Develop labeling module (#25)

* refactor: remove db table management from LS adapter (mv to scripts later); change adapter to use the same MySQL DB as other modules.

* refactor: Rename LS Adapter module to datamate-python

runtime/datamate-python/app/models/collection/__init__.py

@@ -0,0 +1,13 @@
+# app/models/collection/__init__.py
+
+from .task_execution import TaskExecution
+from .collection_task import CollectionTask
+from .task_log import TaskLog
+from .datax_template import DataxTemplate
+
+__all__ = [
+    "TaskExecution",
+    "CollectionTask",
+    "TaskLog",
+    "DataxTemplate"
+]

runtime/datamate-python/app/models/collection/collection_task.py

@@ -0,0 +1,28 @@
+from sqlalchemy import Column, String, Text, Integer, BigInteger, TIMESTAMP
+from sqlalchemy.sql import func
+from app.db.database import Base
+
+class CollectionTask(Base):
+    """数据归集任务模型"""
+    
+    __tablename__ = "t_dc_collection_tasks"
+    
+    id = Column(String(36), primary_key=True, comment="任务ID（UUID）")
+    name = Column(String(255), nullable=False, comment="任务名称")
+    description = Column(Text, nullable=True, comment="任务描述")
+    sync_mode = Column(String(20), default='ONCE', comment="同步模式：ONCE/SCHEDULED")
+    config = Column(Text, nullable=False, comment="归集配置（DataX配置），包含源端和目标端配置信息")
+    schedule_expression = Column(String(255), nullable=True, comment="Cron调度表达式")
+    status = Column(String(20), default='DRAFT', comment="任务状态：DRAFT/READY/RUNNING/SUCCESS/FAILED/STOPPED")
+    retry_count = Column(Integer, default=3, comment="重试次数")
+    timeout_seconds = Column(Integer, default=3600, comment="超时时间（秒）")
+    max_records = Column(BigInteger, nullable=True, comment="最大处理记录数")
+    sort_field = Column(String(100), nullable=True, comment="增量字段")
+    last_execution_id = Column(String(36), nullable=True, comment="最后执行ID（UUID）")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    updated_at = Column(TIMESTAMP, server_default=func.current_timestamp(), onupdate=func.current_timestamp(), comment="更新时间")
+    created_by = Column(String(255), nullable=True, comment="创建者")
+    updated_by = Column(String(255), nullable=True, comment="更新者")
+    
+    def __repr__(self):
+        return f"<CollectionTask(id={self.id}, name={self.name}, status={self.status})>"

runtime/datamate-python/app/models/collection/datax_template.py

@@ -0,0 +1,23 @@
+from sqlalchemy import Column, String, Text, Boolean, TIMESTAMP
+from sqlalchemy.sql import func
+from app.db.database import Base
+
+class DataxTemplate(Base):
+    """DataX模板配置模型"""
+    
+    __tablename__ = "t_dc_datax_templates"
+    
+    id = Column(String(36), primary_key=True, comment="模板ID（UUID）")
+    name = Column(String(255), nullable=False, unique=True, comment="模板名称")
+    source_type = Column(String(50), nullable=False, comment="源数据源类型")
+    target_type = Column(String(50), nullable=False, comment="目标数据源类型")
+    template_content = Column(Text, nullable=False, comment="模板内容")
+    description = Column(Text, nullable=True, comment="模板描述")
+    version = Column(String(20), default='1.0.0', comment="版本号")
+    is_system = Column(Boolean, default=False, comment="是否系统模板")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    updated_at = Column(TIMESTAMP, server_default=func.current_timestamp(), onupdate=func.current_timestamp(), comment="更新时间")
+    created_by = Column(String(255), nullable=True, comment="创建者")
+    
+    def __repr__(self):
+        return f"<DataxTemplate(id={self.id}, name={self.name}, source={self.source_type}, target={self.target_type})>"

runtime/datamate-python/app/models/collection/task_execution.py

@@ -0,0 +1,34 @@
+from sqlalchemy import Column, String, Text, Integer, BigInteger, DECIMAL, JSON, TIMESTAMP
+from sqlalchemy.sql import func
+from app.db.database import Base
+
+class TaskExecution(Base):
+    """任务执行明细模型"""
+    
+    __tablename__ = "t_dc_task_executions"
+    
+    id = Column(String(36), primary_key=True, comment="执行记录ID（UUID）")
+    task_id = Column(String(36), nullable=False, comment="任务ID")
+    task_name = Column(String(255), nullable=False, comment="任务名称")
+    status = Column(String(20), default='RUNNING', comment="执行状态：RUNNING/SUCCESS/FAILED/STOPPED")
+    progress = Column(DECIMAL(5, 2), default=0.00, comment="进度百分比")
+    records_total = Column(BigInteger, default=0, comment="总记录数")
+    records_processed = Column(BigInteger, default=0, comment="已处理记录数")
+    records_success = Column(BigInteger, default=0, comment="成功记录数")
+    records_failed = Column(BigInteger, default=0, comment="失败记录数")
+    throughput = Column(DECIMAL(10, 2), default=0.00, comment="吞吐量（条/秒）")
+    data_size_bytes = Column(BigInteger, default=0, comment="数据量（字节）")
+    started_at = Column(TIMESTAMP, nullable=True, comment="开始时间")
+    completed_at = Column(TIMESTAMP, nullable=True, comment="完成时间")
+    duration_seconds = Column(Integer, default=0, comment="执行时长（秒）")
+    config = Column(JSON, nullable=True, comment="执行配置")
+    error_message = Column(Text, nullable=True, comment="错误信息")
+    datax_job_id = Column(Text, nullable=True, comment="datax任务ID")
+    result = Column(Text, nullable=True, comment="执行结果")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    updated_at = Column(TIMESTAMP, server_default=func.current_timestamp(), onupdate=func.current_timestamp(), comment="更新时间")
+    created_by = Column(String(255), nullable=True, comment="创建者")
+    updated_by = Column(String(255), nullable=True, comment="更新者")
+    
+    def __repr__(self):
+        return f"<TaskExecution(id={self.id}, task_id={self.task_id}, status={self.status})>"

runtime/datamate-python/app/models/collection/task_log.py

@@ -0,0 +1,26 @@
+from sqlalchemy import Column, String, Text, Integer, BigInteger, TIMESTAMP
+from sqlalchemy.sql import func
+from app.db.database import Base
+
+class TaskLog(Base):
+    """任务执行记录模型"""
+    
+    __tablename__ = "t_dc_task_log"
+    
+    id = Column(String(36), primary_key=True, comment="执行记录ID（UUID）")
+    task_id = Column(String(36), nullable=False, comment="任务ID")
+    task_name = Column(String(255), nullable=False, comment="任务名称")
+    sync_mode = Column(String(20), default='FULL', comment="同步模式：FULL/INCREMENTAL")
+    status = Column(String(20), default='RUNNING', comment="执行状态：RUNNING/SUCCESS/FAILED/STOPPED")
+    start_time = Column(TIMESTAMP, nullable=True, comment="开始时间")
+    end_time = Column(TIMESTAMP, nullable=True, comment="结束时间")
+    duration = Column(BigInteger, nullable=True, comment="执行时长(毫秒)")
+    process_id = Column(String(50), nullable=True, comment="进程ID")
+    log_path = Column(String(500), nullable=True, comment="日志文件路径")
+    error_msg = Column(Text, nullable=True, comment="错误信息")
+    result = Column(Text, nullable=True, comment="执行结果")
+    retry_times = Column(Integer, default=0, comment="重试次数")
+    create_time = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    
+    def __repr__(self):
+        return f"<TaskLog(id={self.id}, task_id={self.task_id}, status={self.status})>"