chore: Remove obsolete build and GUI files, and introduce a new Docker Compose setup for microservices architecture. This change streamlines the project structure by eliminating legacy scripts and integrating a PostgreSQL database service alongside extraction and cleaning services.

2026-07-03 18:33:50 -05:00
parent 78df190d27
commit b55ad74539
75 changed files with 1699 additions and 8514 deletions
@@ -0,0 +1,77 @@
+services:
+  database:
+    image: postgres:16-alpine
+    container_name: ipsextractor-db
+    environment:
+      POSTGRES_DB: ipsextractor
+      POSTGRES_USER: ipsextractor
+      POSTGRES_PASSWORD: ipsextractor
+    ports:
+      - "5432:5432"
+    volumes:
+      - ipsextractor_db:/var/lib/postgresql/data
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U ipsextractor -d ipsextractor"]
+      interval: 5s
+      timeout: 5s
+      retries: 10
+
+  extraction:
+    build:
+      context: ./services/extraction
+      dockerfile: Dockerfile
+    container_name: ipsextractor-extraction
+    depends_on:
+      database:
+        condition: service_healthy
+    environment:
+      DATABASE_URL: postgresql+psycopg://ipsextractor:ipsextractor@database:5432/ipsextractor
+      EXTRACTOR_PROVIDER: ${EXTRACTOR_PROVIDER:-}
+      EXTRACTOR_BASE_URL: ${EXTRACTOR_BASE_URL:-}
+      EXTRACTOR_API_KEY: ${EXTRACTOR_API_KEY:-}
+      EXTRACTOR_MODEL: ${EXTRACTOR_MODEL:-}
+      NORMALIZER_PROVIDER: ${NORMALIZER_PROVIDER:-}
+      NORMALIZER_BASE_URL: ${NORMALIZER_BASE_URL:-}
+      NORMALIZER_API_KEY: ${NORMALIZER_API_KEY:-}
+      NORMALIZER_MODEL: ${NORMALIZER_MODEL:-}
+      NORMALIZER_TIMEOUT_S: ${NORMALIZER_TIMEOUT_S:-0}
+      EXTRACTOR_TIMEOUT_S: ${EXTRACTOR_TIMEOUT_S:-120}
+      PDF_DPI: ${PDF_DPI:-600}
+      PDF_THREAD_COUNT: ${PDF_THREAD_COUNT:-10}
+      POPPLER_PATH: ${POPPLER_PATH:-}
+      CONCURRENCY_LIMIT: ${CONCURRENCY_LIMIT:-4}
+      ALLOWED_DOCUMENT_TYPES: ${ALLOWED_DOCUMENT_TYPES:-CC,TI,CE,RC,PA,PPT,PEP,EPT,NUIP}
+    ports:
+      - "8001:8001"
+    volumes:
+      - ${IPS_DOCUMENT_ROOT:-./data}:/data
+
+  cleaning:
+    build:
+      context: ./services/cleaning
+      dockerfile: Dockerfile
+    container_name: ipsextractor-cleaning
+    depends_on:
+      database:
+        condition: service_healthy
+    environment:
+      DATABASE_URL: postgresql+psycopg://ipsextractor:ipsextractor@database:5432/ipsextractor
+    ports:
+      - "8002:8002"
+
+  api_gateway:
+    build:
+      context: ./services/api_gateway
+      dockerfile: Dockerfile
+    container_name: ipsextractor-api
+    depends_on:
+      - extraction
+      - cleaning
+    environment:
+      EXTRACTION_URL: http://extraction:8001
+      CLEANING_URL: http://cleaning:8002
+    ports:
+      - "8000:8000"
+
+volumes:
+  ipsextractor_db: