Skip to content

More Features

AutoDLA includes a variety of features to improve quality of life for the developer.

Let's create a base Object called User and work with it in this page to explore different concepts: 

from autodla import Object, primary_key
from autodla.dbs import PostgresDB, MemoryDB

class User(Object):
    id: primary_key = primary_key.auto_increment()
    name: str
    age: int

# MemoryDB stores everything in RAM. PostgresDB uses SQLite locally and
# syncs changes to PostgreSQL.
db = MemoryDB()  # or PostgresDB()
db.attach([User])

Data Generation

AutoDLA can automatically generate data using the DataGenerator class.

Usage: 

from autodla.utils import DataGenerator
this class includes a variety of staticmethods that allow you to generate distinct types of data: 
# Create 3 Users and save them into a list of users
users = []
for i in range(3):
    usr = User.new(
        name=DataGenerator.name(),
        age=DataGenerator.age()
    )
    users.append(usr)
This generates 3 users and saves them into the users variable

id name age
ee102517-ea1b-466f-8c2a-61d49c1e7425 Karen 20
c0e9b676-32dc-4add-b1a5-b60e464afd30 William 10
d4c12093-756a-4438-9bf1-c7fc0cb563ed Susan 15

Data Audit by design

Tables generated by AutoDLA are transactional tables by design. This allows for more in depth data analysis without the hussle of generating and mantaining this infrastructures.

For the example we are working with, in the DB the table public.users was generated defined with the next columns :

column_name data_type
dla_object_id uuid
dla_modified_at timestamp
dla_modified_by text
dla_operation text
dla_is_current boolean
dla_is_active boolean
id uuid
age integer
name text

These dla_* fields power AutoDLA's versioning and audit features. Every object tracks its modification history automatically.

The rows currently in the table are these:

id name age dla_object_id dla_modified_at dla_modified_by dla_operation dla_is_current dla_is_active
ee102517-ea1b-466f-8c2a-61d49c1e7425 Karen 20 b83e5016-acd7-476f-8b30-f19154cdbbbf '03-04-2025 00:00:00' SYSTEM INSERT True True
c0e9b676-32dc-4add-b1a5-b60e464afd30 William 10 a300842d-da73-41b0-bb97-998601254769 '03-04-2025 00:00:00' SYSTEM INSERT True True
d4c12093-756a-4438-9bf1-c7fc0cb563ed Susan 15 81b5ba22-f381-416f-865f-997c6970d664 '03-04-2025 00:00:00' SYSTEM INSERT True True

but if we apply some change, its going to be applied in the table, saving the changes 

# get Susan
usr = users[-1]
# modify Susan
usr.update(age=25)
Resulting in the table now being (see how a row was added and the previous last in the is_current field was modified):

id name age dla_object_id dla_modified_at dla_modified_by dla_operation dla_is_current dla_is_active
ee102517-ea1b-466f-8c2a-61d49c1e7425 Karen 20 b83e5016-acd7-476f-8b30-f19154cdbbbf '03-04-2025 00:00:00' SYSTEM INSERT True True
c0e9b676-32dc-4add-b1a5-b60e464afd30 William 10 a300842d-da73-41b0-bb97-998601254769 '03-04-2025 00:00:00' SYSTEM INSERT True True
d4c12093-756a-4438-9bf1-c7fc0cb563ed Susan 15 81b5ba22-f381-416f-865f-997c6970d664 '03-04-2025 00:00:00' SYSTEM INSERT False True
531a0f7b-fc95-4e71-90c9-ad7d7c00720d Susan 25 81b5ba22-f381-416f-865f-997c6970d664 '03-04-2025 00:01:15' SYSTEM UPDATE True True

This data can be accessed in 2 ways:

  • User.get_table_res(): returning the complete table for the model User
  • or for each user with the method usr.history(): returning the complete table for the specific instance

The working memory will have always the last version of each object:

id name age
ee102517-ea1b-466f-8c2a-61d49c1e7425 Karen 20
c0e9b676-32dc-4add-b1a5-b60e464afd30 William 10
d4c12093-756a-4438-9bf1-c7fc0cb563ed Susan 25

Automatic relationships

Because AutoDLA uses the Model abstraction instead of the Table abstraction (used by most ORMs), each Model can have multiple tables representing it's data. AutoDLA handles the background tables required to fullfil the specified relationship.

To show this, let's create a new Model: 

class Group(Object):
    id: primary_key = primary_key.auto_increment()
    group_name: str
    participants: list[User]
as you see, participants is a list of Users, to function AutoDLA is creating 2 tables for this model:

  • public.group: This represents the data that doesn't uses a relationship
  • public.group__participants__user: This creates a table for the specific relationship with the other Model

When the program memory is updated for you to use the python objects, a JOIN clause is automatically applied in the background.

In practice, once you setup the models, you can just use them as any other Python object, AutoDLA will handle everything. 

# Group creation
grp = Group.new(
    group_name = 'Group 1',
    participants = [
        users[0]
    ]
)

# Group modification: add another user
grp.update(participants=grp.participants + [users[1]])

# Reference the user from the group
usr = grp.participants[0]
print(usr)
### User( name=Karen , age=20 )

# Modify user
usr.update(age=usr.age + 1)
print(users[0]) # previously defined list
### User( name=Karen , age=21 )

Filtering

AutoDLA includes a built-in method to turn lambda functions into SQL Querys, thanks to this, to filter data in AutoDLA all you need to do is to write a function for the condition you want to apply

# get by name
filtered_users = User.filter(lambda x: x.name == 'Karen')

# get by age
filtered_users = User.filter(lambda x: x.age > 18)

# use python functions
filtered_users = User.filter(lambda x: x.name.startswith('K'))

# boolean expressions
filtered_users = User.filter(lambda x: x.name == 'Karen' and x.age > 18)

In-memory object integrity

Something that others ORMs fail to accomplish is the integrity between the observed objects in the DB and the object you are working with in memory, leading to data issues down the line like duplicated values and others. AutoDLA handles this integrity, the objects you are working with, are always the same as they are in the DB. 

u1 = users[0] # User ( name=Karen , age=21 )
u2 = grp.participants[0] # User ( name=Karen , age=21 )
u3 = User.filter(lambda x: x.name == 'Karen')[0] # User ( name=Karen , age=21 )

u1 == u2 # True
u1 == u3 # True
u2 == u3 # True
id(u1) == id(u2) #True

Auto-generated admin panel

AutoDLA can automatically generate an admin panel for CRUD operations, learn more about this in AutoDLA WEB