User Access Management (UAM), Role-Based (RBAC) Models [SQLAlchemy | Flask-SQLAlchemy Ready]

Data Models:

• User

• UserEmail

• Role

• Permission

• UserPermission

• Group

• GroupMember

• GroupPermission

Craft Data Clean (Object-Oriented) Way

from db.models import (
    User, UserEmail, UserPermission, Role, Permission
)
from db.models.utils import Session

session = Session()

dev_role = Role(name='PYTHON_DEV')
aws_permission = Permission(name='AWS_ACCOUNT')

session.add_all([dev_role, aws_permission])
session.commit()

prem = User(
    first_name = 'Premkumar',
    last_name = 'Chalmeti',
    username = 'premkumar30',
    avatar = 'https://avatars0.githubusercontent.com/u/22867776',
    bio = 'Software Craftsman | Python | Django | Elasticsearch | VueJS | AWS',
    role = dev_role.id,
    password = 'PBKDF2uuid5AWSPrem112'
)

session.add(prem)
session.commit()

emails = UserEmail(user_id=prem.id, primary_email='premkumar@gmail.com')

user_rights = UserPermission(
    role_id = dev_role.id,
    permission_id = aws_permission.id
)

session.add_all([emails, user_rights])
session.commit()

Flexible and Powerful Querying

1. RBAC

In [1]: session = Session()

In [2]: session.query(
         User.first_name, User.last_name, User.username, User.avatar, 
         User.bio, Role.name, User.last_login, UserEmail.primary_email
      ).join(UserEmail).join(Role).filter(
         User.username=='premkumar30'
      ).first()._asdict()
Out[3]: 
{'first_name': 'Premkumar',
 'last_name': 'Chalmeti',
 'username': 'premkumar30',
 'avatar': 'https://avatars0.githubusercontent.com/u/22867776',
 'bio': 'Software Craftsman | Python | Django | Elasticsearch | VueJS | AWS',
 'name': 'PYTHON_DEV',
 'last_login': None,
 'primary_email': 'premkumar@gmail.com'}

2. Group UAM

In [1]: from sqlalchemy import func

In [2]: from db.models import Group, GroupMember

In [3]: from db.models.utils import Session

In [4]: session = Session()

In [5]: session.query(
    Group.name, func.count(GroupMember.id).label('member_count')
).join(GroupMember).group_by(Group.name).first()._asdict()
Out[5]: {'name': 'DevOps Team', 'member_count': 1}