Skip to content

Getting Started with PostgreSQL on JVM

First apply the gradle plugin in your project and set your database's dialect accordingly. 

plugins {
  id("app.cash.sqldelight") version "2.0.2"
}

repositories {
  google()
  mavenCentral()
}

sqldelight {
  databases {
    create("Database") {
      packageName.set("com.example")
      dialect("app.cash.sqldelight:postgresql-dialect:2.0.2")
    }
  }
}

plugins {
  id "app.cash.sqldelight" version "2.0.2"
}

repositories {
  google()
  mavenCentral()
}

sqldelight {
  databases {
    Database { // This will be the name of the generated database class.
      packageName = "com.example"
      dialect "app.cash.sqldelight:postgresql-dialect:2.0.2"
    }
  }
}

SQLDelight needs to know the schema of your database. There are typically two approaches to setting up your database's schema. The "Fresh Schema" approach assumes that you are starting with an empty database, and that all the statements necessary to bring it to the desired state will be applied all at once. The "Migration Schema" approach on the other hand assumes that you already have a database and schema set up (e.g. an existing production database), and that you'll progressively apply migrations over time to update your database's schema.

In SQLDelight, these approaches translate to either writing your table definitions in .sq files for a "Fresh Schema", or by writing migration statements in .sqm files for a "Migration Schema". In both cases, your SQL queries will be written in .sq files (as shown here).

Fresh Schema

Write your SQL statements in a .sq file under src/main/sqldelight. Typically the first statement in the .sq file creates a table, but you can also create indexes or set up default content.

src/main/sqldelight/com/example/sqldelight/hockey/data/Player.sq
CREATE TABLE hockeyPlayer (
  player_number INTEGER PRIMARY KEY NOT NULL,
  full_name TEXT NOT NULL
);

CREATE INDEX hockeyPlayer_full_name ON hockeyPlayer(full_name);

INSERT INTO hockeyPlayer (player_number, full_name)
VALUES (15, 'Ryan Getzlaf');

In the same .sq files you can start placing your sql statements to be executed at runtime.

Migration Schema

First, configure gradle to use migrations to assemble the schema:

sqldelight {
  databases {
    create("Database") {
      ...
      srcDirs("sqldelight")
      deriveSchemaFromMigrations.set(true)
    }
  }
}

sqldelight {
  databases {
    Database {
      ...
      srcDirs "sqldelight"
      deriveSchemaFromMigrations = true
    }
  }
}

Migration files have the extension .sqm, and must have a number in their file name indicating what order the migration file runs in. For example, given this hierarchy:

src
`-- main
    `-- sqldelight
        |-- v1__backend.sqm
        `-- v2__backend.sqm

SQLDelight will create the schema by applying v1__backend.sqm and then v2__backend.sqm. Place your normal SQL CREATE/ALTER statements in these files. If another service reads from your migrations files (like flyway), make sure to read the info on migrations and how to output valid SQL.

Typesafe SQL

Before you're able to execute SQL statements at runtime, you need to create a SqlDriver to connect to your database. The easiest way is off of a DataSource that you would get from hikari or other connection managers.

dependencies {
  implementation("app.cash.sqldelight:jdbc-driver:2.0.2")
}

dependencies {
  implementation "app.cash.sqldelight:jdbc-driver:2.0.2"
}

val driver: SqlDriver = dataSource.asJdbcDriver()

Regardless of if you specify the schema as fresh create table statements or through migrations, runtime SQL goes in .sq files.

Defining Typesafe Queries

SQLDelight will generate a typesafe function for any labeled SQL statement in a .sq file.

src/main/sqldelight/com/example/sqldelight/hockey/data/Player.sq
selectAll:
SELECT *
FROM hockeyPlayer;

insert:
INSERT INTO hockeyPlayer(player_number, full_name)
VALUES (?, ?);

insertFullPlayerObject:
INSERT INTO hockeyPlayer(player_number, full_name)
VALUES ?;

A "Queries" object will be generated for each .sq file containing labeled statements. For example, a PlayerQueries object will be generated for the Player.sq file shown above. This object can be used to call the generated typesafe functions which will execute the actual SQL statements.

fun doDatabaseThings(driver: SqlDriver) {
  val database = Database(driver)
  val playerQueries: PlayerQueries = database.playerQueries

  println(playerQueries.selectAll().executeAsList()) 
  // [HockeyPlayer(15, "Ryan Getzlaf")]

  playerQueries.insert(player_number = 10, full_name = "Corey Perry")
  println(playerQueries.selectAll().executeAsList()) 
  // [HockeyPlayer(15, "Ryan Getzlaf"), HockeyPlayer(10, "Corey Perry")]

  val player = HockeyPlayer(10, "Ronald McDonald")
  playerQueries.insertFullPlayerObject(player)
}

And that's it! Check out the other pages on the sidebar for other functionality.