Observed since 21 Oct, 11:43 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
Activealchemy is a Python library that simplifies database interactions by providing an Active Record pattern and automatic session management. It supports both synchronous and asynchronous operations, making it flexible for various application needs.
pip install activealchemy
Activealchemy uses a configuration schema to manage database connection details. Here's an example using the PostgreSQLConfigSchema:
from activealchemy.config import PostgreSQLConfigSchema
config = PostgreSQLConfigSchema(
db="mydatabase",
user="myuser",
password="mypassword",
host="localhost",
port=5432,
)
See mise.local.toml and other examples for different configuration options.
The ActiveEngine class manages database connections. You can set it up for synchronous or asynchronous operations:
Synchronous:
from activealchemy import ActiveEngine
engine = ActiveEngine(config)
Asynchronous:
from activealchemy.aio import ActiveEngine
engine = ActiveEngine(config)
Set the engine for your models:
from activealchemy import ActiveRecord
from myapp.models import MyModel # Assuming 'MyModel' inherits from ActiveRecord
MyModel.set_engine(engine)
from myapp.models import User
# Method 1: Create and save.
user = User(name="Alice", email="alice@example.com")
user.save()
# Method 2: Using add.
user = User(name="Bob")
User.add(user, commit=True) # commit=True commits the change immediately.
# Retrieve all users
users = User.all()
# Find a user by primary key.
user_pk = uuid.UUID("some uuid here")
user = User.find(user_pk) # using PKMixin which adds find().
# Find a user by email
user = User.find_by(email="alice@example.com")
# Find users with a specific name
users = User.where(User.name == "Alice").all()
# Find the first user in the database.
first_user = User.first() # Orders by the primary key.
# Find the last user in the database, ordered by a column.
last_user = User.last(User.name) # Orders by User.name DESC.
# Use a custom query with advanced selects
query = User.select().where(User.created_at < some_date)
users = User.all(query=query) # Can combine custom queries with other operations
# Find the first created user
first_created_user = User.first_created()
# Find the last created user
last_created_user = User.last_created()
# Count users
user_count = User.count()
user = User.find_by(name="Alice")
if user:
user.name = "Alicia"
user.save()
user = User.find_by(name="Bob")
if user:
User.delete(user)
Activealchemy supports asynchronous operations using asyncio and async/await. Use activealchemy.aio.ActiveRecord, activealchemy.aio.ActiveEngine, and activealchemy.aio.Schema:
import asyncio
from activealchemy.aio import ActiveRecord, Schema
# ... (setup engine as described above, using the async ActiveEngine) ...
async def create_user():
user = User(name="Dave")
await user.save()
print(f"Created user: {user}")
asyncio.run(create_user())
# Use async methods for querying and operations
users = await User.all()
# Convert model instances to dictionaries and vice-versa
user_data = user.dump_model()
restored_user = User.load(**user_data)
schema = User.Schema() # Convert schemas to model instances and back
user_instance = schema.to_model(User)
user_schema = User.Schema.model_validate(user_instance)
See activealchemy/demo/amodels.py for a detailed example.
session = User.new_session()
try:
user = User(name='Bob')
session.add(user)
session.commit()
finally:
session.close()
# async
async def async_example():
async_session = await User.new_session()
try:
user = User(name='Bob')
async_session.add(user)
await async_session.commit()
finally:
await async_session.close()
async def context_session():
async with User.new_session() as session:
user = User(name='Bob')
async_session.add(user)
await async_session.commit(session)
# or
user2 = User(name='Alice')
user2 = await Use.add(user2, commit=True, session=session)
Transactions are automatically handled for add(), delete(), and save().
For fine-grained control use a context manager with begin() and begin_nested():
with User.new_session() as session
try:
# Operations here
# Rollback will happen automatically
session.add(user1)
session.add(user2)
1 / 0 # Will rollback to here
except Exception as e:
print(f"Error: {e}") # Will rollback to here
session.commit()
## Session is closed on context exit
See activealchemy/utils/retry.py for how the Q context manager implements retries and transactions.
Create custom schemas by inheriting from activealchemy.Schema or activealchemy.aio.Schema and using Pydantic's features.
For more comprehensive examples, refer to the following:
activealchemy/demo for sample models and usage.models.py and amodels.py for synchronous and asynchronous model definitions.This documentation provides a starting point for using Activealchemy. Explore the code examples and docstrings for more in-depth information.
Download counts and popularity will appear here as the collector receives them.
| Version | Published | Direct dependencies | Open version |
|---|---|---|---|
| 0.2.0 | 18 Apr 2025 | 5 |
This describes our collected data; it is not a security assessment.
Observed since 21 Oct, 11:43 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
package:13f33f883c3e4adecc2482a89921308d68b6eb851a396a5cdaa01408947b27ec