Asynchronous SQLAlchemy Adapter is the SQLAlchemy adapter for PyCasbin. With this library, Casbin can load policy from SQLAlchemy supported database or save policy to it.
Based on Officially Supported Databases, The current supported databases are:
- PostgreSQL
- MySQL
- MariaDB
- SQLite
- Oracle
- Microsoft SQL Server
- Firebird
pip install casbin_async_sqlalchemy_adapter
import casbin_async_sqlalchemy_adapter
import casbin
adapter = casbin_async_sqlalchemy_adapter.Adapter('sqlite+aiosqlite:///test.db')
# or mysql example
# adapter = casbin_async_sqlalchemy_adapter.Adapter('mysql+aiomysql://user:pwd@127.0.0.1:3306/exampledb')
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
sub = "alice" # the user that wants to access a resource.
obj = "data1" # the resource that is going to be accessed.
act = "read" # the operation that the user performs on the resource.
if e.enforce(sub, obj, act):
# permit alice to read data1
pass
else:
# deny the request, show an error
passNote that AsyncAdapter must be used for AsyncEnforcer.
For production applications, you'll want to manage database schema using Alembic migrations instead of calling create_table() at runtime. The adapter provides create_casbin_rule_model() to integrate with your existing migration workflow.
# In your alembic/env.py or models file
from casbin_async_sqlalchemy_adapter import create_casbin_rule_model
from sqlalchemy.ext.declarative import declarative_base
# Use your application's declarative base
Base = declarative_base()
# Create the CasbinRule model using your base
CasbinRule = create_casbin_rule_model(Base)
# Now Alembic can auto-generate migrations for the casbin_rule table
# Run: alembic revision --autogenerate -m "Add casbin_rule table"
# Then: alembic upgrade headWhen using the adapter with Alembic-managed tables, pass your custom model:
from your_app.models import CasbinRule
import casbin_async_sqlalchemy_adapter
import casbin
adapter = casbin_async_sqlalchemy_adapter.Adapter(
'sqlite+aiosqlite:///test.db',
db_class=CasbinRule
)
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)The adapter supports using externally managed SQLAlchemy sessions. This feature is useful for:
- Better transaction control in complex scenarios
- Reducing database connections and communications
- Supporting advanced database features like two-phase commits
- Integrating with existing database session management
import casbin_async_sqlalchemy_adapter
import casbin
from sqlalchemy.ext.asyncio import create_async_engine, AsyncSession
from sqlalchemy.orm import sessionmaker
# Create your own database session
engine = create_async_engine('sqlite+aiosqlite:///test.db')
async_session = sessionmaker(engine, class_=AsyncSession, expire_on_commit=False)
# Create adapter with external session
session = async_session()
adapter = casbin_async_sqlalchemy_adapter.Adapter(
'sqlite+aiosqlite:///test.db',
db_session=session
)
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
# Now you have full control over the session
# The adapter will not auto-commit or auto-rollback when using external sessions# Example: Manual transaction control
async with async_session() as session:
adapter = casbin_async_sqlalchemy_adapter.Adapter(
'sqlite+aiosqlite:///test.db',
db_session=session
)
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
# Add multiple policies in a single transaction
await e.add_policy("alice", "data1", "read")
await e.add_policy("bob", "data2", "write")
# Commit or rollback as needed
await session.commit()# Example: Efficient batch operations
async with async_session() as session:
adapter = casbin_async_sqlalchemy_adapter.Adapter(
'sqlite+aiosqlite:///test.db',
db_session=session
)
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
# Batch add multiple policies efficiently
policies = [
["alice", "data1", "read"],
["bob", "data2", "write"],
["carol", "data3", "read"]
]
await e.add_policies(policies)
# Commit the transaction
await session.commit()The adapter provides a clear_policy() method to remove all policy records from the database directly:
import casbin_async_sqlalchemy_adapter
import casbin
from sqlalchemy.ext.asyncio import create_async_engine
# Setup
engine = create_async_engine('sqlite+aiosqlite:///test.db')
adapter = casbin_async_sqlalchemy_adapter.Adapter(engine)
await adapter.create_table()
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
await e.load_policy()
# Add some policies
await e.add_policy("alice", "data1", "read")
await e.add_policy("bob", "data2", "write")
# Clear all policies from the database
await adapter.clear_policy()
# Reload to verify - the enforcer will have no policies
await e.load_policy()When soft deletion is enabled, clear_policy() marks all records as deleted instead of physically removing them.
The adapter supports soft deletion, which marks records as deleted instead of physically removing them from the database. This is useful for:
- Maintaining audit trails
- Implementing undo functionality
- Preserving historical data
- Debugging and compliance requirements
To enable soft deletion, you need to:
- Create a custom database model with a boolean
is_deletedcolumn - Pass the soft delete attribute to the adapter
import casbin_async_sqlalchemy_adapter
import casbin
from sqlalchemy import Column, Boolean, Integer, String
from sqlalchemy.ext.asyncio import create_async_engine
# Define a custom model with soft delete support
class CasbinRuleSoftDelete(casbin_async_sqlalchemy_adapter.Base):
__tablename__ = "casbin_rule"
id = Column(Integer, primary_key=True)
ptype = Column(String(255))
v0 = Column(String(255))
v1 = Column(String(255))
v2 = Column(String(255))
v3 = Column(String(255))
v4 = Column(String(255))
v5 = Column(String(255))
# Add the soft delete column
is_deleted = Column(Boolean, default=False, index=True, nullable=False)
# Create adapter with soft delete support
engine = create_async_engine('sqlite+aiosqlite:///test.db')
adapter = casbin_async_sqlalchemy_adapter.Adapter(
engine,
db_class=CasbinRuleSoftDelete,
db_class_softdelete_attribute=CasbinRuleSoftDelete.is_deleted
)
# Create the table
await adapter.create_table()
e = casbin.AsyncEnforcer('path/to/model.conf', adapter)
# When you delete a policy, it will be soft-deleted (marked as deleted)
await e.delete_permission_for_user("alice", "data1", "read")
# The record remains in the database with is_deleted=True
# Load policy will automatically filter out soft-deleted records
await e.load_policy()When soft deletion is enabled:
- Delete operations set the
is_deletedflag toTrueinstead of removing records - Load operations automatically filter out records where
is_deleted=True - Save policy marks removed rules as deleted while preserving the records
- Update operations only affect non-deleted records
This feature maintains full backward compatibility - when db_class_softdelete_attribute is not provided, the adapter functions with hard deletion as before.
This project is licensed under the Apache 2.0 license.