Skip to main content
Version: Next

Schema support

You can specify the database schemas to organize tables and other database objects, Platformatic DB will create entities using these schemas.

Example Configuration

Consider a database setup with two schemas, each containing a different set of tables:

CREATE SCHEMA IF NOT EXISTS "test1";
CREATE TABLE IF NOT EXISTS test1.movies (
id INTEGER PRIMARY KEY,
title TEXT NOT NULL
);

CREATE SCHEMA IF NOT EXISTS "test2";
CREATE TABLE IF NOT EXISTS test2.users (
id INTEGER PRIMARY KEY,
title TEXT NOT NULL
);
info

Note that if we use schemas and migrations, we must specify the schema in the migrations table as well (with postgresql, we assume we use the default public schema).

Configuring Schemas

To utilize multiple schemas, you must specify them in the schema section of the configuration file as follows:

  ...
"db": {
"connectionString": "(...)",
"schema": [
"test1", "test2"
],
"ignore": {
"versions": true
}
},
"migrations": {
"dir": "migrations",
"table": "test1.versions"
},

...
  • schema: An array specifying which schemas Platformatic DB should inspect to create entities.
  • ignore: Here, "versions": true specifies to ignore version tracking tables from entity generation.

Entity Naming

The entity names are then generated in the form schemaName + entityName, PascalCase (this is necessary to avoid name collisions in case there are tables with same name in different schemas). So for instance for the example above we generate the Test1Movie and Test2User entities.

Entity Names and Authorization: When using schemas, it's important to refer to entities by their full generated names (e.g., Test1Movies) when setting up authorization rules.

info

When using schemas, it's important to refer to entities by their full generated names (e.g., Test1Movies) when setting up authorization rules.

Issues

If you run into a bug or have a suggestion for improvement, please raise an issue on GitHub or join our Discord feedback channel.