contents

Overview

Warning: In general, it is always better to perform database actions through Repository methods. Directly executing native database commands may lead to unexpected results and other issues.

When you project outgrows built-in Repository methods, you can execute native database commands to implement more complex data queries and manipulations, for example execute a custom SQL query or invoke a MongoDB command.

LoopBack provides two APIs:

DataSource-level execute() method
Repository-level execute() method

Both methods offer the same set of signatures, the implementation in Repository is just a thin wrapper delegating the task to DataSource.

Example use:

const result = await repository.execute('SELECT * FROM Products');

Examples

MySQL

Use parameterized queries to avoid SQL injection attacks.

const result = await repository.execute('SELECT * FROM Products WHERE id = ?', [
  productId,
]);

PostgreSQL

const result = await repository.execute(
  'SELECT * FROM Products WHERE id = $1',
  [productId],
);

MongoDB

const result = await repository.execute('MyCollection', 'aggregate', [
  {$unwind: '$data'},
  {$out: 'tempData'},
]);

Use parameterized queries whenever possible instead of constructing SQL strings manually.

See API docs for parameter reference, additional information and examples:

Important: Each connector implements a slightly different flavor of execute() to match the capabilities supported by the database engine. Please refer to connector documentation to learn more about the expected parameters and command/query syntax to use.

Supported connectors

Not all connectors support execution of native database commands. Check your connector’s documentation to learn more. The following connectors are known to support execute method.