refactor: Reorganize datamate-python (#34)

refactor: Reorganize datamate-python (previously label-studio-adapter) into a DDD style structure.
2025-10-30 01:32:59 +08:00
parent 0614157c0b
commit 2f7341dc1f
79 changed files with 1077 additions and 1577 deletions
--- a/runtime/datamate-python/app/db/models/init.py
+++ b/runtime/datamate-python/app/db/models/init.py
@@ -0,0 +1,28 @@
+
+from .dataset_management import (
+    Dataset,
+    DatasetTag,
+    DatasetFiles,
+    DatasetStatistics,
+    Tag
+)
+
+from .user_management import (
+    User
+)
+
+from .annotation_management import (
+    AnnotationTemplate,
+    LabelingProject
+)
+
+__all__ = [
+    "Dataset",
+    "DatasetTag",
+    "DatasetFiles",
+    "DatasetStatistics",
+    "Tag",
+    "User",
+    "AnnotationTemplate",
+    "LabelingProject",
+]
--- a/runtime/datamate-python/app/db/models/annotation_management.py
+++ b/runtime/datamate-python/app/db/models/annotation_management.py
@@ -0,0 +1,51 @@
+"""
+Tables of Annotation Management Module
+"""
+
+import uuid
+from sqlalchemy import Column, String, BigInteger, Boolean, TIMESTAMP, Text, Integer, JSON, Date
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+class AnnotationTemplate(Base):
+    """标注模板模型"""
+    
+    __tablename__ = "t_dm_annotation_templates"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID主键ID")
+    name = Column(String(32), nullable=False, comment="模板名称")
+    description = Column(String(255), nullable=True, comment="模板描述")
+    configuration = Column(JSON, nullable=True, comment="配置信息（JSON格式）")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    deleted_at = Column(TIMESTAMP, nullable=True, comment="删除时间（软删除）")
+    
+    def __repr__(self):
+        return f"<AnnotationTemplate(id={self.id}, name={self.name})>"
+    
+    @property
+    def is_deleted(self) -> bool:
+        """检查是否已被软删除"""
+        return self.deleted_at is not None
+    
+class LabelingProject(Base):
+    """标注工程表"""
+    
+    __tablename__ = "t_dm_labeling_projects"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID主键ID")
+    dataset_id = Column(String(36), nullable=False, comment="数据集ID")
+    name = Column(String(32), nullable=False, comment="项目名称")
+    labeling_project_id = Column(String(8), nullable=False, comment="Label Studio项目ID")
+    configuration = Column(JSON, nullable=True, comment="标签配置")
+    progress = Column(JSON, nullable=True, comment="标注进度统计")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    deleted_at = Column(TIMESTAMP, nullable=True, comment="删除时间（软删除）")
+    
+    def __repr__(self):
+        return f"<LabelingProject(id={self.id}, dataset_id={self.dataset_id}, name={self.name})>"
+    
+    @property
+    def is_deleted(self) -> bool:
+        """检查是否已被软删除"""
+        return self.deleted_at is not None
--- a/runtime/datamate-python/app/db/models/dataset_management.py
+++ b/runtime/datamate-python/app/db/models/dataset_management.py
@@ -0,0 +1,113 @@
+"""
+Tables of Dataset Management Module
+"""
+
+import uuid
+from sqlalchemy import Column, String, BigInteger, Boolean, TIMESTAMP, Text, Integer, JSON, Date
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+class Dataset(Base):
+    """数据集模型（支持医学影像、文本、问答等多种类型）"""
+    
+    __tablename__ = "t_dm_datasets"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID")
+    name = Column(String(255), nullable=False, comment="数据集名称")
+    description = Column(Text, nullable=True, comment="数据集描述")
+    dataset_type = Column(String(50), nullable=False, comment="数据集类型：IMAGE/TEXT/QA/MULTIMODAL/OTHER")
+    category = Column(String(100), nullable=True, comment="数据集分类：医学影像/问答/文献等")
+    path = Column(String(500), nullable=True, comment="数据存储路径")
+    format = Column(String(50), nullable=True, comment="数据格式：DCM/JPG/JSON/CSV等")
+    schema_info = Column(JSON, nullable=True, comment="数据结构信息")
+    size_bytes = Column(BigInteger, default=0, comment="数据大小(字节)")
+    file_count = Column(BigInteger, default=0, comment="文件数量")
+    record_count = Column(BigInteger, default=0, comment="记录数量")
+    retention_days = Column(Integer, default=0, comment="数据保留天数（0表示长期保留）")
+    tags = Column(JSON, nullable=True, comment="标签列表")
+    dataset_metadata = Column("metadata", JSON, nullable=True, comment="元数据信息")
+    status = Column(String(50), default='DRAFT', comment="状态：DRAFT/ACTIVE/ARCHIVED")
+    is_public = Column(Boolean, default=False, comment="是否公开")
+    is_featured = Column(Boolean, default=False, comment="是否推荐")
+    version = Column(BigInteger, nullable=False, default=0, comment="版本号")
+    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"<Dataset(id={self.id}, name={self.name}, type={self.dataset_type})>"
+
+class DatasetTag(Base):
+    """数据集标签关联模型"""
+    
+    __tablename__ = "t_dm_dataset_tags"
+    
+    dataset_id = Column(String(36), primary_key=True, comment="数据集ID（UUID）")
+    tag_id = Column(String(36), primary_key=True, comment="标签ID（UUID）")
+    created_at = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="创建时间")
+    
+    def __repr__(self):
+        return f"<DatasetTag(dataset_id={self.dataset_id}, tag_id={self.tag_id})>"
+
+class DatasetFiles(Base):
+    """DM数据集文件模型"""
+    
+    __tablename__ = "t_dm_dataset_files"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID")
+    dataset_id = Column(String(36), nullable=False, comment="所属数据集ID（UUID）")
+    file_name = Column(String(255), nullable=False, comment="文件名")
+    file_path = Column(String(1000), nullable=False, comment="文件路径")
+    file_type = Column(String(50), nullable=True, comment="文件格式：JPG/PNG/DCM/TXT等")
+    file_size = Column(BigInteger, default=0, comment="文件大小(字节)")
+    check_sum = Column(String(64), nullable=True, comment="文件校验和")
+    tags = Column(JSON, nullable=True, comment="文件标签信息")
+    dataset_filemetadata = Column("metadata", JSON, nullable=True, comment="文件元数据")
+    status = Column(String(50), default='ACTIVE', comment="文件状态：ACTIVE/DELETED/PROCESSING")
+    upload_time = Column(TIMESTAMP, server_default=func.current_timestamp(), comment="上传时间")
+    last_access_time = Column(TIMESTAMP, nullable=True, comment="最后访问时间")
+    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")
+    
+    def __repr__(self):
+        return f"<DatasetFiles(id={self.id}, dataset_id={self.dataset_id}, file_name={self.file_name})>"
+    
+class DatasetStatistics(Base):
+    """数据集统计信息模型"""
+    
+    __tablename__ = "t_dm_dataset_statistics"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID")
+    dataset_id = Column(String(36), nullable=False, comment="数据集ID（UUID）")
+    stat_date = Column(Date, nullable=False, comment="统计日期")
+    total_files = Column(BigInteger, default=0, comment="总文件数")
+    total_size = Column(BigInteger, default=0, comment="总大小(字节)")
+    processed_files = Column(BigInteger, default=0, comment="已处理文件数")
+    error_files = Column(BigInteger, default=0, comment="错误文件数")
+    download_count = Column(BigInteger, default=0, comment="下载次数")
+    view_count = Column(BigInteger, default=0, comment="查看次数")
+    quality_metrics = Column(JSON, nullable=True, comment="质量指标")
+    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")
+    
+    def __repr__(self):
+        return f"<DatasetStatistics(id={self.id}, dataset_id={self.dataset_id}, date={self.stat_date})>"
+
+class Tag(Base):
+    """标签集合模型"""
+    
+    __tablename__ = "t_dm_tags"
+    
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID")
+    name = Column(String(100), nullable=False, unique=True, comment="标签名称")
+    description = Column(Text, nullable=True, comment="标签描述")
+    category = Column(String(50), nullable=True, comment="标签分类")
+    color = Column(String(7), nullable=True, comment="标签颜色(十六进制)")
+    usage_count = Column(BigInteger, default=0, comment="使用次数")
+    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")
+    
+    def __repr__(self):
+        return f"<Tag(id={self.id}, name={self.name}, category={self.category})>"
--- a/runtime/datamate-python/app/db/models/user_management.py
+++ b/runtime/datamate-python/app/db/models/user_management.py
@@ -0,0 +1,29 @@
+"""
+Tables of User Management Module
+"""
+
+from sqlalchemy import Column, String, BigInteger, Boolean, TIMESTAMP
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+class User(Base):
+    """用户模型"""
+    
+    __tablename__ = "users"
+    
+    id = Column(BigInteger, primary_key=True, autoincrement=True, comment="用户ID")
+    username = Column(String(255), nullable=False, unique=True, comment="用户名")
+    email = Column(String(255), nullable=False, unique=True, comment="邮箱")
+    password_hash = Column(String(255), nullable=False, comment="密码哈希")
+    full_name = Column(String(255), nullable=True, comment="真实姓名")
+    avatar_url = Column(String(500), nullable=True, comment="头像URL")
+    role = Column(String(50), nullable=False, default='USER', comment="角色：ADMIN/USER")
+    organization = Column(String(255), nullable=True, comment="所属机构")
+    enabled = Column(Boolean, nullable=False, default=True, comment="是否启用")
+    last_login_at = Column(TIMESTAMP, nullable=True, comment="最后登录时间")
+    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")
+    
+    def __repr__(self):
+        return f"<User(id={self.id}, username={self.username}, role={self.role})>"