docker-compose.yml

version: "3.9"

# ─────────────────────────────────────────────────────────────────────────────
# SentimentPulse — docker-compose
#
# Infrastructure (Kafka, Redis) is defined here for standalone development.
# If you already have FinStream running, point KAFKA_BOOTSTRAP_SERVERS and
# REDIS_HOST at your existing containers via .env instead of starting these.
#
# Services:
#   Infrastructure:        zookeeper, kafka, redis, kafdrop
#   Phase 2 (this PR):     sentiment-ingestor
#   Phase 3 (next):        sentiment-classifier
#   Phase 4 (next):        sentiment-persistence, sentiment-api
#   Phase 5 (next):        sentiment-dashboard
# ─────────────────────────────────────────────────────────────────────────────

networks:
  sentimentpulse:
    driver: bridge

volumes:
  kafka_data:
  zookeeper_data:
  redis_data:
  timescaledb_data:

services:

  # ── Infrastructure ──────────────────────────────────────────────────────────

  zookeeper:
    image: confluentinc/cp-zookeeper:7.6.1
    container_name: sp-zookeeper
    networks: [sentimentpulse]
    environment:
      ZOOKEEPER_CLIENT_PORT: 2181
      ZOOKEEPER_TICK_TIME: 2000
    volumes:
      - zookeeper_data:/var/lib/zookeeper/data
    healthcheck:
      test: ["CMD", "echo", "ruok", "|", "nc", "localhost", "2181"]
      interval: 10s
      timeout: 5s
      retries: 5

  kafka:
    image: confluentinc/cp-kafka:7.6.1
    container_name: sp-kafka
    networks: [sentimentpulse]
    depends_on:
      zookeeper:
        condition: service_healthy
    ports:
      - "9092:9092"      # host access for local development
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: zookeeper:2181
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://kafka:29092,PLAINTEXT_HOST://localhost:9092
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
      KAFKA_GROUP_INITIAL_REBALANCE_DELAY_MS: 0
      KAFKA_AUTO_CREATE_TOPICS_ENABLE: "true"
    volumes:
      - kafka_data:/var/lib/kafka/data
    healthcheck:
      test: ["CMD", "kafka-broker-api-versions", "--bootstrap-server", "localhost:29092"]
      interval: 15s
      timeout: 10s
      retries: 10

  # Kafdrop — Kafka UI for verifying messages (Phase 2 exit criteria)
  kafdrop:
    image: obsidiandynamics/kafdrop:latest
    container_name: sp-kafdrop
    networks: [sentimentpulse]
    depends_on:
      kafka:
        condition: service_healthy
    ports:
      - "9000:9000"
    environment:
      KAFKA_BROKERCONNECT: kafka:29092
      JVM_OPTS: "-Xms32M -Xmx64M"

  redis:
    image: redis:7.2-alpine
    container_name: sp-redis
    networks: [sentimentpulse]
    ports:
      - "6379:6379"
    volumes:
      - redis_data:/data
    command: redis-server --appendonly yes
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 3s
      retries: 5

  # Kafka topic initialisation — creates raw.headlines with correct partition count
  kafka-init:
    image: confluentinc/cp-kafka:7.6.1
    container_name: sp-kafka-init
    networks: [sentimentpulse]
    depends_on:
      kafka:
        condition: service_healthy
    entrypoint: ["/bin/sh", "-c"]
    command: |
      "
      echo 'Creating Kafka topics...'
      kafka-topics --bootstrap-server kafka:29092 --create --if-not-exists \
        --topic raw.headlines \
        --partitions 4 \
        --replication-factor 1 \
        --config retention.ms=86400000

      kafka-topics --bootstrap-server kafka:29092 --create --if-not-exists \
        --topic enriched.sentiment \
        --partitions 12 \
        --replication-factor 1 \
        --config retention.ms=604800000

      kafka-topics --bootstrap-server kafka:29092 --create --if-not-exists \
        --topic sentiment.dlq \
        --partitions 2 \
        --replication-factor 1 \
        --config retention.ms=2592000000

      echo 'Topics created.'
      kafka-topics --bootstrap-server kafka:29092 --list
      "
    restart: "no"

  # ── Phase 2 — News Ingestor ─────────────────────────────────────────────────

  sentiment-ingestor:
    build:
      context: .
      dockerfile: sentiment_ingestor/Dockerfile
    container_name: sp-ingestor
    networks: [sentimentpulse]
    depends_on:
      kafka:
        condition: service_healthy
      redis:
        condition: service_healthy
      kafka-init:
        condition: service_completed_successfully
    environment:
      KAFKA_BOOTSTRAP_SERVERS: kafka:29092
      REDIS_HOST: redis
      REDIS_PORT: 6379
      EDGAR_USER_AGENT: "SentimentPulse dev@example.com"
      GNEWS_API_KEY: "${GNEWS_API_KEY:-}"    # optional — set in .env
      TICKERS_CSV_PATH: /app/data/sp500_tickers.csv
    restart: unless-stopped

  # ── Phase 3 — Sentiment Classifier ─────────────────────────────────────────

  sentiment-classifier:
    build:
      context: .
      dockerfile: sentiment_classifier/Dockerfile
    container_name: sp-classifier
    networks: [sentimentpulse]
    depends_on:
      kafka:
        condition: service_healthy
      kafka-init:
        condition: service_completed_successfully
    environment:
      KAFKA_BOOTSTRAP_SERVERS: kafka:29092
      MODEL_DIR: /app/models/finbert-finetuned
      SP500_PATTERNS_PATH: /app/data/sp500_patterns.jsonl
      SP500_TICKERS_CSV: /app/data/sp500_tickers.csv
      MODEL_VERSION: finbert-finetuned-v1.0
      PROMETHEUS_PORT: 8000
    ports:
      - "8000:8000"    # Prometheus /metrics
    restart: unless-stopped

  # ── Phase 4 — TimescaleDB ────────────────────────────────────────────────────

  timescaledb:
    image: timescale/timescaledb:latest-pg16
    container_name: sp-timescaledb
    networks: [sentimentpulse]
    ports:
      - "5432:5432"
    environment:
      POSTGRES_DB: sentimentpulse
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: postgres
    volumes:
      - timescaledb_data:/var/lib/postgresql/data
      - ./db/migrations/V1__create_sentiment_events.sql:/docker-entrypoint-initdb.d/V1__create_sentiment_events.sql
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U postgres -d sentimentpulse"]
      interval: 10s
      timeout: 5s
      retries: 10

  # ── Phase 4 — Persistence Worker ─────────────────────────────────────────────

  sentiment-persistence:
    build:
      context: .
      dockerfile: sentiment_persistence/Dockerfile
    container_name: sp-persistence
    networks: [sentimentpulse]
    depends_on:
      kafka:
        condition: service_healthy
      timescaledb:
        condition: service_healthy
      redis:
        condition: service_healthy
      kafka-init:
        condition: service_completed_successfully
    environment:
      KAFKA_BOOTSTRAP_SERVERS: kafka:29092
      DATABASE_URL: postgresql://postgres:postgres@timescaledb:5432/sentimentpulse
      REDIS_HOST: redis
      REDIS_PORT: 6379
    restart: unless-stopped

  # ── Phase 4 — Sentiment API ───────────────────────────────────────────────────

  sentiment-api:
    build:
      context: .
      dockerfile: sentiment_api/Dockerfile
    container_name: sp-api
    networks: [sentimentpulse]
    depends_on:
      timescaledb:
        condition: service_healthy
      redis:
        condition: service_healthy
    ports:
      - "8081:8081"    # avoids conflict with FinStream API on 8080
    environment:
      DATABASE_URL: postgresql://postgres:postgres@timescaledb:5432/sentimentpulse
      REDIS_HOST: redis
      REDIS_PORT: 6379
    restart: unless-stopped

  # ── Phase 5 — React Dashboard ────────────────────────────────────────────────

  sentiment-dashboard:
    build:
      context: .
      dockerfile: sentiment_dashboard/Dockerfile
    container_name: sp-dashboard
    networks: [sentimentpulse]
    depends_on:
      sentiment-api:
        condition: service_started
    ports:
      - "3001:3001"
    restart: unless-stopped