添加Ai 能力，重构项目结构

.gitignore

@@ -1,4 +1,7 @@
 .venv/
+.env
+.env.test
+.env.prod
 .idea/
 .vscode/
 .history/

app/__init__.py

@@ -0,0 +1 @@
+"""OfferPie 爬虫应用包。"""

app/ai/__init__.py

@@ -0,0 +1 @@
+"""AI 模型配置。"""

app/ai/model_config.py

@@ -0,0 +1,13 @@
+"""AI 模型场景配置。
+
+集中管理爬虫服务使用的模型，切换模型只需修改此文件。
+"""
+
+from app.ai.models import LLM
+
+
+class SpiderModel:
+    """爬虫 AI 场景模型。"""
+
+    # 页面内容理解与结构化提取
+    CONTENT_EXTRACT = LLM.CLAUDE_OPUS.create(temperature=0)

app/ai/models.py

@@ -0,0 +1,64 @@
+"""LLM 模型枚举与实例获取。
+
+Usage:
+    from app.ai.models import LLM
+
+    llm = LLM.DOUBAO_SEED_LITE.create(temperature=0)
+    llm = LLM.CLAUDE_OPUS.create(temperature=0)
+"""
+
+from enum import Enum
+from typing import Callable
+
+from langchain_anthropic import ChatAnthropic
+from langchain_core.language_models import BaseChatModel
+from langchain_openai import ChatOpenAI
+
+from app.config import settings
+
+ConfigGetter = Callable[[], str]
+
+# 供应商连接配置 = (api_key函数, base_url函数)
+_VOLCENGINE = (
+    lambda: settings.volcengine_api_key,
+    lambda: settings.volcengine_base_url,
+)
+_ANTHROPIC = (
+    lambda: settings.anthropic_api_key,
+    lambda: settings.anthropic_base_url,
+)
+
+
+class LLM(Enum):
+    """所有可用模型，每个枚举值 = (模型名, 封装类, api_key函数, base_url函数)。"""
+
+    # 火山引擎（OpenAI 兼容）
+    DOUBAO_PRO_32K = ("doubao-1-5-pro-32k-250115", ChatOpenAI, *_VOLCENGINE)
+    DOUBAO_LITE_32K = ("doubao-1-5-lite-32k-250115", ChatOpenAI, *_VOLCENGINE)
+    DOUBAO_SEED_LITE = ("doubao-seed-2-0-lite-260215", ChatOpenAI, *_VOLCENGINE)
+    DOUBAO_SEED_PRO = ("doubao-seed-2-0-pro-260215", ChatOpenAI, *_VOLCENGINE)
+    DEEPSEEK_V4_FLASH = ("deepseek-v4-flash-260425", ChatOpenAI, *_VOLCENGINE)
+
+    # Claude（Anthropic 风格）
+    CLAUDE_OPUS = ("claude-opus-4-6", ChatAnthropic, *_ANTHROPIC)
+
+    def __init__(
+        self,
+        model_name: str,
+        model_class: type[BaseChatModel],
+        api_key_getter: ConfigGetter,
+        base_url_getter: ConfigGetter,
+    ) -> None:
+        self.model_name = model_name
+        self._model_class = model_class
+        self._api_key_getter = api_key_getter
+        self._base_url_getter = base_url_getter
+
+    def create(self, **kwargs) -> BaseChatModel:
+        """创建模型实例，其他参数透传给 LangChain 模型类。"""
+        return self._model_class(
+            model=self.model_name,
+            api_key=self._api_key_getter(),
+            base_url=self._base_url_getter(),
+            **kwargs,
+        )

app/config/__init__.py

@@ -0,0 +1,18 @@
+import os
+from pathlib import Path
+
+from .settings import Settings
+
+_env = os.getenv("ENV", "dev")
+_env_files = {"dev": ".env", "test": ".env.test", "prod": ".env.prod"}
+
+# 定位项目根目录（config 上两级）
+_project_root = Path(__file__).resolve().parent.parent.parent
+_env_file = _project_root / _env_files.get(_env, ".env")
+
+if not _env_file.exists():
+    raise FileNotFoundError(f".env 文件不存在: {_env_file}")
+
+settings = Settings(_env_file=str(_env_file))
+
+__all__ = ["settings"]

app/config/settings.py

@@ -0,0 +1,22 @@
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    """项目配置，通过环境变量或项目根目录下的 .env 文件覆盖。"""
+
+    env: str = "dev"
+
+    # 火山引擎（OpenAI 兼容风格）
+    volcengine_api_key: str = ""
+    volcengine_base_url: str = "https://ark.cn-beijing.volces.com/api/v3"
+
+    # Claude（Anthropic 风格）
+    anthropic_api_key: str = ""
+    anthropic_base_url: str = ""
+
+    model_config = SettingsConfigDict(
+        env_file=".env",
+        env_file_encoding="utf-8",
+        case_sensitive=False,
+        extra="ignore",
+    )

app/tool/__init__.py

@@ -0,0 +1 @@
+"""爬虫工具集。"""

requirements.txt

@@ -3,3 +3,7 @@ rapidocr>=3.9
 onnxruntime>=1.17
 opencv-python>=4.5
 playwright>=1.49
+pydantic-settings>=2.0
+langchain-openai>=0.3
+langchain-anthropic>=0.3
+langchain-core>=0.3