updb 0.1.0

Simple DB migration manager for non-bloated projects.

Not production ready at all. Use at your own risk.

Updb - simple DB migration manager for non-bloated projects

How it works? In short, a single migrations.sql file is used to store recent migrations. There is no need to create tons of up.sql and down.sql.

Only PostgreSQL is supported.

Key points

1. Do not create VCS inside of another VCS. There is no need to create tons of up.sql and down.sql files. It doesn't help. You can store all the migrations in a single migrations.sql file.
2. Oldest migrations are not needed so you can keep only 10-20 recent migrations in migrations.sql. All migrations are stored in Git any way.
3. Using a single file for storing 10-20 recent migrations can help you to observe what is happening to the DB schema. It's easier to read one file instead of 10-20 separated ones. You can read migrations.sql from time to time to gain a better understanding of how your project is going.
4. How often do you use a down.sql?

Installation

pip install updb

Usage

1. Create updb_conf.py and specify DBNAME and other settings (see example.updb_conf.py).
2. Write some migrations in migrations.sql file (just pure SQL, see example1.migrations.sql and example2.migrations.sql). Migrations should be separated by an empty line.
3. Run updb command. The script will enumerate new migrations and apply them.
4. The number of last applied migration will be stored in public.last_migration_n() function. Additionally it will appeared in migrations.sql at the first line.
5. Optionally per-table schema will be dumped to schema/{table}.sql files and full DB schema - to schema/schema.sql .
6. When you want to create new migrations, just add it to migrations.sql and run updb again.
7. If you need to create an empty DB for tests, use schema/schema.sql . Older versions are available via Git.

Suggested directory structure:

./project/db/migrations.sql
./project/db/schema/{table}.sql