precis-cli 0.1.3__py3-none-any.whl

app/api/routers/ai/models.py

@@ -0,0 +1,352 @@
+# SPDX-License-Identifier: Apache-2.0
+#
+# Copyright 2026 Precis Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+@fileoverview AI API 请求/响应 Pydantic 模型定义模块
+
+功能概述:
+- 定义 AI 聊天、Provider 管理、配置生成等 API 的输入输出模型
+- 提供硬件诊断相关的数据模型
+- 支持异步任务的创建、状态、结果模型
+
+架构设计:
+- 按功能区域分块：Provider 管理、配置生成、异步任务、硬件诊断
+- 所有模型继承自 Pydantic BaseModel，支持自动校验和序列化
+- 使用 Field 定义字段描述和默认值
+
+输入示例:
+    class ConfigGenerateRequest(BaseModel):
+        file_paths: list[str]
+        project_name: str
+        options: ConfigGenerateOptions = Field(default_factory=ConfigGenerateOptions)
+
+输出示例:
+    class ConfigGenerateResponse(BaseModel):
+        success: bool
+        yaml_preview: str
+        manifest: Optional[dict[str, Any]] = None
+"""
+
+from typing import Any
+
+from pydantic import BaseModel, ConfigDict, Field
+
+# =============================================================================
+# Provider 管理模型
+# =============================================================================
+
+
+class ChatMessageInput(BaseModel):
+    """聊天消息输入"""
+
+    role: str
+    content: str
+
+
+class ChatRequestInput(BaseModel):
+    """聊天请求输入"""
+
+    provider_id: str | None = Field(default=None, description="指定 Provider，不指定则使用默认")
+    messages: list[ChatMessageInput]
+    stream: bool = False
+    model: str | None = Field(default=None, description="覆盖默认模型")
+    temperature: float = 0.7
+
+
+class ChatResponseOutput(BaseModel):
+    """聊天响应输出"""
+
+    content: str
+    model: str
+
+
+class AiChatContextNode(BaseModel):
+    """AI Chat 上下文节点"""
+
+    id: str
+    type: str
+    data: dict[str, Any]
+    label: str | None = None
+
+
+class AiChatContext(BaseModel):
+    """AI Chat 上下文"""
+
+    hasContext: bool
+    selectedNodes: list[AiChatContextNode]
+    # 画布全部业务节点快照（前端 serializeCanvasForAI 裁剪后携带），
+    # 供 read_canvas 工具查询画布真实状态。默认空 list 保证旧请求/CLI 不破坏。
+    canvasNodes: list[AiChatContextNode] = Field(default_factory=list)
+
+
+class AiChatHistoryMessage(BaseModel):
+    """AI Chat 历史消息"""
+
+    role: str
+    content: str
+
+
+class AiChatRequest(BaseModel):
+    """AI Chat 请求输入 (对应前端请求)"""
+
+    message: str
+    context: AiChatContext
+    history: list[AiChatHistoryMessage] | None = None
+    agent_mode: bool = Field(default=True, description="是否启用 Agent 深度模式")
+
+
+class AiChatConfirmRequest(BaseModel):
+    """AI Chat apply_actions 确认/拒绝请求"""
+
+    decision: str = Field(..., description="confirm(确认并落盘)或 reject(拒绝,不落盘)")
+    apply_id: str | None = Field(
+        default=None,
+        description="本次 apply 的 ID（{job_id}#{seq}）。为空时回退到该 job 当前唯一挂起项。",
+    )
+
+
+class AiChatRespondRequest(BaseModel):
+    """AI Chat ask_user 提问的回答请求"""
+
+    ask_id: str = Field(..., description="本次 ask 的 ID（{job_id}#ask#{seq}）")
+    response: dict[str, Any] = Field(
+        ...,
+        description="用户回答。形如 {answer: ...} 或 {skipped: true, reason: ...}",
+    )
+
+
+class AgentMeta(BaseModel):
+    """Agent 模式执行元数据（仅 agent_mode=true 时填充）"""
+
+    iterations: int = 0
+    tool_steps: list[dict[str, Any]] = Field(default_factory=list, description="工具调用轨迹")
+
+
+class AiChatResponse(BaseModel):
+    """AI Chat 响应输出 (对应前端期望)"""
+
+    status: str
+    reply: str
+    actions: list[Any]
+    frontend_instructions: list[Any]
+    agent_meta: AgentMeta | None = None
+    error: str | None = None
+
+
+class ProviderResponse(BaseModel):
+    """Provider 响应"""
+
+    model_config = ConfigDict(populate_by_name=True)
+
+    id: str
+    name: str
+    provider: str = Field(alias="type")
+    deployment: str
+    base_url: str
+    model: str
+    context_window: int | None = Field(default=None, description="用户指定的上下文窗口，None 表示自动探测/回退")
+    health: dict[str, Any]
+    is_configured: bool = False
+
+
+class DiscoverResponse(BaseModel):
+    """发现服务响应"""
+
+    services: list[dict[str, Any]]
+    count: int
+
+
+class TestProviderResponse(BaseModel):
+    """测试 Provider 响应"""
+
+    provider_id: str
+    health: dict[str, Any]
+    available_models: list[str]
+
+
+class CreateProviderRequest(BaseModel):
+    """创建 Provider 请求"""
+
+    name: str = Field(..., description="显示名称")
+    type: str = Field(default="openai", description="Provider 类型（openai 或 ollama）")
+    base_url: str = Field(..., description="API 基础 URL")
+    api_key: str | None = Field(default=None, description="API 密钥，本地服务可留空")
+    model: str = Field(..., description="默认模型名称")
+    context_window: int | None = Field(
+        default=None,
+        ge=1024,
+        description="模型上下文窗口（tokens），留空则自动探测或回退到 200k",
+    )
+
+
+class UpdateProviderRequest(BaseModel):
+    """更新 Provider 请求（所有字段可选，仅传递需要更新的字段）"""
+
+    name: str | None = Field(default=None, description="显示名称")
+    type: str | None = Field(default=None, description="Provider 类型")
+    base_url: str | None = Field(default=None, description="API 基础 URL")
+    api_key: str | None = Field(default=None, description="API 密钥，传空字符串表示清空")
+    model: str | None = Field(default=None, description="默认模型名称")
+    context_window: int | None = Field(
+        default=None,
+        ge=1024,
+        description="模型上下文窗口（tokens），不传则保持原值，传值则更新",
+    )
+
+
+class ProviderPresetResponse(BaseModel):
+    """服务商预设响应"""
+
+    id: str
+    name: str
+    type: str
+    base_url: str
+    default_model: str
+    models: list[str]
+
+
+# =============================================================================
+# 配置生成模型 (新架构)
+# =============================================================================
+
+
+class ConfigGenerateOptions(BaseModel):
+    """配置生成选项"""
+
+    sample_rows: int = Field(default=100, description="每文件采样行数")
+    sample_values_per_column: int = Field(default=100, description="每列采样值数量")
+    max_files: int = Field(default=50, description="最大处理文件数")
+    max_cell_chars: int = Field(default=500, description="单元格最大字符数")
+    generate_schemas: bool = Field(default=True, description="生成表结构")
+    generate_constraints: bool = Field(default=True, description="生成约束规则")
+    generate_regex_nodes: bool = Field(default=False, description="生成正则节点")
+    keep_existing: bool = Field(default=True, description="保留现有配置")
+    target_files: list[str] | None = Field(default=None, description="指定目标文件")
+    agent_mode: bool = Field(default=True, description="启用 Agent 多轮优化模式")
+    max_iterations: int = Field(default=2, ge=1, le=5, description="Agent 最大迭代轮数")
+    validation_sample_size: int = Field(default=1000, ge=100, le=10000, description="校验采样行数")
+    auto_chunking: bool = Field(default=True, description="大数据量自动分块处理")
+    chunk_max_columns: int = Field(default=20, ge=5, le=100, description="分块最大列数")
+    chunk_max_files: int = Field(default=5, ge=1, le=20, description="分块最大文件数")
+
+
+class ConfigGenerateRequest(BaseModel):
+    """配置生成请求"""
+
+    file_paths: list[str] = Field(..., description="数据文件路径列表")
+    project_name: str = Field(..., description="项目名称")
+    project_id: str = Field(..., description="项目标识")
+    provider_id: str | None = Field(default=None, description="指定AI Provider，不指定使用默认")
+    options: ConfigGenerateOptions = Field(default_factory=ConfigGenerateOptions)
+
+
+class ConfigGenerateResponse(BaseModel):
+    """配置生成响应（同步）"""
+
+    success: bool
+    yaml_preview: str = Field(..., description="生成的配置YAML预览")
+    manifest: dict[str, Any] | None = Field(default=None, description="项目清单")
+    schemas: dict[str, Any] | None = Field(default=None, description="表结构定义")
+    constraints: dict[str, Any] | None = Field(default=None, description="约束规则")
+    regex_nodes: dict[str, Any] | None = Field(default=None, description="正则节点")
+    warnings: list[str] = Field(default_factory=list)
+    error: str | None = None
+    iterations: int | None = Field(default=None, description="Agent 迭代轮数")
+    metrics: dict[str, Any] | None = Field(default=None, description="Agent 校验指标")
+
+
+class ConfigMigrateSource(BaseModel):
+    """单个迁移脚本来源"""
+
+    content: str = Field(..., description="脚本内容或自然语言描述")
+    language: str = Field(default="python", description="脚本类型: python/natural_language/excel_formula/sql")
+    name: str | None = Field(default=None, description="来源名称或文件路径，用于展示")
+
+
+class ConfigMigrateRequest(BaseModel):
+    """配置迁移请求"""
+
+    script_content: str = Field(default="", description="脚本内容或自然语言描述（兼容单来源）")
+    language: str = Field(default="python", description="脚本类型: python/natural_language/excel_formula/sql")
+    sources: list[ConfigMigrateSource] | None = Field(default=None, description="批量脚本来源列表")
+    file_paths: list[str] = Field(default_factory=list, description="数据文件路径列表")
+    project_name: str = Field(..., description="项目名称")
+    project_id: str = Field(..., description="项目标识")
+    provider_id: str | None = Field(default=None, description="指定AI Provider，不指定使用默认")
+    options: ConfigGenerateOptions = Field(default_factory=ConfigGenerateOptions)
+
+
+# =============================================================================
+# 异步任务模型
+# =============================================================================
+
+
+class ConfigGenerateJobCreateResponse(BaseModel):
+    """配置生成任务创建响应"""
+
+    job_id: str
+
+
+class ConfigGenerateJobStatus(BaseModel):
+    """配置生成任务状态"""
+
+    job_id: str
+    status: str = Field(..., description="pending/running/completed/failed/cancelled")
+    stage: str | None = Field(default=None, description="当前阶段")
+    message: str | None = Field(default=None, description="状态消息")
+    progress: float | None = Field(default=None, description="进度 0-100")
+    iterations: int | None = Field(default=None, description="当前已执行迭代轮数")
+    max_iterations: int | None = Field(default=None, description="最大迭代轮数")
+    metrics: dict[str, Any] | None = Field(default=None, description="校验指标")
+    current_plan: list[dict[str, Any]] | None = Field(default=None, description="当前执行计划")
+    checkpoints: list[dict[str, Any]] | None = Field(default=None, description="已保存的 checkpoints")
+    created_at: str
+    updated_at: str
+    warnings: list[str] = Field(default_factory=list)
+    error: str | None = None
+    result: ConfigGenerateResponse | None = None
+
+
+# =============================================================================
+# 硬件诊断模型
+# =============================================================================
+
+
+class HardwareInfo(BaseModel):
+    """硬件信息"""
+
+    name: str
+    value: Any
+
+
+class HardwareRequirement(BaseModel):
+    """硬件需求检查项"""
+
+    name: str
+    required: str
+    current: str | None = None
+    satisfied: bool = False
+
+
+class HardwareDiagnoseResponse(BaseModel):
+    """硬件诊断响应"""
+
+    platform: str
+    cpu: HardwareInfo | None = None
+    memory: HardwareInfo | None = None
+    disk: HardwareInfo | None = None
+    gpu: list[HardwareInfo] | None = None
+    requirements: list[HardwareRequirement] = Field(default_factory=list)
+    recommendations: list[str] = Field(default_factory=list)

app/api/routers/ai/ollama.py

@@ -0,0 +1,108 @@
+# SPDX-License-Identifier: Apache-2.0
+#
+# Copyright 2026 Precis Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+@fileoverview Ollama 服务管理 API 路由模块
+
+功能概述:
+- 获取 Ollama 本地服务的可用模型列表
+- 检查 Ollama 服务的健康状态
+- 从配置的 Ollama Provider 自动发现服务
+
+架构设计:
+- 扫描配置文件中的 ollama 类型 Provider
+- 通过 Provider 抽象层调用 list_models 和 health 方法
+- 失败时优雅降级，返回空列表或错误信息
+
+输入示例:
+    GET /ai/ollama/models
+
+输出示例:
+    ["llama3.2", "qwen2.5", "deepseek-coder"]
+"""
+
+import logging
+from typing import Any
+
+from app.shared.services.llm.config import loader
+from app.shared.services.llm.providers import create
+
+from .router import router
+
+
+@router.get(
+    "/ollama/models",
+    response_model=list[str],
+    summary="获取 Ollama 可用模型列表",
+    responses={
+        500: {"description": "服务器内部错误"},
+    },
+)
+async def list_ollama_models() -> list[str]:
+    """
+    获取 Ollama 可用模型列表
+
+    从配置的 Ollama Provider 获取模型列表
+    """
+    config = loader.load()
+
+    # 查找 Ollama Provider
+    ollama_provider = None
+    for p in config.providers:
+        if p.type.value == "ollama":
+            ollama_provider = p
+            break
+
+    if not ollama_provider:
+        return []
+
+    try:
+        provider = create(ollama_provider)
+        models = await provider.list_models()
+        return models
+    except Exception as exc:
+        logging.getLogger(__name__).warning("Failed to list Ollama models: %s", exc)
+        return []
+
+
+@router.get(
+    "/ollama/health",
+    summary="检查 Ollama 服务健康状态",
+    responses={
+        500: {"description": "服务器内部错误"},
+    },
+)
+async def ollama_health() -> dict[str, Any]:
+    """
+    检查 Ollama 服务健康状态
+    """
+    config = loader.load()
+
+    # 查找 Ollama Provider
+    ollama_provider = None
+    for p in config.providers:
+        if p.type.value == "ollama":
+            ollama_provider = p
+            break
+
+    if not ollama_provider:
+        return {"success": False, "error": "No Ollama provider configured"}
+
+    try:
+        provider = create(ollama_provider)
+        health = await provider.health()
+        return {"success": health.get("status") == "ok", "data": health}
+    except Exception as e:
+        return {"success": False, "error": str(e)}