Asynchronous SQLAlchemy Object Relational Mapper.

This is an ORM for accessing SQLAlchemy using asyncio. Working on top of SQLAlchemy Core.

It presents a method of associating user-defined Python classes with database tables, and instances of those classes (objects) with rows in their corresponding tables.

Usage

Initialization:

from sqlalchemy import MetaData, Integer, String, DateTime
from aiopg.sa import create_engine
from aiosqlalchemy_miniorm import RowModel, RowModelDeclarativeMeta, BaseModelManager

metadata = MetaData()
BaseModel = declarative_base(metadata=metadata, cls=RowModel, metaclass=RowModelDeclarativeMeta)    

async def setup():
    metadata.bind = await create_engine(**database_settings)

class MyEntityManager(BaseModelManager):
    async def get_with_products(self):
        return await self.get_items(where_list=[(MyEntity.c.num_products > 0)])
    
class MyEntity(BaseModel):
    __tablename__ = 'my_entity'
    __model_manager_class__ = MyEntityManager

    id = Column(Integer, primary_key=True)
    name = Column(String(100), nullable=False)
    num_products = Column(Integer)
    created_at = Column(DateTime(), server_default=text('now()'), nullable=False)

Query:

objects = await MyEntity.objects.get_instances(
    where_list=[(MyEntity.c.name == 'foo')],
    order_by=['name', '-num_products']
)

num_objects = await MyEntity.objects.count(
    where_list=[(MyEntity.c.name == 'foo'), (MyEntity.c.num_products > 3)]
)

or (low-level):

objects = await MyEntity.objects \
    .set_sql(MyEntity.table.select()) \
    .where([(MyEntity.c.name == 'foo')]) \
    .limit(10) \
    .fetchall()

Management:

record = await MyEntity.objects.insert(
    name='bar',
    num_products=0,
)

await record.update(name='baz')
await record.delete()

Transactions:

async with MyEntity.objects.transaction() as my_entity_objects:
    await my_entity_objects.insert(name='bar', num_products=0)
    await my_entity_objects.delete([(MyEntity.c.name == 'foo')])