Using Database Triggers in GORM

Triggers are useful tools in relational databases that allow you to execute custom code when specific events occur on a table. For instance, triggers can automatically populate the audit log table whenever a new mutation is applied to a different table. This way we ensure that all changes (including those made by other applications) are meticulously recorded, enabling the enforcement on the database-level and reducing the need for additional code in the applications.

This guide explains how to attach triggers to your GORM types (objects) and configure the schema migration to manage both the triggers and the GORM model as a single migration unit using Atlas.

Atlas Pro Feature

Atlas support for Triggers used in this guide is available exclusively to Pro users. To use this feature, run:

atlas login

Getting started with Atlas and GORM

Before we continue, ensure you have installed the Atlas GORM Provider on your GORM project.

To set up, follow along the getting started guide for GORM and Atlas.

Composite Schema

The GORM package is mostly used for defining tables (our Go types) and interacting with the database. Table triggers or any other database native objects do not have representation in GORM models. A trigger function can be defined once, and used in multiple triggers in different tables.

In order to extend our PostgreSQL schema to include both our GORM types and their triggers, we configure Atlas to read the state of the schema from a Composite Schema data source. Follow the steps below to configure this for your project:

1. Let's define a simple model with two types (tables): users and user_audit_logs:

models.go

package models

import "gorm.io/gorm"

type User struct {
    gorm.Model
    ID   uint
    Name string
}

type UserAuditLogs struct {
    gorm.Model
    OperationType   string
    OperationTime   string
    OldValue        string
    NewValue        string
}

Now, suppose we want to log every change to the users table and save it in the user_audit_logs table. To achieve this, we need to create a trigger function on INSERT, UPDATE and DELETE operations and attach it to the users table.

2. Next step, we define a trigger function ( audit_users_changes) and attach it to the users table using the CREATE TRIGGER commands:

schema.sql

-- Function to audit changes in the users table.
CREATE OR REPLACE FUNCTION audit_users_changes()
RETURNS TRIGGER AS $$
BEGIN
    IF (TG_OP = 'INSERT') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(NEW));
        RETURN NEW;
    ELSIF (TG_OP = 'UPDATE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD), row_to_json(NEW));
        RETURN NEW;
    ELSIF (TG_OP = 'DELETE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD));
        RETURN OLD;
    END IF;
    RETURN NULL;
END;
$$ LANGUAGE plpgsql;

-- Trigger for INSERT operations.
CREATE TRIGGER users_insert_audit AFTER INSERT ON users FOR EACH ROW EXECUTE FUNCTION audit_users_changes();

-- Trigger for UPDATE operations.
CREATE TRIGGER users_update_audit AFTER UPDATE ON users FOR EACH ROW EXECUTE FUNCTION audit_users_changes();

-- Trigger for DELETE operations.
CREATE TRIGGER users_delete_audit AFTER DELETE ON users FOR EACH ROW EXECUTE FUNCTION audit_users_changes();

3. In your atlas.hcl config file, add a composite_schema that includes both our GORM model and the custom triggers defined in schema.sql:

atlas.hcl

data "composite_schema" "app" {
  # Load the GORM model first
  schema "public" {
    url = data.external_schema.gorm.url
  }
  # Next, load the trigger.
  schema "public" {
    url = "file://schema.sql"
  }
}

env "local" {
  src = data.composite_schema.app.url
  dev = "docker://postgres/15/dev?search_path=public"
}

Usage

After setting up our composite schema, we can get its representation using the atlas schema inspect command, generate schema migrations for it, apply them to a database, and more. Below are a few commands to get you started with Atlas:

Inspect the Schema

The atlas schema inspect command is commonly used to inspect databases. However, we can also use it to inspect our composite_schema and print the SQL representation of it:

atlas schema inspect \
  --env local \
  --url env://src \
  --format '{{ sql . }}'

The command above prints the following SQL. Note, the audit_users_changes function and the triggers are defined after the users and user_audit_logs tables:

-- Create "user_audit_logs" table
CREATE TABLE "user_audit_logs" ("id" bigserial NOT NULL, "created_at" timestamptz NULL, "updated_at" timestamptz NULL, "deleted_at" timestamptz NULL, "operation_type" text NULL, "operation_time" text NULL, "old_value" text NULL, "new_value" text NULL, PRIMARY KEY ("id"));
-- Create index "idx_user_audit_logs_deleted_at" to table: "user_audit_logs"
CREATE INDEX "idx_user_audit_logs_deleted_at" ON "user_audit_logs" ("deleted_at");
-- Create "users" table
CREATE TABLE "users" ("id" bigserial NOT NULL, "created_at" timestamptz NULL, "updated_at" timestamptz NULL, "deleted_at" timestamptz NULL, "name" text NULL, PRIMARY KEY ("id"));
-- Create index "idx_users_deleted_at" to table: "users"
CREATE INDEX "idx_users_deleted_at" ON "users" ("deleted_at");
-- Create "audit_users_changes" function
CREATE FUNCTION "audit_users_changes" () RETURNS trigger LANGUAGE plpgsql AS $$
BEGIN
    IF (TG_OP = 'INSERT') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(NEW));
RETURN NEW;
ELSIF (TG_OP = 'UPDATE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD), row_to_json(NEW));
RETURN NEW;
ELSIF (TG_OP = 'DELETE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD));
RETURN OLD;
END IF;
RETURN NULL;
END;
$$;
-- Create trigger "users_delete_audit"
CREATE TRIGGER "users_delete_audit" AFTER DELETE ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();
-- Create trigger "users_insert_audit"
CREATE TRIGGER "users_insert_audit" AFTER INSERT ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();
-- Create trigger "users_update_audit"
CREATE TRIGGER "users_update_audit" AFTER UPDATE ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();```

Generate Migrations For the Schema

To generate a migration for the schema, run the following command:

atlas migrate diff \
  --env local

Note that a new migration file is created with the following contents:

migrations/20240712090543.sql

-- Create "user_audit_logs" table
CREATE TABLE "user_audit_logs" ("id" bigserial NOT NULL, "created_at" timestamptz NULL, "updated_at" timestamptz NULL, "deleted_at" timestamptz NULL, "operation_type" text NULL, "operation_time" text NULL, "old_value" text NULL, "new_value" text NULL, PRIMARY KEY ("id"));
-- Create index "idx_user_audit_logs_deleted_at" to table: "user_audit_logs"
CREATE INDEX "idx_user_audit_logs_deleted_at" ON "user_audit_logs" ("deleted_at");
-- Create "users" table
CREATE TABLE "users" ("id" bigserial NOT NULL, "created_at" timestamptz NULL, "updated_at" timestamptz NULL, "deleted_at" timestamptz NULL, "name" text NULL, PRIMARY KEY ("id"));
-- Create index "idx_users_deleted_at" to table: "users"
CREATE INDEX "idx_users_deleted_at" ON "users" ("deleted_at");
-- Create "audit_users_changes" function
CREATE FUNCTION "audit_users_changes" () RETURNS trigger LANGUAGE plpgsql AS $$
BEGIN
    IF (TG_OP = 'INSERT') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(NEW));
RETURN NEW;
ELSIF (TG_OP = 'UPDATE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value, new_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD), row_to_json(NEW));
RETURN NEW;
ELSIF (TG_OP = 'DELETE') THEN
        INSERT INTO user_audit_logs(operation_type, operation_time, old_value)
        VALUES (TG_OP, CURRENT_TIMESTAMP, row_to_json(OLD));
RETURN OLD;
END IF;
RETURN NULL;
END;
$$;
-- Create trigger "users_delete_audit"
CREATE TRIGGER "users_delete_audit" AFTER DELETE ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();
-- Create trigger "users_insert_audit"
CREATE TRIGGER "users_insert_audit" AFTER INSERT ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();
-- Create trigger "users_update_audit"
CREATE TRIGGER "users_update_audit" AFTER UPDATE ON "users" FOR EACH ROW EXECUTE FUNCTION "audit_users_changes"();

Apply the Migrations

To apply the migration generated above to a database, run the following command:

atlas migrate apply \
  --env local \
  --url "postgres://postgres:pass@localhost:5432/database?search_path=public&sslmode=disable"

Apply the Schema Directly on the Database

Sometimes, there is a need to apply the schema directly to the database without generating a migration file. For example, when experimenting with schema changes, spinning up a database for testing, etc. In such cases, you can use the command below to apply the schema directly to the database:

atlas schema apply \
  --env local \
  --url "postgres://postgres:pass@localhost:5432/database?search_path=public&sslmode=disable"

Or, using the Atlas Go SDK:

ac, err := atlasexec.NewClient(".", "atlas")
if err != nil {
    log.Fatalf("failed to initialize client: %w", err)
}
// Automatically update the database with the desired schema.
// Another option, is to use 'migrate apply' or 'schema apply' manually.
if _, err := ac.SchemaApply(ctx, &atlasexec.SchemaApplyParams{
    Env: "local",
    URL: "postgres://postgres:pass@localhost:5432/database?search_path=public&sslmode=disable",
}); err != nil {
    log.Fatalf("failed to apply schema changes: %w", err)
}