DataMate/editions/community/config/application.yml

# 数据引擎平台 - 主应用配置
spring:
  application:
    name: data-mate-platform

  # 暂时排除Spring Security自动配置（开发阶段使用）
  autoconfigure:
    exclude:
      - org.springframework.boot.autoconfigure.security.servlet.SecurityAutoConfiguration
      - org.springframework.boot.autoconfigure.security.servlet.UserDetailsServiceAutoConfiguration

  # 数据源配置
  datasource:
    driver-class-name: com.mysql.cj.jdbc.Driver
    url: jdbc:mysql://datamate-database:3306/datamate?useUnicode=true&characterEncoding=utf8&useSSL=false&serverTimezone=Asia/Shanghai&allowPublicKeyRetrieval=true
    username: ${DB_USERNAME:root}
    password: ${DB_PASSWORD:Huawei@123}
    hikari:
      maximum-pool-size: 20
      minimum-idle: 5
      connection-timeout: 30000
      idle-timeout: 600000
      max-lifetime: 1800000

  # Elasticsearch配置
  elasticsearch:
    uris: ${ES_URIS:http://localhost:9200}
    username: ${ES_USERNAME:}
    password: ${ES_PASSWORD:}
    connection-timeout: 10s
    socket-timeout: 30s

  # Jackson配置
  jackson:
    time-zone: Asia/Shanghai
    date-format: yyyy-MM-dd HH:mm:ss
    serialization:
      write-dates-as-timestamps: false
    deserialization:
      fail-on-unknown-properties: false

  # 文件上传配置
  servlet:
    multipart:
      max-file-size: 100MB
      max-request-size: 100MB

  # 任务调度配置
  task:
    execution:
      pool:
        core-size: ${TASK_EXECUTION_CORE_SIZE:10}
        max-size: ${TASK_EXECUTION_MAX_SIZE:20}
        queue-capacity: ${TASK_EXECUTION_QUEUE_CAPACITY:100}
        keep-alive: ${TASK_EXECUTION_KEEP_ALIVE:60s}
    scheduling:
      pool:
        size: ${TASK_SCHEDULING_POOL_SIZE:5}
  config:
    import:
      - classpath:config/application-datacollection.yml
      - classpath:config/application-datamanagement.yml

# MyBatis配置（需在顶层，不在 spring 下）
mybatis-plus:
  configuration:
    map-underscore-to-camel-case: true
    default-fetch-size: 100
    default-statement-timeout: 30
    use-generated-keys: true
    cache-enabled: true
    lazy-loading-enabled: false
    multiple-result-sets-enabled: true
    use-column-label: true
    auto-mapping-behavior: partial
    auto-mapping-unknown-column-behavior: none
    default-executor-type: simple
    call-setters-on-nulls: false
    return-instance-for-empty-row: false
    log-impl: org.apache.ibatis.logging.slf4j.Slf4jImpl
  mapper-locations:
    - classpath*:mappers/**/*.xml
  type-aliases-package:
    - com.datamate.collection.domain.model
    - com.datamate.datamanagement.domain.model.dataset

# 应用配置
server:
  port: ${SERVER_PORT:8080}
  servlet:
    context-path: /api
    encoding:
      charset: UTF-8
      enabled: true
      force: true

# 日志配置
logging:
  config: file:/opt/backend/log4j2.xml

# Actuator配置
management:
  endpoints:
    web:
      exposure:
        include: health,info,metrics,prometheus
  endpoint:
    health:
      show-details: when-authorized
  health:
    elasticsearch:
      enabled: false  # 禁用Elasticsearch健康检查

# 平台配置
datamate:
  # JWT配置
  jwt:
    secret: ${JWT_SECRET:dataMateSecretKey2024ForJWTTokenGeneration}
    expiration: ${JWT_EXPIRATION:86400} # 24小时，单位秒
    header: Authorization
    prefix: "Bearer "

  # 文件存储配置
  storage:
    type: ${STORAGE_TYPE:local} # local, minio, s3
    local:
      base-path: ${STORAGE_LOCAL_PATH:./data/storage}
    minio:
      endpoint: ${MINIO_ENDPOINT:http://localhost:9000}
      access-key: ${MINIO_ACCESS_KEY:minioadmin}
      secret-key: ${MINIO_SECRET_KEY:minioadmin}
      bucket-name: ${MINIO_BUCKET:data-mate}

  # Ray执行器配置
  ray:
    enabled: ${RAY_ENABLED:false}
    address: ${RAY_ADDRESS:ray://localhost:10001}
    runtime-env:
      working-dir: ${RAY_WORKING_DIR:./runtime/python-executor}
      pip-packages:
        - "ray[default]==2.7.0"
        - "pandas"
        - "numpy"
        - "data-juicer"

  # 数据归集服务配置（可由模块导入叠加）
  data-collection: {}

  # 算子市场配置
  operator-market:
    repository-path: ${OPERATOR_REPO_PATH:./runtime/operators}
    registry-url: ${OPERATOR_REGISTRY_URL:}
    max-upload-size: ${OPERATOR_MAX_UPLOAD_SIZE:50MB}

  # 数据处理配置
  data-processing:
    max-file-size: ${MAX_FILE_SIZE:1GB}
    temp-dir: ${TEMP_DIR:./data/temp}
    batch-size: ${BATCH_SIZE:1000}

  # 标注配置
  annotation:
    auto-annotation:
      enabled: ${AUTO_ANNOTATION_ENABLED:true}
      model-endpoint: ${ANNOTATION_MODEL_ENDPOINT:}
    quality-control:
      enabled: ${QC_ENABLED:true}
      threshold: ${QC_THRESHOLD:0.8}

  # RAG配置
  rag:
    embedding:
      model: ${RAG_EMBEDDING_MODEL:text-embedding-ada-002}
      api-key: ${RAG_API_KEY:}
      dimension: ${RAG_DIMENSION:1536}
    chunk:
      size: ${RAG_CHUNK_SIZE:512}
      overlap: ${RAG_CHUNK_OVERLAP:50}
    retrieval:
      top-k: ${RAG_TOP_K:5}
      score-threshold: ${RAG_SCORE_THRESHOLD:0.7}