Skip to content

CLI

Experimental

The drift cli tool is experimental at the moment. Please report all issues you can find.

Usage

If your app depends on drift_dev, you're ready to use the CLI tool. In this article, we'll use dart run drift_dev to start the tool. The tool should be run from the same folder where you keep your pubspec.yaml.

Analyze

Runs drift's analyzer and linter across all .drift files in your project.

$ dart run drift_dev analyze

WARNING: For file test/data/tables/tables.drift:
WARNING: line 38, column 28: This table has columns without default values, so defaults can't be used for insert.
   ╷
38 │ defaultConfig: INSERT INTO config DEFAULT VALUES;
   │                            ^^^^^^
   ╵
INFO: Found 1 errors or problems

Exits with error code 1 if any error was found.

Identify databases

This is more of a test command to verify that drift's analyzer is working correctly. It will identify all databases or daos defined in your project.

$ dart run drift_dev identify-databases

Starting to scan in /tmp/example/ ...
INFO: example/example.dart has drift databases or daos: Database
INFO: lib/src/data/database.dart has drift databases or daos: AppDatabase
INFO: test/fake_db.dart has drift databases or daos: TodoDb, SomeDao

Schema tools

Dump for version control

This subcommand expects two paths, a Dart file and a target. The Dart file should contain exactly one class annotated with @DriftDatabase. Running the following command will export the database schema to json.

$ dart run drift_dev schema dump path/to/database.dart schema.json

The generated file (schema.json in this case) contains information about all

  • tables, including detailed information about columns
  • triggers
  • indices
  • @create-queries from included drift files
  • dependencies thereof

Exporting a schema can be used to generate test code for your schema migrations. For details, see the guide.

Exporting

In some cases, it can be beneficial to export a list of CREATE statements that define a drift database. The schema export command does just that. It takes a path to a Dart source file defining a drift database as an argument:

$ dart run drift_dev schema dump path/to/database.dart

It will output all statements that would be run by drift if the database were freshly created, each statement on its own line.

The optional --dialect option on schema export can be used to control the target dialect of the generated statements. It defaults to sqlite and experimentally also supports postgres and mariadb.