☕ Barrista Admin

Admin panel for managing coffee and drink recipes, dishes, categories, and brew methods. Supports multiple languages with translation via OpenAI API.

Features

• Login — username and password (JWT, protected routes)
• Drinks — full CRUD, ingredients, instructions, categories, dish, photo
• Dishes — CRUD, description, volume, photo
• Categories — CRUD, photo
• Brew methods — CRUD (title, description, main info, how to prepare, pro tips, common mistakes), photo
• Multi-language — separate SQLite DB per language (barrista_en.db, barrista_ru.db, etc.)
• Translations — translate from English to other languages via OpenAI API (gpt-4o-mini, batched for cost efficiency); configurable language list and “override existing” or merge
• Dashboard
  • JSON — import JSON (as any configured language), export DB or JSON for selected languages, check translation integrity (compare translated DBs with en), delete translated DB
  • Images — import drink/category/dish/brew method images (by filename convention), download all images as ZIP
  • Translations — select entities (brew methods, drinks, categories, dishes), target language, override or merge, Start translation with progress
• Translation languages — separate tab to configure which languages appear in the translation selector (code + English label)
• SQLite — all data in server/data/; easy backup and deploy with a volume

Installation

1. Install dependencies:

npm install

2. Configure environment. Copy the example and edit:

cp server/.env.example .env

Or create .env in the project root. Set:

• ADMIN_USER — login (case-insensitive)
• ADMIN_PASSWORD — password
• JWT_SECRET — secret for JWT (use a long random string in production)
• OPENAI_API_KEY — required for translations (Dashboard → Start translation). Get a key at OpenAI API.

Optional: JWT_EXPIRES_IN (default 7d).

3. Start the app (API and frontend together):

npm run dev

4. Open in the browser:
   • Frontend: http://localhost:5173
   • API: http://localhost:3001

Log in with credentials from .env.

Data structure

Drink

• id, title, ingredients (array of { name, amount?, unit? }), instructions (array of strings), dishId, portionsAmount, categories (array of category ids).

Dish

• id, title, description, volume.

Category

• id, title. A drink’s categories is an array of category ids.

Brew method

• id, title, description, info (coffee, water, temperature, time), howToPrepare, proTips, commonMistakes (arrays of strings).

Usage

Dashboard

• JSON

  • Import as — choose language (Main en or any from Translation languages), then Import JSON to create/overwrite barrista_<lang>.db.
  • Languages (for export) — checkboxes for existing DBs; Export selected databases (download .db files) or Export JSON for selected languages.
  • Check translation integrity — compares each translated DB with barrista_en.db; report shows “All good” or missing entities per DB; log is shown and downloaded as .txt.
  • Delete translated DB — select a non‑en language and delete its DB file (en cannot be deleted).

• Images

  • Import by filename: drink_<id>.jpg, category_<id>.png, dish_<id>.png, brew_<id>.png. Download all images saves a ZIP of server/data/images/.

• Translations

  • Check which to translate: Brew methods, Drinks, Categories, Dishes.
  • Choose Language (from Translation languages tab).
  • Override existing — if checked, target tables are cleared and refilled; if unchecked, only missing entities are added.
  • Start translation — runs OpenAI translation from barrista_en.db into the selected language; progress is shown next to the button. Requires OPENAI_API_KEY in .env.

Translation languages (tab)

• Add/edit/remove languages (code + English label). This list is used in the Dashboard translation selector and in the JSON “Import as” dropdown. Save to server/data/translation-languages.json.

Export

• Dashboard: download DB or JSON for selected languages.
• Drinks / Dishes / Categories / Brew methods: each section has its own management and data; export is via Dashboard JSON per language.

Build and run (single port)

npm run build
npm run start

Open http://localhost:3001 — admin UI and API from one port. Data in server/data/ (SQLite files and images).

Deploy (e.g. Railway)

1. Push to GitHub and create a project on Railway; deploy from the repo (Dockerfile builds frontend and runs the Node server).
2. Variables: set ADMIN_USER, ADMIN_PASSWORD, JWT_SECRET, and OPENAI_API_KEY (for translations).
3. Persistent volume — mount a volume at /app/server/data so DBs and images survive redeploys.
4. Generate a domain in Settings → Networking.

The app uses PORT from the platform. With the volume mounted, barrista_*.db and images persist.

Scripts

• npm run dev — API + Vite dev server
• npm run build — TypeScript + Vite build
• npm run start — run production server
• npm run sanitize-db — sanitize main DB (see script)
• npm run import-dishes / import-categories — import from JSON in repo
• npm run translate-db-es — CLI script to build barrista_es.db from en (uses same OpenAI translator if OPENAI_API_KEY is set)

Tech stack

• Frontend: React 18, TypeScript, Vite, Lucide React, React Router
• Backend: Node.js, Express, SQLite (better-sqlite3), JWT, dotenv, OpenAI (translations), multer, archiver