Pink Diary (Next.js New_version)

This is a modern Next.js rewrite of the original Pink Diary app. It supports Node 18/20/22 and uses MongoDB via Mongoose.

Features

• Sign up, log in (JWT cookie)
• Home calendar showing logged period days and simple predictions
• Log symptoms: flow, mood, pain, sex
• Help requests (basic stub)

Getting started (development)

1. Copy env and install dependencies:

Copy-Item .env.example .env.local
# Edit .env.local to set MONGODB_URI, JWT_SECRET, GEMINI_API_KEY (optional)
npm install
npm run dev

Dev server runs at http://127.0.0.1:3500

Required environment variables

Name	Purpose
MONGODB_URI	Connection string for MongoDB database
JWT_SECRET	Secret used to sign JWT auth cookie
GEMINI_API_KEY	(Optional) Key for AI advice/chat; if omitted, fallbacks show static tips

Provide these in .env.local for local dev or as host/container environment variables in production.

Production build

npm run build
npm run start

next start serves the optimized build on port 3500 as configured in package.json.

Docker deployment

Build and run the included Dockerfile:

docker build -t pink-diary:prod ./New_version
docker run -d -p 3500:3500 --name pink-diary ^
	-e MONGODB_URI="mongodb://host.docker.internal:27017/pink-diary-next" ^
	-e JWT_SECRET="replace_me" ^
	-e GEMINI_API_KEY="optional_key" ^
	pink-diary:prod

On Linux/macOS replace PowerShell line continuation ^ with \.

Vercel deployment

1. Push repository to GitHub.
2. In Vercel dashboard: New Project → Import → Select New_version directory as root.
3. Add Environment Variables (Production & Preview): MONGODB_URI, JWT_SECRET, optional GEMINI_API_KEY.
4. Deploy. Vercel will run npm install then next build automatically.

PM2 (optional) process manager

You can create an ecosystem.config.js like:

module.exports = {
	apps: [{
		name: 'pink-diary',
		script: 'npm',
		args: 'run start',
		env: { PORT: 3500 }
	}]
};

Then run:

pm2 start ecosystem.config.js
pm2 save

Health checks

• Mongo connectivity errors: ensure MONGODB_URI is reachable from host/container.
• JWT issues: confirm JWT_SECRET identical across all running instances.
• AI endpoints returning 404: supply GEMINI_API_KEY; otherwise static fallback advice is used.

Folder structure (production relevance)

• app/ – Next.js App Router pages & API routes
• models/ – Mongoose schemas
• lib/db.js – Lazy Mongo connection (pool size 10)
• Dockerfile – Multi-stage build producing minimal runtime image

Security notes

• JWT cookie is httpOnly; ensure NODE_ENV=production so Next sets secure flags when served over HTTPS.
• Never commit real secrets; keep only .env.example in version control.

Scaling tips

• Use MongoDB Atlas for managed DB (set MONGODB_URI).
• Add an index on Entry collection if query volume grows (e.g., { userId: 1, date: 1 }).
• Consider enabling image optimization domains in next.config.js if external avatars are added.

Troubleshooting

Symptom	Likely Cause	Fix
App boots but API 401	Missing JWT cookie	Log in; check JWT_SECRET consistency
Build fails on Mongo var	MONGODB_URI absent	Set in environment before build/start
AI endpoints 500/404	Missing/invalid Gemini key	Set GEMINI_API_KEY or rely on fallback
Container can’t reach DB	Network bridge / host mapping	Use host.docker.internal (Mac/Win) or link container

Deployment summary

Minimum for production: set env vars, npm run build, run npm run start (or use Docker image). For a managed platform, point root to New_version and configure environment variables.

Notes

• API routes live under app/api/* (App Router)
• Mongo connection is cached for hot reload
• Auth uses httpOnly JWT cookie token
• Styling uses Tailwind CSS; utility classes are available out of the box
• Fallback AI logic activates automatically when GEMINI_API_KEY missing