Document Kafka Streams implementation architecture and patterns

[Copilot]

Added technical documentation for the Kafka Streams implementation in quarkus-api, which serves as the real-time data processing and caching layer between Flink jobs and the frontend.

Documentation Added

Kafka Streams Architecture (quarkus-api/src/main/java/com/ververica/composable_job/quarkus/kafka/streams/README.md)

• KafkaStreamsTopology: Stream processing pipeline consuming inventory-events, maintaining materialized KTable for product cache, processing chat/data point streams
• ProductCacheService: Dual-source cache manager handling full products (from products topic) and partial updates (from inventory events), implements intelligent merge logic to prevent race conditions
• ProcessingFanoutProcessor: Custom processor for bounded state stores with automatic trimming
• CacheQueryService: Interactive Queries interface for state store access

Key Patterns Documented

• CQRS read model via KTable materialization
• Event sourcing from inventory event streams
• Race condition mitigation: filters PRODUCT_ADDED events, merges partial updates only into existing products
• WebSocket fanout for real-time client synchronization

Integration Overview

Flink → inventory-events → KafkaStreamsTopology → KTable (products-cache)
                                ↓
                        ProductCacheService (merge logic)
                                ↓
                        WebSocket → Frontend

Main README (quarkus-api/README.md)

• Added architecture overview and Kafka Streams integration summary
• Linked to detailed documentation

Original prompt

please provide a short explanation of the Kafka streams code in this directory

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.