aury-boot 0.0.2__py3-none-any.whl → 0.0.3__py3-none-any.whl

aury/boot/commands/generate.py

@@ -0,0 +1,1277 @@
+"""代码生成器命令。
+
+生成符合 Aury 规范的代码文件：
+- model: SQLAlchemy 模型
+- repo: Repository 数据访问层
+- service: Service 业务逻辑层
+- api: FastAPI 路由
+- schema: Pydantic 模型
+- crud: 一键生成以上所有
+
+支持两种字段定义模式：
+1. 命令行参数（AI 友好）：
+   aury generate model user email:str:unique age:int? status:str=active
+
+2. 交互式（人类友好）：
+   aury generate model user -i
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from pathlib import Path
+import re
+from typing import Annotated
+
+from rich.console import Console
+from rich.panel import Panel
+from rich.prompt import Confirm, Prompt
+from rich.table import Table
+import typer
+
+from .config import get_project_config
+
+console = Console()
+
+# 模板目录
+GENERATE_TEMPLATES_DIR = Path(__file__).parent / "templates" / "generate"
+
+# 创建代码生成器子应用
+app = typer.Typer(
+    name="generate",
+    help="代码生成器 - 生成 model/repo/service/api/schema",
+    no_args_is_help=True,
+)
+
+
+# ============================================================
+# 字段解析
+# ============================================================
+
+# 类型映射：简写 -> (SQLAlchemy 类型, Pydantic 类型, 需要的导入)
+TYPE_MAPPING: dict[str, tuple[str, str, list[str]]] = {
+    # 字符串
+    "str": ("String(255)", "str", ["String"]),
+    "string": ("String(255)", "str", ["String"]),
+    "text": ("Text", "str", ["Text"]),
+    # 数字
+    "int": ("Integer", "int", ["Integer"]),
+    "integer": ("Integer", "int", ["Integer"]),
+    "bigint": ("BigInteger", "int", ["BigInteger"]),
+    "float": ("Float", "float", ["Float"]),
+    "decimal": ("Numeric(10, 2)", "Decimal", ["Numeric"]),
+    # 布尔
+    "bool": ("Boolean", "bool", ["Boolean"]),
+    "boolean": ("Boolean", "bool", ["Boolean"]),
+    # 日期时间
+    "datetime": ("DateTime", "datetime", ["DateTime"]),
+    "date": ("Date", "date", ["Date"]),
+    "time": ("Time", "time", ["Time"]),
+    # JSON
+    "json": ("JSON", "dict", ["JSON"]),
+    "dict": ("JSON", "dict", ["JSON"]),
+    # UUID
+    "uuid": ("GUID", "str", []),  # 使用框架内置 GUID
+}
+
+
+@dataclass
+class FieldDefinition:
+    """字段定义。"""
+
+    name: str
+    type_name: str = "str"
+    nullable: bool = False
+    unique: bool = False
+    index: bool = False
+    default: str | None = None
+    max_length: int | None = None  # 用于 str 类型
+    comment: str | None = None
+
+    @classmethod
+    def parse(cls, spec: str) -> "FieldDefinition":
+        """解析字段定义字符串。
+
+        格式: name:type:modifiers
+        修饰符:
+        - ? 或 nullable: 可空
+        - unique: 唯一
+        - index: 索引
+        - =value: 默认值
+        - (length): 长度限制
+
+        示例:
+        - email:str:unique
+        - age:int?
+        - status:str=active
+        - name:str(100)
+        - bio:text?
+        - price:decimal:index
+        """
+        parts = spec.split(":")
+        name = parts[0]
+
+        # 默认类型为 str
+        type_name = "str"
+        nullable = False
+        unique = False
+        index = False
+        default = None
+        max_length = None
+
+        for part in parts[1:]:
+            # 检查是否是类型定义
+            type_match = re.match(r"^([a-z]+)(\((\d+)\))?$", part.lower())
+            if type_match and type_match.group(1) in TYPE_MAPPING:
+                type_name = type_match.group(1)
+                if type_match.group(3):
+                    max_length = int(type_match.group(3))
+                continue
+
+            # 检查修饰符
+            part_lower = part.lower()
+            if part_lower in ("?", "nullable"):
+                nullable = True
+            elif part_lower == "unique":
+                unique = True
+            elif part_lower == "index":
+                index = True
+            elif part.startswith("="):
+                default = part[1:]
+            elif part.endswith("?"):
+                # 处理 age:int? 这种格式
+                type_check = part[:-1].lower()
+                if type_check in TYPE_MAPPING:
+                    type_name = type_check
+                    nullable = True
+
+        # 处理名字后面直接跟 ? 的情况，如 "age?"
+        if name.endswith("?"):
+            name = name[:-1]
+            nullable = True
+
+        return cls(
+            name=name,
+            type_name=type_name,
+            nullable=nullable,
+            unique=unique,
+            index=index,
+            default=default,
+            max_length=max_length,
+        )
+
+
+# 可用的模型基类
+# id_type: "int" | "uuid" 用于生成正确的 Schema/Service/API 类型
+MODEL_BASE_CLASSES = {
+    "IDOnlyModel": {
+        "desc": "纯 int 主键（无时间戳，适合关系表）",
+        "features": ["id: int"],
+        "id_type": "int",
+        "has_timestamps": False,
+    },
+    "UUIDOnlyModel": {
+        "desc": "纯 UUID 主键（无时间戳，适合关系表）",
+        "features": ["id: UUID"],
+        "id_type": "uuid",
+        "has_timestamps": False,
+    },
+    "Model": {
+        "desc": "标准模型（int主键 + 时间戳）",
+        "features": ["id: int", "created_at", "updated_at"],
+        "id_type": "int",
+        "has_timestamps": True,
+    },
+    "AuditableStateModel": {
+        "desc": "标准模型 + 软删除",
+        "features": ["id: int", "created_at", "updated_at", "deleted_at"],
+        "id_type": "int",
+        "has_timestamps": True,
+    },
+    "UUIDModel": {
+        "desc": "UUID 主键模型",
+        "features": ["id: UUID", "created_at", "updated_at"],
+        "id_type": "uuid",
+        "has_timestamps": True,
+    },
+    "UUIDAuditableStateModel": {
+        "desc": "UUID 主键 + 软删除（推荐）",
+        "features": ["id: UUID", "created_at", "updated_at", "deleted_at"],
+        "id_type": "uuid",
+        "has_timestamps": True,
+    },
+    "VersionedModel": {
+        "desc": "乐观锁模型（int主键 + version）",
+        "features": ["id: int", "version"],
+        "id_type": "int",
+        "has_timestamps": False,
+    },
+    "VersionedTimestampedModel": {
+        "desc": "乐观锁 + 时间戳",
+        "features": ["id: int", "created_at", "updated_at", "version"],
+        "id_type": "int",
+        "has_timestamps": True,
+    },
+    "VersionedUUIDModel": {
+        "desc": "UUID + 乐观锁 + 时间戳",
+        "features": ["id: UUID", "created_at", "updated_at", "version"],
+        "id_type": "uuid",
+        "has_timestamps": True,
+    },
+    "FullFeaturedModel": {
+        "desc": "完整功能 int版",
+        "features": ["id: int", "created_at", "updated_at", "deleted_at", "version"],
+        "id_type": "int",
+        "has_timestamps": True,
+    },
+    "FullFeaturedUUIDModel": {
+        "desc": "完整功能 UUID版",
+        "features": ["id: UUID", "created_at", "updated_at", "deleted_at", "version"],
+        "id_type": "uuid",
+        "has_timestamps": True,
+    },
+}
+
+# UUID 类型的基类名称
+UUID_BASE_CLASSES = {k for k, v in MODEL_BASE_CLASSES.items() if v.get("id_type") == "uuid"}
+
+
+@dataclass
+class ModelDefinition:
+    """模型定义。"""
+
+    name: str
+    fields: list[FieldDefinition] = field(default_factory=list)
+    soft_delete: bool = True
+    timestamps: bool = True
+    base_class: str | None = None  # 用户指定的基类
+
+    @property
+    def id_type(self) -> str:
+        """获取 id 类型：'int' 或 'uuid'。"""
+        if self.base_class:
+            return MODEL_BASE_CLASSES.get(self.base_class, {}).get("id_type", "uuid")
+        # 默认通过 soft_delete/timestamps 推断时使用 UUID
+        return "uuid"
+
+    @property
+    def id_py_type(self) -> str:
+        """获取 Python/Pydantic 的 id 类型。"""
+        return "int" if self.id_type == "int" else "UUID"
+
+    @property
+    def has_timestamps(self) -> bool:
+        """是否有时间戳字段。"""
+        if self.base_class:
+            return MODEL_BASE_CLASSES.get(self.base_class, {}).get("has_timestamps", True)
+        return self.timestamps
+
+    @property
+    def class_name(self) -> str:
+        """PascalCase 类名。"""
+        return _to_pascal_case(self.name)
+
+    @property
+    def file_name(self) -> str:
+        """snake_case 文件名。"""
+        return _to_snake_case(self.name)
+
+    @property
+    def table_name(self) -> str:
+        """snake_case 复数表名。"""
+        return _to_plural(self.file_name)
+
+    @property
+    def var_name(self) -> str:
+        """变量名。"""
+        return self.file_name
+
+    @property
+    def var_name_plural(self) -> str:
+        """复数变量名。"""
+        return _to_plural(self.file_name)
+
+
+# ============================================================
+# 工具函数
+# ============================================================
+
+
+def _to_snake_case(name: str) -> str:
+    """转换为 snake_case。"""
+    s1 = re.sub(r"(.)([A-Z][a-z]+)", r"\1_\2", name)
+    return re.sub(r"([a-z0-9])([A-Z])", r"\1_\2", s1).lower()
+
+
+def _to_pascal_case(name: str) -> str:
+    """转换为 PascalCase。"""
+    snake = _to_snake_case(name)
+    return "".join(word.capitalize() for word in snake.split("_"))
+
+
+def _to_plural(name: str) -> str:
+    """简单的复数转换。"""
+    if name.endswith("y"):
+        return name[:-1] + "ies"
+    if name.endswith(("s", "x", "ch", "sh")):
+        return name + "es"
+    return name + "s"
+
+
+def _create_file(path: Path, content: str, force: bool = False) -> bool:
+    """创建文件。"""
+    if path.exists() and not force:
+        return False
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(content, encoding="utf-8")
+    return True
+
+
+def _update_init_file(init_path: Path, import_line: str, export_name: str) -> None:
+    """更新 __init__.py 文件。"""
+    if not init_path.exists():
+        init_path.write_text(
+            f'{import_line}\n\n__all__ = ["{export_name}"]\n', encoding="utf-8"
+        )
+        return
+
+    content = init_path.read_text(encoding="utf-8")
+
+    # 检查是否已导入
+    if import_line in content:
+        return
+
+    # 添加导入
+    if "__all__" in content:
+        content = content.replace("__all__", f"{import_line}\n\n__all__")
+        if f'"{export_name}"' not in content:
+            content = content.replace("__all__ = [", f'__all__ = [\n    "{export_name}",')
+    else:
+        content += f'\n{import_line}\n\n__all__ = ["{export_name}"]\n'
+
+    init_path.write_text(content, encoding="utf-8")
+
+
+# ============================================================
+# 交互式字段收集
+# ============================================================
+
+
+def _collect_base_class_interactive() -> str:
+    """交互式选择模型基类。"""
+    console.print("\n[bold cyan]📚 选择模型基类[/bold cyan]\n")
+    
+    # 显示可用基类
+    table = Table(title="可用模型基类", show_header=True, header_style="bold magenta")
+    table.add_column("序号", style="dim", width=4)
+    table.add_column("基类名", style="cyan")
+    table.add_column("说明")
+    table.add_column("自动继承的字段")
+    
+    base_names = list(MODEL_BASE_CLASSES.keys())
+    for i, name in enumerate(base_names, 1):
+        info = MODEL_BASE_CLASSES[name]
+        # 推荐的加标记
+        desc = info["desc"]
+        if name == "UUIDAuditableStateModel":
+            desc = f"[bold green]★ {desc}[/bold green]"
+        table.add_row(str(i), name, desc, ", ".join(info["features"]))
+    
+    console.print(table)
+    console.print()
+    
+    # 默认选择 UUIDAuditableStateModel（第 4 个）
+    choice = Prompt.ask(
+        "请选择基类序号",
+        default="4",
+        choices=[str(i) for i in range(1, len(base_names) + 1)],
+    )
+    
+    selected = base_names[int(choice) - 1]
+    console.print(f"  [green]✓ 已选择: {selected}[/green]\n")
+    return selected
+
+
+def _collect_fields_interactive() -> list[FieldDefinition]:
+    """交互式收集字段定义。"""
+    fields: list[FieldDefinition] = []
+
+    console.print("\n[bold cyan]📝 添加字段[/bold cyan] (输入空名称结束)\n")
+
+    # 显示类型帮助
+    table = Table(title="支持的类型", show_header=True, header_style="bold magenta")
+    table.add_column("类型", style="cyan")
+    table.add_column("说明")
+    table.add_row("str, string", "字符串 (默认)")
+    table.add_row("text", "长文本")
+    table.add_row("int, integer", "整数")
+    table.add_row("bigint", "大整数")
+    table.add_row("float", "浮点数")
+    table.add_row("decimal", "精确小数")
+    table.add_row("bool, boolean", "布尔值")
+    table.add_row("datetime", "日期时间")
+    table.add_row("date", "日期")
+    table.add_row("json, dict", "JSON 对象")
+    console.print(table)
+    console.print()
+
+    while True:
+        name = Prompt.ask("[bold]字段名[/bold]", default="")
+        if not name:
+            break
+
+        type_name = Prompt.ask(
+            "  类型",
+            default="str",
+            choices=list(TYPE_MAPPING.keys()),
+        )
+
+        nullable = Confirm.ask("  可空?", default=False)
+        unique = Confirm.ask("  唯一?", default=False)
+        index = Confirm.ask("  索引?", default=False)
+        default = Prompt.ask("  默认值 (留空无默认)", default="")
+
+        max_length = None
+        if type_name in ("str", "string"):
+            length_str = Prompt.ask("  最大长度", default="255")
+            max_length = int(length_str) if length_str.isdigit() else 255
+
+        fields.append(
+            FieldDefinition(
+                name=name,
+                type_name=type_name,
+                nullable=nullable,
+                unique=unique,
+                index=index,
+                default=default if default else None,
+                max_length=max_length,
+            )
+        )
+
+        console.print(f"  [green]✓ 已添加: {name}:{type_name}[/green]\n")
+
+    return fields
+
+
+# ============================================================
+# 模板读取
+# ============================================================
+
+
+def _read_generate_template(name: str) -> str:
+    """读取代码生成模板文件。"""
+    template_path = GENERATE_TEMPLATES_DIR / name
+    if not template_path.exists():
+        raise FileNotFoundError(f"模板文件不存在: {name} (查找路径: {GENERATE_TEMPLATES_DIR})")
+    return template_path.read_text(encoding="utf-8")
+
+
+def _get_base_class_from_model_file(code_root: Path, model_name: str) -> str | None:
+    """从已生成的模型文件中读取基类名称。
+    
+    Args:
+        code_root: 代码根目录
+        model_name: 模型名称（snake_case）
+    
+    Returns:
+        基类名称，如果无法读取则返回 None
+    """
+    model_file = code_root / "models" / f"{model_name}.py"
+    if not model_file.exists():
+        return None
+    
+    try:
+        content = model_file.read_text(encoding="utf-8")
+        # 查找继承的基类，例如: class User(UUIDAuditableStateModel):
+        pattern = r"class\s+\w+\s*\((\w+)\)"
+        match = re.search(pattern, content)
+        if match:
+            base_class = match.group(1)
+            # 验证是否是有效的基类
+            if base_class in MODEL_BASE_CLASSES:
+                return base_class
+    except Exception:
+        pass
+    
+    return None
+
+
+# ============================================================
+# 模板生成
+# ============================================================
+
+
+def _generate_model_content(model: ModelDefinition) -> str:
+    """生成 Model 内容。"""
+    # 收集需要的导入
+    imports: set[str] = {"String"}  # 默认总是有 String
+    for f in model.fields:
+        type_info = TYPE_MAPPING.get(f.type_name, ("String(255)", "str", ["String"]))
+        imports.update(type_info[2])
+
+    imports_str = ", ".join(sorted(imports))
+
+    # 生成字段定义
+    field_lines = []
+    for f in model.fields:
+        type_info = TYPE_MAPPING.get(f.type_name, ("String(255)", "str", ["String"]))
+        sa_type = type_info[0]
+        py_type = type_info[1]
+
+        # 处理字符串长度
+        if f.type_name in ("str", "string") and f.max_length:
+            sa_type = f"String({f.max_length})"
+
+        # 构建 Mapped 类型
+        if f.nullable:
+            mapped_type = f"Mapped[{py_type} | None]"
+        else:
+            mapped_type = f"Mapped[{py_type}]"
+
+        # 构建 mapped_column 参数
+        col_args = [sa_type]
+        if f.unique:
+            col_args.append("unique=True")
+        if f.index:
+            col_args.append("index=True")
+        if f.nullable:
+            col_args.append("nullable=True")
+        if f.default is not None:
+            if f.type_name in ("str", "string", "text"):
+                col_args.append(f'default="{f.default}"')
+            elif f.type_name in ("bool", "boolean"):
+                col_args.append(f"default={f.default.capitalize()}")
+            else:
+                col_args.append(f"default={f.default}")
+
+        col_args_str = ", ".join(col_args)
+        field_lines.append(f"    {f.name}: {mapped_type} = mapped_column({col_args_str})")
+
+    fields_str = "\n".join(field_lines) if field_lines else "    # 添加字段"
+
+    # 选择基类：优先使用用户指定的，否则根据选项推断
+    if model.base_class:
+        base_class = model.base_class
+        base_info = MODEL_BASE_CLASSES.get(base_class, {})
+        features = base_info.get("features", [])
+        base_doc = f"继承 {base_class} 自动获得：\n    - " + "\n    - ".join(features) if features else f"继承 {base_class} 基类。"
+    elif model.soft_delete and model.timestamps:
+        base_class = "UUIDAuditableStateModel"
+        base_doc = """继承 UUIDAuditableStateModel 自动获得：
+    - id: UUID 主键
+    - created_at: 创建时间
+    - updated_at: 更新时间
+    - deleted_at: 软删除时间戳"""
+    elif model.timestamps:
+        base_class = "UUIDModel"
+        base_doc = """继承 UUIDModel 自动获得：
+    - id: UUID 主键
+    - created_at: 创建时间
+    - updated_at: 更新时间"""
+    else:
+        base_class = "Model"
+        base_doc = "继承 Model 基类。"
+
+    template = _read_generate_template("model.py.tpl")
+    return template.format(
+        class_name=model.class_name,
+        imports_str=imports_str,
+        base_class=base_class,
+        base_doc=base_doc,
+        table_name=model.table_name,
+        fields_str=fields_str,
+    )
+
+
+def _generate_schema_content(model: ModelDefinition) -> str:
+    """生成 Schema 内容。"""
+    # 基础字段
+    base_fields = []
+    update_fields = []
+
+    for f in model.fields:
+        type_info = TYPE_MAPPING.get(f.type_name, ("String(255)", "str", ["String"]))
+        py_type = type_info[1]
+
+        # Base 字段（Create 继承）
+        if f.nullable:
+            field_type = f"{py_type} | None"
+            default = "None"
+        elif f.default is not None:
+            field_type = py_type
+            if f.type_name in ("str", "string", "text"):
+                default = f'"{f.default}"'
+            elif f.type_name in ("bool", "boolean"):
+                default = f.default.capitalize()
+            else:
+                default = f.default
+        else:
+            field_type = py_type
+            default = "..."
+
+        # 构建 Field 参数
+        field_args = [default]
+        if f.type_name in ("str", "string") and f.max_length:
+            field_args.append(f"max_length={f.max_length}")
+        field_args.append(f'description="{f.name}"')
+
+        field_args_str = ", ".join(field_args)
+        base_fields.append(f"    {f.name}: {field_type} = Field({field_args_str})")
+
+        # Update 字段（全部可选）
+        update_field_args = ["None"]
+        if f.type_name in ("str", "string") and f.max_length:
+            update_field_args.append(f"max_length={f.max_length}")
+        update_field_args.append(f'description="{f.name}"')
+        update_field_args_str = ", ".join(update_field_args)
+        update_fields.append(
+            f"    {f.name}: {py_type} | None = Field({update_field_args_str})"
+        )
+
+    base_fields_str = "\n".join(base_fields) if base_fields else "    pass"
+    update_fields_str = "\n".join(update_fields) if update_fields else "    pass"
+
+    # Response 字段（继承 Base，添加 id 和时间戳）
+    id_type = model.id_py_type  # "int" 或 "UUID"
+    
+    # 根据是否有时间戳生成不同的 response 字段
+    if model.has_timestamps:
+        response_extra = f'''    id: {id_type} = Field(..., description="ID")
+    created_at: datetime = Field(..., description="创建时间")
+    updated_at: datetime = Field(..., description="更新时间")'''
+    else:
+        response_extra = f'    id: {id_type} = Field(..., description="ID")'
+
+    # 导入语句
+    imports = ["from datetime import datetime"] if model.has_timestamps else []
+    if id_type == "UUID":
+        imports.append("from uuid import UUID")
+    imports_str = "\n".join(imports)
+    if imports_str:
+        imports_str += "\n"
+
+    template = _read_generate_template("schema.py.tpl")
+    return template.format(
+        class_name=model.class_name,
+        imports_str=imports_str,
+        base_fields_str=base_fields_str,
+        update_fields_str=update_fields_str,
+        response_extra=response_extra,
+    )
+
+
+def _generate_repository_content(model: ModelDefinition, import_prefix: str = "") -> str:
+    """生成 Repository 内容。
+
+    Args:
+        model: 模型定义
+        import_prefix: import 前缀，如 "mypackage." 或 ""
+    """
+    # 查找唯一字段生成 get_by_xxx 方法
+    get_by_methods = []
+    for f in model.fields:
+        if f.unique:
+            type_info = TYPE_MAPPING.get(f.type_name, ("String(255)", "str", ["String"]))
+            py_type = type_info[1]
+            get_by_methods.append(f'''
+    async def get_by_{f.name}(self, {f.name}: {py_type}) -> {model.class_name} | None:
+        """按 {f.name} 获取。"""
+        return await self.get_by({f.name}={f.name})''')
+
+    methods_str = "\n".join(get_by_methods) if get_by_methods else ""
+
+    template = _read_generate_template("repository.py.tpl")
+    return template.format(
+        class_name=model.class_name,
+        import_prefix=import_prefix,
+        file_name=model.file_name,
+        methods_str=methods_str,
+    )
+
+
+def _generate_service_content(model: ModelDefinition, import_prefix: str = "") -> str:
+    """生成 Service 内容。
+
+    Args:
+        model: 模型定义
+        import_prefix: import 前缀，如 "mypackage." 或 ""
+    """
+    # 检查唯一字段，生成重复检测
+    unique_checks = []
+    for f in model.fields:
+        if f.unique:
+            unique_checks.append(
+                f'''        # 检查 {f.name} 是否已存在
+        existing = await self.repo.get_by_{f.name}(data.{f.name})
+        if existing:
+            raise AlreadyExistsError(f"{model.class_name} 已存在: {{data.{f.name}}}")
+'''
+            )
+
+    unique_check_str = "\n".join(unique_checks) if unique_checks else ""
+    
+    # UUID 类型需要导入
+    id_type = model.id_py_type
+    uuid_import = "from uuid import UUID\n\n" if id_type == "UUID" else ""
+
+    template = _read_generate_template("service.py.tpl")
+    return template.format(
+        class_name=model.class_name,
+        uuid_import=uuid_import,
+        import_prefix=import_prefix,
+        file_name=model.file_name,
+        id_py_type=model.id_py_type,
+        unique_check_str=unique_check_str,
+    )
+
+
+def _generate_api_content(model: ModelDefinition, import_prefix: str = "") -> str:
+    """生成 API 内容。
+
+    Args:
+        model: 模型定义
+        import_prefix: import 前缀，如 "mypackage." 或 ""
+    """
+    id_type = model.id_py_type  # "int" 或 "UUID"
+    
+    # 导入语句
+    imports = []
+    if id_type == "UUID":
+        imports.append("from uuid import UUID")
+    imports_str = "\n".join(imports)
+    if imports_str:
+        imports_str += "\n"
+    
+    template = _read_generate_template("api.py.tpl")
+    return template.format(
+        class_name=model.class_name,
+        uuid_import=imports_str,  # 模板中使用 uuid_import 占位符，但实际传入的是 imports_str（可能为空）
+        import_prefix=import_prefix,
+        file_name=model.file_name,
+        var_name_plural=model.var_name_plural,
+        var_name=model.var_name,
+        id_type=id_type,
+    )
+
+
+# ============================================================
+# 命令
+# ============================================================
+
+
+@app.command(name="model")
+def generate_model(
+    name: str = typer.Argument(..., help="模型名称（如 user, UserProfile）"),
+    fields: Annotated[
+        list[str] | None,
+        typer.Argument(
+            help="字段定义，格式: name:type:modifiers（如 email:str:unique age:int?）"
+        ),
+    ] = None,
+    interactive: bool = typer.Option(
+        False, "--interactive", "-i", help="交互式添加字段"
+    ),
+    base: str | None = typer.Option(
+        None, "--base", "-b",
+        help="模型基类（Model/UUIDModel/UUIDAuditableStateModel/VersionedModel 等）"
+    ),
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+    no_soft_delete: bool = typer.Option(False, "--no-soft-delete", help="禁用软删除"),
+    no_timestamps: bool = typer.Option(False, "--no-timestamps", help="禁用时间戳"),
+) -> None:
+    """生成 SQLAlchemy 模型。
+
+    支持两种模式：
+
+    1. 命令行参数（AI 友好）:
+        aury generate model user email:str:unique age:int? status:str=active
+
+    2. 交互式（人类友好）:
+        aury generate model user -i
+
+    字段语法：
+        name:type:modifiers
+
+    支持的类型：
+        str, text, int, bigint, float, decimal, bool, datetime, date, json
+
+    修饰符：
+        ? 或 nullable  - 可空
+        unique         - 唯一约束
+        index          - 索引
+        =value         - 默认值
+        (length)       - 字符串长度，如 str(100)
+
+    可用基类：
+        Model, AuditableStateModel, UUIDModel, UUIDAuditableStateModel,
+        VersionedModel, VersionedTimestampedModel, VersionedUUIDModel,
+        FullFeaturedModel, FullFeaturedUUIDModel
+
+    示例：
+        aury generate model user
+        aury generate model user -b VersionedUUIDModel
+        aury generate model user email:str:unique age:int?
+        aury generate model article title:str(200) content:text status:str=draft
+    """
+    base_path = Path.cwd()
+
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+
+    # 解析字段
+    field_defs: list[FieldDefinition] = []
+    selected_base_class: str | None = base
+
+    if interactive:
+        # 交互式模式：先选择基类，再添加字段
+        selected_base_class = _collect_base_class_interactive()
+        field_defs = _collect_fields_interactive()
+    elif fields:
+        for spec in fields:
+            try:
+                field_defs.append(FieldDefinition.parse(spec))
+            except Exception as e:
+                console.print(f"[red]❌ 解析字段失败: {spec} - {e}[/red]")
+                raise typer.Exit(1) from e
+
+    # 验证基类名称
+    if selected_base_class and selected_base_class not in MODEL_BASE_CLASSES:
+        console.print(f"[red]❌ 无效的基类: {selected_base_class}[/red]")
+        console.print(f"[dim]可用基类: {', '.join(MODEL_BASE_CLASSES.keys())}[/dim]")
+        raise typer.Exit(1)
+
+    model = ModelDefinition(
+        name=name,
+        fields=field_defs,
+        soft_delete=not no_soft_delete,
+        timestamps=not no_timestamps,
+        base_class=selected_base_class,
+    )
+
+    content = _generate_model_content(model)
+    file_path = code_root / "models" / f"{model.file_name}.py"
+    rel_path = file_path.relative_to(base_path)
+
+    if _create_file(file_path, content, force):
+        console.print(f"[green]✅ 创建模型: {rel_path}[/green]")
+        _update_init_file(
+            code_root / "models" / "__init__.py",
+            f"from .{model.file_name} import {model.class_name}",
+            model.class_name,
+        )
+    else:
+        console.print(
+            f"[yellow]⚠️  文件已存在: {rel_path}（使用 --force 覆盖）[/yellow]"
+        )
+
+
+@app.command(name="repo")
+def generate_repo(
+    name: str = typer.Argument(..., help="模型名称"),
+    fields: Annotated[
+        list[str] | None,
+        typer.Argument(help="字段定义（用于生成 get_by_xxx 方法）"),
+    ] = None,
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+) -> None:
+    """生成 Repository 数据访问层。
+
+    示例：
+        aury generate repo user
+        aury generate repo user email:str:unique  # 生成 get_by_email 方法
+    """
+    base_path = Path.cwd()
+
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+    import_prefix = config.get_import_prefix()
+
+    field_defs = [FieldDefinition.parse(spec) for spec in (fields or [])]
+    # 尝试从已生成的模型文件中读取基类信息
+    base_class = _get_base_class_from_model_file(code_root, _to_snake_case(name))
+    model = ModelDefinition(name=name, fields=field_defs, base_class=base_class)
+
+    content = _generate_repository_content(model, import_prefix)
+    file_path = code_root / "repositories" / f"{model.file_name}_repository.py"
+    rel_path = file_path.relative_to(base_path)
+
+    if _create_file(file_path, content, force):
+        console.print(f"[green]✅ 创建仓储: {rel_path}[/green]")
+        _update_init_file(
+            code_root / "repositories" / "__init__.py",
+            f"from .{model.file_name}_repository import {model.class_name}Repository",
+            f"{model.class_name}Repository",
+        )
+    else:
+        console.print("[yellow]⚠️  文件已存在（使用 --force 覆盖）[/yellow]")
+
+
+@app.command(name="service")
+def generate_service(
+    name: str = typer.Argument(..., help="模型名称"),
+    fields: Annotated[
+        list[str] | None,
+        typer.Argument(help="字段定义（用于生成重复检测）"),
+    ] = None,
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+) -> None:
+    """生成 Service 业务逻辑层。
+
+    示例：
+        aury generate service user
+        aury generate service user email:str:unique  # 创建时检查 email 重复
+    """
+    base_path = Path.cwd()
+
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+    import_prefix = config.get_import_prefix()
+
+    field_defs = [FieldDefinition.parse(spec) for spec in (fields or [])]
+    # 尝试从已生成的模型文件中读取基类信息
+    base_class = _get_base_class_from_model_file(code_root, _to_snake_case(name))
+    model = ModelDefinition(name=name, fields=field_defs, base_class=base_class)
+
+    content = _generate_service_content(model, import_prefix)
+    file_path = code_root / "services" / f"{model.file_name}_service.py"
+    rel_path = file_path.relative_to(base_path)
+
+    if _create_file(file_path, content, force):
+        console.print(f"[green]✅ 创建服务: {rel_path}[/green]")
+        _update_init_file(
+            code_root / "services" / "__init__.py",
+            f"from .{model.file_name}_service import {model.class_name}Service",
+            f"{model.class_name}Service",
+        )
+    else:
+        console.print("[yellow]⚠️  文件已存在（使用 --force 覆盖）[/yellow]")
+
+
+def _register_router_in_api_init(code_root: Path, model: ModelDefinition) -> bool:
+    """自动在 api/__init__.py 中注册路由。
+    
+    Args:
+        code_root: 代码根目录（包含 api/ 的目录）
+        model: 模型定义
+    
+    Returns:
+        是否成功注册
+    """
+    api_init_path = code_root / "api" / "__init__.py"
+    if not api_init_path.exists():
+        return False
+    
+    content = api_init_path.read_text(encoding="utf-8")
+    
+    # 检查是否已经注册
+    import_line = f"from . import {model.file_name}"
+    router_line = f"router.include_router({model.file_name}.router)"
+    
+    if import_line in content or f"{model.file_name}.router" in content:
+        return False  # 已经注册
+    
+    # 查找插入位置：在 "# 注册子路由" 标记之后
+    marker = "# 注册子路由"
+    
+    if marker not in content:
+        return False
+    
+    try:
+        # 在标记之后插入
+        lines = content.split("\n")
+        new_lines = []
+        inserted = False
+        
+        for line in lines:
+            new_lines.append(line)
+            if marker in line and not inserted:
+                new_lines.append(import_line)
+                new_lines.append(router_line)
+                inserted = True
+        
+        if inserted:
+            api_init_path.write_text("\n".join(new_lines), encoding="utf-8")
+            return True
+    except Exception:
+        pass  # 插入失败不影响主流程
+    
+    return False
+
+
+@app.command(name="api")
+def generate_api(
+    name: str = typer.Argument(..., help="模型名称"),
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+    no_register: bool = typer.Option(False, "--no-register", help="不自动注册到 api/__init__.py"),
+) -> None:
+    """生成 FastAPI 路由。
+
+    示例：
+        aury generate api user
+        aury generate api user --no-register  # 不自动注册到 api/__init__.py
+    """
+    base_path = Path.cwd()
+
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+    import_prefix = config.get_import_prefix()
+
+    # 尝试从已生成的模型文件中读取基类信息
+    base_class = _get_base_class_from_model_file(code_root, _to_snake_case(name))
+    model = ModelDefinition(name=name, base_class=base_class)
+
+    content = _generate_api_content(model, import_prefix)
+    file_path = code_root / "api" / f"{model.file_name}.py"
+    rel_path = file_path.relative_to(base_path)
+
+    if _create_file(file_path, content, force):
+        console.print(f"[green]✅ 创建 API: {rel_path}[/green]")
+        
+        # 自动注册到 api/__init__.py
+        if not no_register:
+            if _register_router_in_api_init(code_root, model):
+                console.print("[green]✅ 已自动注册到 api/__init__.py[/green]")
+            else:
+                console.print("[dim]提示: 请在 api/__init__.py 中注册路由:[/dim]")
+                console.print(f"[dim]  from . import {model.file_name}[/dim]")
+                console.print(f"[dim]  router.include_router({model.file_name}.router)[/dim]")
+    else:
+        console.print("[yellow]⚠️  文件已存在（使用 --force 覆盖）[/yellow]")
+
+
+@app.command(name="schema")
+def generate_schema(
+    name: str = typer.Argument(..., help="模型名称"),
+    fields: Annotated[
+        list[str] | None,
+        typer.Argument(help="字段定义"),
+    ] = None,
+    interactive: bool = typer.Option(
+        False, "--interactive", "-i", help="交互式添加字段"
+    ),
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+) -> None:
+    """生成 Pydantic Schema。
+
+    示例：
+        aury generate schema user
+        aury generate schema user email:str:unique age:int?
+    """
+    base_path = Path.cwd()
+
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+
+    field_defs: list[FieldDefinition] = []
+    if interactive:
+        field_defs = _collect_fields_interactive()
+    elif fields:
+        field_defs = [FieldDefinition.parse(spec) for spec in fields]
+
+    # 尝试从已生成的模型文件中读取基类信息
+    base_class = _get_base_class_from_model_file(code_root, _to_snake_case(name))
+    model = ModelDefinition(name=name, fields=field_defs, base_class=base_class)
+
+    content = _generate_schema_content(model)
+    file_path = code_root / "schemas" / f"{model.file_name}.py"
+    rel_path = file_path.relative_to(base_path)
+
+    if _create_file(file_path, content, force):
+        console.print(f"[green]✅ 创建 Schema: {rel_path}[/green]")
+        _update_init_file(
+            code_root / "schemas" / "__init__.py",
+            f"from .{model.file_name} import {model.class_name}Create, {model.class_name}Response, {model.class_name}Update",
+            f"{model.class_name}Create",
+        )
+    else:
+        console.print("[yellow]⚠️  文件已存在（使用 --force 覆盖）[/yellow]")
+
+
+@app.command(name="crud")
+def generate_crud(
+    name: str = typer.Argument(..., help="模型名称"),
+    fields: Annotated[
+        list[str] | None,
+        typer.Argument(help="字段定义"),
+    ] = None,
+    interactive: bool = typer.Option(
+        False, "--interactive", "-i", help="交互式添加字段"
+    ),
+    base: str | None = typer.Option(
+        None, "--base", "-b",
+        help="模型基类（Model/UUIDModel/UUIDAuditableStateModel/VersionedModel 等）"
+    ),
+    force: bool = typer.Option(False, "--force", "-f", help="强制覆盖"),
+    no_soft_delete: bool = typer.Option(False, "--no-soft-delete", help="禁用软删除"),
+    no_timestamps: bool = typer.Option(False, "--no-timestamps", help="禁用时间戳"),
+) -> None:
+    """一键生成完整 CRUD（model + repo + service + api + schema）。
+
+    支持两种模式：
+
+    1. 命令行参数（AI 友好）:
+        aury generate crud user email:str:unique age:int? status:str=active
+
+    2. 交互式（人类友好）:
+        aury generate crud user -i
+
+    示例：
+        aury generate crud user
+        aury generate crud user --base Model  # 使用 int 主键
+        aury generate crud user --base UUIDAuditableStateModel  # 使用 UUID 主键（推荐）
+        aury generate crud user email:str:unique age:int? --force
+        aury generate crud article title:str(200) content:text status:str=draft
+    """
+    base_path = Path.cwd()
+    
+    # 读取项目配置
+    config = get_project_config(base_path)
+    code_root = config.get_package_dir(base_path)
+    
+    # 解析字段和基类
+    field_defs: list[FieldDefinition] = []
+    selected_base_class: str | None = base
+
+    if interactive:
+        # 交互式模式：先选择基类，再添加字段
+        selected_base_class = _collect_base_class_interactive()
+        field_defs = _collect_fields_interactive()
+    elif fields:
+        for spec in fields:
+            try:
+                field_defs.append(FieldDefinition.parse(spec))
+            except Exception as e:
+                console.print(f"[red]❌ 解析字段失败: {spec} - {e}[/red]")
+                raise typer.Exit(1) from e
+
+    # 验证基类名称
+    if selected_base_class and selected_base_class not in MODEL_BASE_CLASSES:
+        console.print(f"[red]❌ 无效的基类: {selected_base_class}[/red]")
+        console.print(f"[dim]可用基类: {', '.join(MODEL_BASE_CLASSES.keys())}[/dim]")
+        raise typer.Exit(1)
+
+    model = ModelDefinition(
+        name=name,
+        fields=field_defs,
+        soft_delete=not no_soft_delete,
+        timestamps=not no_timestamps,
+        base_class=selected_base_class,
+    )
+
+    console.print(
+        Panel.fit(
+            f"[bold cyan]⚡ 生成 CRUD: {model.class_name}[/bold cyan]",
+            border_style="cyan",
+        )
+    )
+
+    # 显示字段信息
+    if model.fields:
+        console.print("\n[bold]字段列表:[/bold]")
+        for f in model.fields:
+            modifiers = []
+            if f.nullable:
+                modifiers.append("nullable")
+            if f.unique:
+                modifiers.append("unique")
+            if f.index:
+                modifiers.append("index")
+            if f.default:
+                modifiers.append(f"default={f.default}")
+            mod_str = f" [{', '.join(modifiers)}]" if modifiers else ""
+            console.print(f"  • {f.name}: {f.type_name}{mod_str}")
+
+    console.print()
+
+    # 获取 import_prefix（项目配置已在前面读取）
+    import_prefix = config.get_import_prefix()
+
+    # 生成所有文件，直接使用已创建的 ModelDefinition 对象，确保基类信息一致
+    # 1. 生成 Model
+    model_content = _generate_model_content(model)
+    model_file_path = code_root / "models" / f"{model.file_name}.py"
+    model_rel_path = model_file_path.relative_to(base_path)
+    if _create_file(model_file_path, model_content, force):
+        console.print(f"[green]✅ 创建模型: {model_rel_path}[/green]")
+        _update_init_file(
+            code_root / "models" / "__init__.py",
+            f"from .{model.file_name} import {model.class_name}",
+            model.class_name,
+        )
+    else:
+        console.print(f"[yellow]⚠️  文件已存在: {model_rel_path}（使用 --force 覆盖）[/yellow]")
+
+    # 2. 生成 Repository
+    repo_content = _generate_repository_content(model, import_prefix)
+    repo_file_path = code_root / "repositories" / f"{model.file_name}_repository.py"
+    repo_rel_path = repo_file_path.relative_to(base_path)
+    if _create_file(repo_file_path, repo_content, force):
+        console.print(f"[green]✅ 创建仓储: {repo_rel_path}[/green]")
+        _update_init_file(
+            code_root / "repositories" / "__init__.py",
+            f"from .{model.file_name}_repository import {model.class_name}Repository",
+            f"{model.class_name}Repository",
+        )
+    else:
+        console.print(f"[yellow]⚠️  文件已存在: {repo_rel_path}（使用 --force 覆盖）[/yellow]")
+
+    # 3. 生成 Service
+    service_content = _generate_service_content(model, import_prefix)
+    service_file_path = code_root / "services" / f"{model.file_name}_service.py"
+    service_rel_path = service_file_path.relative_to(base_path)
+    if _create_file(service_file_path, service_content, force):
+        console.print(f"[green]✅ 创建服务: {service_rel_path}[/green]")
+        _update_init_file(
+            code_root / "services" / "__init__.py",
+            f"from .{model.file_name}_service import {model.class_name}Service",
+            f"{model.class_name}Service",
+        )
+    else:
+        console.print(f"[yellow]⚠️  文件已存在: {service_rel_path}（使用 --force 覆盖）[/yellow]")
+
+    # 4. 生成 Schema
+    schema_content = _generate_schema_content(model)
+    schema_file_path = code_root / "schemas" / f"{model.file_name}.py"
+    schema_rel_path = schema_file_path.relative_to(base_path)
+    if _create_file(schema_file_path, schema_content, force):
+        console.print(f"[green]✅ 创建 Schema: {schema_rel_path}[/green]")
+        _update_init_file(
+            code_root / "schemas" / "__init__.py",
+            f"from .{model.file_name} import {model.class_name}Create, {model.class_name}Response, {model.class_name}Update",
+            f"{model.class_name}Create",
+        )
+    else:
+        console.print(f"[yellow]⚠️  文件已存在: {schema_rel_path}（使用 --force 覆盖）[/yellow]")
+
+    # 5. 生成 API
+    api_content = _generate_api_content(model, import_prefix)
+    api_file_path = code_root / "api" / f"{model.file_name}.py"
+    api_rel_path = api_file_path.relative_to(base_path)
+    if _create_file(api_file_path, api_content, force):
+        console.print(f"[green]✅ 创建 API: {api_rel_path}[/green]")
+        # 自动注册到 api/__init__.py
+        if _register_router_in_api_init(code_root, model):
+            console.print("[green]✅ 已自动注册到 api/__init__.py[/green]")
+        else:
+            console.print("[dim]提示: 请在 api/__init__.py 中注册路由:[/dim]")
+            console.print(f"[dim]  from . import {model.file_name}[/dim]")
+            console.print(f"[dim]  router.include_router({model.file_name}.router)[/dim]")
+    else:
+        console.print(f"[yellow]⚠️  文件已存在: {api_rel_path}（使用 --force 覆盖）[/yellow]")
+
+    console.print()
+    console.print("[bold green]✨ CRUD 生成完成！[/bold green]")
+    console.print()
+    console.print("[bold]下一步：[/bold]")
+    console.print("  1. 生成数据库迁移：")
+    console.print(f'     [cyan]aury migrate make -m "add {model.file_name} table"[/cyan]')
+    console.print("  2. 执行迁移：")
+    console.print("     [cyan]aury migrate up[/cyan]")
+
+
+__all__ = ["app"]