what is smart-drink-fridge fr?

Smart Drink Fridge is an inventory system for a drink fridge, built around a Raspberry Pi and a USB camera. The core idea is straightforward: you add your drinks and their barcodes to the system. When you take a bottle or can from the fridge, you hold its barcode in front of the camera. The scanner detects the barcode, reduces the stock count by one, and records the transaction with a timestamp. You can view the current stock and transaction history through a local web interface. The system separates products from barcodes, which allows some useful behavior. A single product can have multiple barcodes, each with its own action and quantity. For example, scanning a six-pack barcode adds six items to stock, while scanning an individual can barcode removes one. This means you can stock a multipack with a single scan and then track individual bottles or cans as they are consumed. When adding a new barcode, the system can look up product details using Open Food Facts, a public food product database. The returned name, brand, and packaging info are treated as suggestions you can edit before saving. Duplicate products can also be merged, combining their stock and barcodes into one entry. On the hardware side, you need a Raspberry Pi or another Linux system, a 1080p USB camera, and optionally a small buzzer for audio feedback after a successful scan. A 1080p camera is recommended because lower-resolution cameras did not reliably read barcodes during development. The buzzer connects to the Raspberry Pi's GPIO pins and provides a sound confirmation each time a scan succeeds. The web interface covers product and barcode management, current stock, transaction history, and consumption statistics over different time periods. It supports both German and English, and can send low-stock notifications through a service called Pushover. Scanner transactions can be cancelled through the web interface using a password. The project uses a SQLite database that is created automatically on first start. Docker support is included for easier deployment. The full README is longer than what was shown.