Confiture

FraiseQL is database-first — you write SQL tables and views, and FraiseQL maps them to GraphQL. But who manages the database itself? You confiture it.

confiture /kɔ̃.fi.tyʁ/ — French for jam, preserve. Sounds like configure. That’s the point.

Confiture is a database schema evolution framework that provides four strategies (“Mediums”) for every database lifecycle scenario. It treats your DDL files as the single source of truth, not migration history. You don’t configure your database — you confiture it.

The 4 Mediums

Medium	What It Does	When to Use
Build	Creates a fresh database from DDL files in under 1s	Development, CI/CD, testing
Migrate	Applies incremental ALTER statements	Production schema changes
Sync	Copies production data with anonymization	Realistic local development
Schema-to-Schema	Zero-downtime migration via FDW	Major production refactoring

FraiseQL + Confiture Workflow

The typical development cycle:

# 1. Write your SQL (tables in 01_write/, views in 02_read/)
vim db/schema/01_write/tb_order.sql
vim db/schema/02_read/v_order.sql

# 2. Build fresh database
confiture build --env local

# 3. Define GraphQL types
vim schema.py

# 4. Compile the mapping
fraiseql compile

# 5. Serve
fraiseql run

For existing databases:

# 1. Update SQL files
vim db/schema/01_write/tb_user.sql  # Add new column

# 2. Create migration for existing data
confiture migrate generate add_avatar_url

# 3. Apply migration
confiture migrate up --env local

# 4. Update view to include new field
vim db/schema/02_read/v_user.sql

# 5. Rebuild views (or apply view migration)
confiture build --env local --views-only

# 6. Recompile and serve
fraiseql run

Configuration

Confiture uses a confiture.yaml configuration file:

confiture.yaml

project:
  name: my-fraiseql-app

environments:
  local:
    database:
      host: localhost
      port: 5432
      database: myapp
      user: postgres
      password: postgres
    schema_dirs:
      - db/schema/00_extensions
      - db/schema/01_write
      - db/schema/02_read
      - db/schema/03_functions
    migrations_dir: db/migrations

  production:
    database:
      host: ${DB_HOST}
      port: ${DB_PORT}
      database: ${DB_NAME}
      user: ${DB_USER}
      password: ${DB_PASSWORD}
    schema_dirs:
      - db/schema/00_extensions
      - db/schema/01_write
      - db/schema/02_read
      - db/schema/03_functions
    migrations_dir: db/migrations

When to Use Which Medium

Scenario	Medium	Command
Fresh dev environment	Build	confiture build --env local
CI/CD test database	Build	confiture build --env ci
Add column to production	Migrate	confiture migrate up
Test with realistic data	Sync	confiture sync --from prod --to local --anonymize
Rename column on 50M row table	Schema-to-Schema	confiture migrate schema-to-schema ...
Reset local database	Build	confiture build --env local

Next Steps

Build from DDL

Instant database creation from your DDL files in under 1 second.

Incremental Migrations

Schema evolution with data for existing production databases.

Production Data Sync

Safe data synchronization with built-in PII anonymization.

Schema-to-Schema

Zero-downtime refactoring for major production changes.