README
¶
D3 ORM
D3 is golang ORM and DataMapper. This project was design with respect to such ORM's like hibernate and doctrine. Main task - give an instrument for create nice domain layer. If your business code not expressive enough, or if you want unit tests without a database, or if in your application, for some reason, you need aggregates, repositories, entities and value objects - d3 can be a good choice.
Motivation. Why another ORM?
In my opinion in GO have a lot of good ORM's. They are pretty fast and may save the developer from a lot of boilerplate code. But, if you want to write code in DDD style (using DDD patterns and philosophy) it's not enough, because DDD approach imposes certain requirements. Main requirement - persistence ignorance. Current GO ORM's do not provide sufficient level of abstraction for this. Other words, we need keep business logic free of data access code. That's why D3 created.
Main futures
- code generation instead of reflection
- db schema auto generation
- one-to-one, one-to-many and many-to-many relations
- lazy and eager relation loading
- query builder
- relation fetch strategies (eager/lazy as above or extract relation in one query with join)
- fetched entity cache (first level cache)
- cascade remove and update of related entities
- application level transaction (UnitOfWork)
- db transactions support
- smart persist layer don't generate redundant queries on entity updates
- UUID support
Documentation
All documentation is on project wiki.
Example of usage
Full example of usage you can see in this project. Note, that is it is a test project with main responsibility show d3 orm in action.
Tests
D3 integration tests require a postgreSQL database. Connect to the database specified in the D3_PG_TEST_DB environment variable.
Roadmap
- add mysql and sqlite support
- composite pk
- not only schema generation but generation of schema diff's
- embedding structures
- index definition in entity comments
- generate fk's for relations
Note: Current project status - is pre-alpha. It can be used in production with some risky.
Directories