feat: 接入 DeepSeek AI 分析公告是否为中国电信可承接项目

- 新增 ai_enabled/ai_api_key/ai_base_url/ai_model 等配置项，通过 .env 管理
- 新增 app/services/ai_analyzer.py — DeepSeek API 调用 + 详情页正文提取
- 新增 extract_page_content() 从详情页抓取正文供 AI 分析
- Announcement 模型新增 ai_relevant / ai_analysis 字段
- 流水线集成 AI 分析步骤（关键词匹配后、通知前）
- AI 标记为可承接的项目额外发送 markdown 着重通知
- 创建 alembic 迁移版本 6e8f4c2d1b0a

app/crawler/parsers.py

@@ -149,6 +149,57 @@ def parse_dahuagov_detail_pubdate(html: str) -> datetime | None:
     return None
 
 
+async def extract_page_content(url: str, timeout: int = 30) -> str | None:
+    """抓取详情页并用 BeautifulSoup 提取正文纯文本"""
+    import httpx
+
+    try:
+        headers = {
+            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
+            "Accept": "text/html,application/xhtml+xml",
+            "Accept-Language": "zh-CN,zh;q=0.9",
+        }
+        async with httpx.AsyncClient(timeout=timeout, follow_redirects=True) as client:
+            resp = await client.get(url, headers=headers)
+            if resp.status_code != 200:
+                return None
+
+            soup = BeautifulSoup(resp.text, "html.parser")
+
+            # 尝试多种常见正文容器选择器
+            selectors = [
+                "div.article-content", "div.content", "div.TRS_Editor",
+                "div.Custom_UnionStyle", "div.pages_content", "div#content",
+                "div.main-content", "article", ".article", ".detail-content",
+                ".text-content", ".news-content", ".detail-article",
+            ]
+            for selector in selectors:
+                container = soup.select_one(selector)
+                if container:
+                    # 移除脚本和样式
+                    for tag in container.find_all(["script", "style"]):
+                        tag.decompose()
+                    text = container.get_text(separator="\n", strip=True)
+                    if len(text) > 50:  # 至少50字才算有效正文
+                        return text
+
+            # 兜底：取 body 内所有文本
+            body = soup.find("body")
+            if body:
+                for tag in body.find_all(["script", "style", "nav", "footer", "header"]):
+                    tag.decompose()
+                text = body.get_text(separator="\n", strip=True)
+                # 移除过长的空白行
+                lines = [l.strip() for l in text.split("\n") if l.strip()]
+                text = "\n".join(lines[:200])  # 最多取前200行
+                if len(text) > 50:
+                    return text
+
+            return None
+    except Exception:
+        return None
+
+
 def _generate_hash(ann: dict[str, Any]) -> str:
     content = (
         f"{ann['title']}|{ann['publish_date'].strftime('%Y-%m-%d')}"