Lafzloom

A poetry/shayari platform built with Django, DRF, PostgreSQL, and TailwindCSS. Users can browse, like, save, share, and submit shayaris. Admins can moderate submissions. A translation API supports Hindi ↔ English ↔ Urdu with pluggable providers.

Tech Notes

• This project targets Django 6.x as requested.
• Database migrations are standard Django migrations (work with PostgreSQL). If you truly need MongoDB, you will need a different backend (e.g., Djongo) and a separate migration strategy.

Features

• Home page with latest shayaris and categories
• Browse/search/filter by category, popularity, or author
• Shayari detail with likes, save, share, copy
• User submission flow with moderation queue
• Admin moderation panel and Django admin
• REST API for shayari CRUD and translation
• JWT auth for API + session auth for web
• Email/password registration and login without OTP verification

Local Setup

1) Create virtualenv and install deps

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2) Configure environment

cp .env.example .env

Update values as needed.

3) Run migrations

python manage.py migrate

4) Seed sample data (optional)

python manage.py seed_data

Demo user: demo / demo1234

5) Create admin

python manage.py createsuperuser

6) Start server

python manage.py runserver

Open: http://127.0.0.1:8000/

Docker

docker compose up --build

Then open http://localhost/

Deploy on Render (Web Service + Render PostgreSQL)

This repo now includes render.yaml, so you can deploy from Render Blueprint.

If you prefer manual setup, use these Render settings:

• Build Command: pip install -r requirements.txt
• Start Command: ./scripts/entrypoint.sh
• Health Check Path: /healthz/

Set environment variables in Render:

• DEBUG=0
• SECRET_KEY=<long-random-secret>
• DATABASE_URL=<Render PostgreSQL Internal Database URL>
• ALLOWED_HOSTS=<your-render-domain> (required only when using a custom domain)
• CSRF_TRUSTED_ORIGINS=https://<your-render-domain> (required only when using a custom domain)
• EMAIL_BACKEND=django.core.mail.backends.smtp.EmailBackend (for password-reset email links)
• EMAIL_HOST, EMAIL_PORT, EMAIL_HOST_USER, EMAIL_HOST_PASSWORD, DEFAULT_FROM_EMAIL
• EMAIL_TIMEOUT=15
• WEB_CONCURRENCY=1
• GUNICORN_TIMEOUT=120

Recommended security vars:

• SESSION_COOKIE_SECURE=1
• CSRF_COOKIE_SECURE=1
• SECURE_SSL_REDIRECT=1
• SECURE_HSTS_SECONDS=31536000
• SECURE_HSTS_INCLUDE_SUBDOMAINS=1
• SECURE_HSTS_PRELOAD=1

If you attach a custom domain, add it to both ALLOWED_HOSTS and CSRF_TRUSTED_ORIGINS.

For uploaded media persistence, add a Render Disk mounted at:

• /opt/render/project/src/media

API Endpoints

• GET /api/shayaris/ - list shayaris (filters: q, category, author, sort=popular|latest|oldest)
• POST /api/shayaris/ - create shayari (auth required)
• POST /api/shayaris/{id}/like/ - like/unlike (auth required)
• POST /api/shayaris/{id}/save/ - save/unsave (auth required)
• GET /api/categories/
• POST /api/translate/

JWT:

• POST /api/token/ (username/password)
• POST /api/token/refresh/

Translation Integration

The translation endpoint is pluggable. Choose one of the providers below.

Option A: HuggingFace Inference API (default LLM provider)

Set environment variables:

TRANSLATION_PROVIDER=hf
HF_API_TOKEN=your_hf_token

Optional overrides per language pair:

HF_MODEL_HI_EN=Helsinki-NLP/opus-mt-hi-en
HF_MODEL_EN_HI=Helsinki-NLP/opus-mt-en-hi
HF_MODEL_HI_UR=Helsinki-NLP/opus-mt-hi-ur
HF_MODEL_UR_HI=Helsinki-NLP/opus-mt-ur-hi
HF_MODEL_EN_UR=Helsinki-NLP/opus-mt-en-ur
HF_MODEL_UR_EN=Helsinki-NLP/opus-mt-ur-en

The backend calls the HuggingFace Inference API and returns translation as JSON.

Option B: Custom LLM HTTP Provider

If you have your own LLM endpoint, set:

TRANSLATION_PROVIDER=http
LLM_API_URL=https://your-llm.example.com/translate
LLM_API_KEY=your-key

The endpoint should accept JSON:

{
  "text": "...",
  "source_lang": "hi",
  "target_lang": "en",
  "style": "poetic"
}

And return:

{ "translation": "..." }

Triggering Translation (Frontend)

Buttons with .btn-translate call /api/translate/ via fetch. The JS lives in:

• static/js/app.js

It cycles translation targets based on the shayari language and updates the UI on success.

Security Notes

• Use strong SECRET_KEY and set DEBUG=0 in production.
• Configure ALLOWED_HOSTS and CSRF_TRUSTED_ORIGINS.
• Session and CSRF cookies can be secured via env vars in settings.py.

Tests

Basic tests can be added in shayari/tests.py. (Optional)

---