first commit

2026-07-26 21:12:38 -03:00
commit e6badb0d79
133 changed files with 38368 additions and 0 deletions
@@ -0,0 +1,120 @@
+name: leads-extractor
+
+services:
+  postgres:
+    image: postgres:18
+    restart: unless-stopped
+    environment:
+      POSTGRES_DB: ${POSTGRES_DB:-leads_extractor}
+      POSTGRES_USER: ${POSTGRES_USER:-leads}
+      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-leads}
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U \"$${POSTGRES_USER}\" -d \"$${POSTGRES_DB}\""]
+      interval: 5s
+      timeout: 5s
+      retries: 20
+      start_period: 10s
+    volumes:
+      - postgres_data:/var/lib/postgresql
+    networks:
+      - internal
+
+  backend:
+    build:
+      context: ./backend
+      dockerfile: Dockerfile
+    restart: unless-stopped
+    depends_on:
+      postgres:
+        condition: service_healthy
+    environment:
+      SERVER_HOST: 0.0.0.0
+      SERVER_PORT: 8080
+      FRONTEND_ORIGIN: ${FRONTEND_URL:-http://localhost:8090}
+      DATABASE_URL: postgres://${POSTGRES_USER:-leads}:${POSTGRES_PASSWORD:-leads}@postgres:5432/${POSTGRES_DB:-leads_extractor}
+      MEDIA_DIR: /data/media
+      RUST_LOG: ${RUST_LOG:-backend=info}
+
+      AUTH_USERNAME: ${AUTH_USERNAME}
+      AUTH_PASSWORD: ${AUTH_PASSWORD}
+      AUTH_SESSION_TTL_SECS: ${AUTH_SESSION_TTL_SECS:-43200}
+      AUTH_COOKIE_SECURE: ${AUTH_COOKIE_SECURE:-false}
+      AUTH_COOKIE_SAME_SITE: ${AUTH_COOKIE_SAME_SITE:-strict}
+
+      RATE_LIMIT_ENABLED: ${RATE_LIMIT_ENABLED:-true}
+      RATE_LIMIT_BURST_SIZE: ${RATE_LIMIT_BURST_SIZE:-120}
+      RATE_LIMIT_PERIOD_MS: ${RATE_LIMIT_PERIOD_MS:-200}
+      RATE_LIMIT_LOGIN_BURST_SIZE: ${RATE_LIMIT_LOGIN_BURST_SIZE:-5}
+      RATE_LIMIT_LOGIN_PERIOD_SECS: ${RATE_LIMIT_LOGIN_PERIOD_SECS:-30}
+
+      OPENAI_API_KEY: ${OPENAI_API_KEY}
+      OPENAI_MODEL: ${OPENAI_MODEL:-gpt-5.6-luna}
+      OPENAI_BASE_URL: ${OPENAI_BASE_URL:-https://api.openai.com/v1}
+      OPENAI_TIMEOUT_SECS: ${OPENAI_TIMEOUT_SECS:-120}
+      OPENAI_MAX_RETRIES: ${OPENAI_MAX_RETRIES:-6}
+      OPENAI_INPUT_COST_PER_1M_USD: ${OPENAI_INPUT_COST_PER_1M_USD:-0.15}
+      OPENAI_OUTPUT_COST_PER_1M_USD: ${OPENAI_OUTPUT_COST_PER_1M_USD:-0.60}
+      OPENAI_MONTHLY_BUDGET_USD: ${OPENAI_MONTHLY_BUDGET_USD:-50.00}
+
+      DATAIMPULSE_PROXY_ENABLED: ${DATAIMPULSE_PROXY_ENABLED:-true}
+      DATAIMPULSE_PROXY_SCHEME: ${DATAIMPULSE_PROXY_SCHEME:-http}
+      DATAIMPULSE_PROXY_HOST: ${DATAIMPULSE_PROXY_HOST:-gw.dataimpulse.com}
+      DATAIMPULSE_PROXY_PORT: ${DATAIMPULSE_PROXY_PORT:-823}
+      DATAIMPULSE_PROXY_USERNAME: ${DATAIMPULSE_PROXY_USERNAME}
+      DATAIMPULSE_PROXY_PASSWORD: ${DATAIMPULSE_PROXY_PASSWORD}
+      DATAIMPULSE_PROXY_COUNTRY: ${DATAIMPULSE_PROXY_COUNTRY:-br}
+      DATAIMPULSE_PROXY_STICKY: ${DATAIMPULSE_PROXY_STICKY:-false}
+
+      WORKER_CONCURRENCY: ${WORKER_CONCURRENCY:-8}
+      BROWSER_CONCURRENCY: ${BROWSER_CONCURRENCY:-4}
+      JOB_POLL_INTERVAL_MS: ${JOB_POLL_INTERVAL_MS:-750}
+      REQUEST_TIMEOUT_SECS: ${REQUEST_TIMEOUT_SECS:-45}
+      BROWSER_TIMEOUT_SECS: ${BROWSER_TIMEOUT_SECS:-75}
+      BROWSER_NO_SANDBOX: ${BROWSER_NO_SANDBOX:-true}
+      BROWSER_MIN_NAVIGATION_DELAY_MS: ${BROWSER_MIN_NAVIGATION_DELAY_MS:-2500}
+      BROWSER_MAX_NAVIGATION_DELAY_MS: ${BROWSER_MAX_NAVIGATION_DELAY_MS:-15000}
+      BROWSER_MACRO_PAUSE_EVERY_MIN: ${BROWSER_MACRO_PAUSE_EVERY_MIN:-15}
+      BROWSER_MACRO_PAUSE_EVERY_MAX: ${BROWSER_MACRO_PAUSE_EVERY_MAX:-20}
+      BROWSER_MACRO_PAUSE_MIN_SECS: ${BROWSER_MACRO_PAUSE_MIN_SECS:-30}
+      BROWSER_MACRO_PAUSE_MAX_SECS: ${BROWSER_MACRO_PAUSE_MAX_SECS:-90}
+      CRAWL_MAX_DEPTH: ${CRAWL_MAX_DEPTH:-5}
+      CRAWL_MAX_PAGES_PER_INTERVIEWEE: ${CRAWL_MAX_PAGES_PER_INTERVIEWEE:-250}
+      MAX_INTERVIEWEES_PER_RUN: ${MAX_INTERVIEWEES_PER_RUN:-0}
+    volumes:
+      - media_data:/data/media
+    ports:
+      - "${BACKEND_PORT:-8081}:8080"
+    healthcheck:
+      # /api/health exige sessão autenticada; 401 já indica que o servidor
+      # está de pé e respondendo (só uma requisição de rede falhando é erro).
+      test: ["CMD-SHELL", "curl -s -o /dev/null -w '%{http_code}' http://127.0.0.1:8080/api/health | grep -qE '^(200|401)$'"]
+      interval: 10s
+      timeout: 5s
+      retries: 15
+      start_period: 10s
+    networks:
+      - internal
+
+  frontend:
+    build:
+      context: ./frontend
+      dockerfile: Dockerfile
+      args:
+        VITE_API_URL: ${BACKEND_URL:-http://localhost:8081}
+        VITE_API_POLL_INTERVAL: ${VITE_API_POLL_INTERVAL:-6000}
+        VITE_DEMO_MODE: ${VITE_DEMO_MODE:-false}
+    restart: unless-stopped
+    depends_on:
+      backend:
+        condition: service_healthy
+    ports:
+      - "${FRONTEND_PORT:-8080}:8080"
+    networks:
+      - internal
+
+networks:
+  internal:
+
+volumes:
+  postgres_data:
+  media_data: