django-dynamic-query 0.1.0

A dynamic ORM query builder for Django with NLP support

Django Dynamic Query

Secure AI-powered Dynamic Django ORM Query Builder

django-dynamic-query is a Django package that allows developers to safely generate Django ORM queries from natural language using AI models such as OpenAI or Ollama.

Instead of writing complex filtering logic manually, users can ask questions in plain English and the package converts them into validated Django ORM queries while protecting against unsafe operations.

---

Features

• AI-powered natural language querying
• Secure ORM query generation
• Schema indexing using Django management command
• OpenAI support
• Ollama support
• Built-in validation layer
• Protection against dangerous ORM operations
• Easy Django integration
• Simple web dashboard
• Extensible architecture

---

Installation

Install using pip:

pip install django-dynamic-query

or

pip install git+https://github.com/sajjad-ch/django-dynamic-query.git

---

Requirements

• Python 3.10+
• Django 4.2+

---

Add to INSTALLED_APPS

Open your project's settings.py

INSTALLED_APPS = [
    ...
    "django_dynamic_query",
]

---

Configuration Add the following configuration to your settings.py to enable and customize the AI-powered features:

DJANGO_DYNAMIC_QUERY = {
    # Set to True to enable AI processing (Natural Language to ORM)
    "ENABLE_NLP": True,
    
    # Choose 'local' for Ollama/Local LLMs or 'api' for OpenAI/Cloud services
    "MODEL_TYPE": "local", 
    
    # Required for 'local': The name of the model (e.g., 'gemma2', 'llama3')
    "MODEL_NAME": "llama3",
    
    # Optional for 'local': Absolute path to a model file if not using a registry
    "MODEL_PATH": None,
    
    # Required for 'api': Your API Key
    "AI_API_KEY": "sk-...",
}

---

Index Your Models

Before using AI queries, index your Django models.

Run:

python manage.py index_schema

This command scans your Django models and builds the schema used by the AI engine.

Whenever your models change, run the command again.

---

URLs

Add the package URLs to your project.

from django.urls import include, path

urlpatterns = [
    ...
    path(
        "dynamic-query/",
        include("django_dynamic_query.urls")
    ),
]

---

Dashboard

After adding the URLs, open:

http://localhost:8000/dynamic-query/dashboard

The dashboard allows testing natural language queries directly from the browser.

---

Basic Usage

Example:

from django_dynamic_query import ...

# Your code here

Natural language example:

Show all active users created this month.

The package converts the request into a validated Django ORM query.

---

Security

The package is designed with security in mind.

Validation includes:

• Preventing dangerous ORM operations
• Validating model names
• Validating field names
• Safe filter generation
• Query sanitization

---

Supported AI Providers

Currently supported:

• OpenAI
• Ollama

The architecture allows adding additional providers in the future.

---

Project Structure

django_dynamic_query/

├── core/
│   ├── builder.py
│   ├── validators.py
│   └── exceptions.py
│
├── nlp/
│   ├── ai_handler.py
│   └── schema_rag.py
│
├── management/
│   └── commands/
│       └── index_schema.py
│
├── templates/
│
├── views.py
├── urls.py
└── utils.py

---

Running Tests

Run all tests:

python -m pytest

or

python runtests.py

---

Development Installation

Clone the repository:

git clone https://github.com/sajjad-ch/django-dynamic-query.git

Install dependencies:

pip install -r requirements.txt

Run tests:

pytest

---

Contributing

Contributions are welcome.

If you would like to improve the package:

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Open a Pull Request

---

License

This project is licensed under the MIT License.

See the LICENSE file for details.

---

Roadmap

Future improvements include:

• Additional AI providers
• Better prompt optimization
• Query explanation
• Query caching
• Admin integration
• Multi-database support

---

Author

Developed with ❤️ for the Django community.