Auto backup \2026-\04-\20

personal-blog/backend/routers/crawl.py

@@ -675,7 +675,67 @@ def download_image(url: str, temp_dir: str) -> tuple:
         return filename, filepath, True
     except Exception as e:
         print(f"下载图片失败 {url}: {e}")
-        return None, None, False
+    
+    # CDN 子域名回退（如 eewimg.cn 的 66 子域名在部分网络无法解析）
+    if 'eewimg.cn' in url and url.startswith('http'):
+        fallback_result = _try_eewimg_cdn_fallback(url, temp_dir)
+        if fallback_result and fallback_result[2]:
+            return fallback_result
+    
+    return None, None, False
+
+
+# eewimg.cn CDN 可解析的子域名（66 子域名在海外/部分网络 NXDOMAIN）
+_EEWIMG_FALLBACK_SUBDOMAINS = ["8", "2", "10", "11", "12", "3", "1"]
+
+
+def _try_eewimg_cdn_fallback(url: str, temp_dir: str) -> tuple:
+    """尝试用其他 eewimg.cn 子域名下载图片"""
+    # 去掉 imageView2 等图片处理参数，获取原始图片
+    clean_url = re.sub(r'\?.*$', '', url)
+    
+    for sub in _EEWIMG_FALLBACK_SUBDOMAINS:
+        try:
+            # 替换子域名（数字开头的子域名）
+            fallback_url = re.sub(r'https?://\d+\.eewimg\.cn', f'https://{sub}.eewimg.cn', clean_url)
+            if fallback_url == clean_url:
+                continue  # 没替换成功，跳过
+            
+            headers = {
+                'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
+                'Accept': 'image/avif,image/webp,image/apng,image/*,*/*;q=0.8',
+                'Referer': 'https://www.eeworld.com.cn/',
+            }
+            response = requests.get(fallback_url, headers=headers, timeout=15, stream=True)
+            response.raise_for_status()
+            
+            content_type = response.headers.get('content-type', '')
+            ext = '.jpg'
+            if 'png' in content_type:
+                ext = '.png'
+            elif 'gif' in content_type:
+                ext = '.gif'
+            elif 'webp' in content_type:
+                ext = '.webp'
+            else:
+                parsed = urlparse(url)
+                path_ext = os.path.splitext(parsed.path)[1]
+                if path_ext:
+                    ext = path_ext
+            
+            filename = f"crawl_{datetime.now().strftime('%Y%m%d_%H%M%S')}_{uuid.uuid4().hex[:8]}{ext}"
+            filepath = os.path.join(temp_dir, filename)
+            
+            with open(filepath, 'wb') as f:
+                for chunk in response.iter_content(chunk_size=8192):
+                    f.write(chunk)
+            
+            print(f"[CDN fallback] 成功通过 {sub}.eewimg.cn 下载: {url}")
+            return filename, filepath, True
+        except Exception as e:
+            continue
+    
+    return None, None, False
 
 
 def is_spa_site(html: str) -> bool:

personal-blog/backend/routers/universal_crawl.py

@@ -40,6 +40,20 @@ from .crawl import (
     CrawlImage,
 )
 
+
+class CrawlSaveRequest(BaseModel):
+    """通用采集保存请求"""
+    title: str
+    content: str
+    content_html: Optional[str] = ""
+    images: List[CrawlImage] = []
+    summary: Optional[str] = ""
+    category_id: Optional[int] = None
+    is_draft: bool = True
+    is_published: bool = False
+    tags: Optional[List[str]] = []
+    source_url: Optional[str] = None
+
 # 复用图片处理
 from .images import generate_thumbnail, get_image_dimensions
 
@@ -71,6 +85,7 @@ def _needs_browser(url: str) -> bool:
         'infoq.cn',
         'infoq.com',
         '51cto.com',
+        'eeworld.com.cn',   # JS WAF 防护，必须用浏览器渲染
     ]
     return any(d in domain for d in browser_domains)
 
@@ -466,29 +481,19 @@ async def universal_fetch(request: UniversalCrawlRequest):
 
 
 @router.post("/save")
-async def universal_save(
-    title: str,
-    content: str,
-    excerpt: Optional[str] = None,
-    category_id: Optional[int] = None,
-    source_url: Optional[str] = None,
-    images: Optional[str] = None,
-    db: Session = Depends(get_db)
-):
-    """保存通用采集的文章（复用 crawl.py 的保存逻辑）"""
+async def universal_save(request: CrawlSaveRequest, db: Session = Depends(get_db)):
+    """保存通用采集的文章，下载图片到本地并替换内容中的远程 URL"""
     try:
-        print(f"[Universal Crawl] Saving article: {title}")
-
-        # 解析图片列表
-        image_list = []
-        if images:
-            try:
-                import json
-                image_data = json.loads(images)
-                for img in image_data:
-                    image_list.append(CrawlImage(**img))
-            except Exception as e:
-                print(f"[Universal Crawl] Parse images error: {e}")
+        title = request.title
+        content = request.content
+        excerpt = request.summary or ""
+        category_id = request.category_id
+        source_url = request.source_url
+        image_list = request.images
+        is_draft = request.is_draft
+        is_published = request.is_published
+        
+        print(f"[Universal Crawl] Saving article: {title}, images: {len(image_list)}")
 
         # 创建临时目录
         temp_dir = tempfile.mkdtemp(dir=TEMP_DIR)
@@ -499,12 +504,17 @@ async def universal_save(
 
         # 下载图片
         downloaded_images = []
+        images_saved = 0
+        images_failed = 0
         for img in image_list:
             try:
-                filename, filepath, success = download_image(str(img.url), temp_dir)
+                img_url = str(img.original_url)
+                if not img_url:
+                    continue
+                filename, filepath, success = download_image(img_url, temp_dir)
                 if success and filepath:
                     # 移动最终文件
-                    final_filename = f"crawl_{datetime.now().strftime('%Y%m%d_%H%M%S')}_{uuid.uuid4().hex[:8]}{os.path.splitext(filename)[1]}"
+                    final_filename = f"crawl_{datetime.now().strftime('%Y%m%d_%H%M%S')}_{uuid.uuid4().hex[:8]}{os.path.splitext(filename or 'img.jpg')[1] or '.jpg'}"
                     final_path = final_dir / final_filename
                     os.rename(filepath, str(final_path))
 
@@ -521,7 +531,7 @@ async def universal_save(
                     # 保存到数据库
                     db_image = Image(
                         filename=final_filename,
-                        original_filename=img.alt or final_filename,
+                        original_filename=final_filename,
                         file_path=f"/uploads/images/{final_filename}",
                         file_size=os.path.getsize(str(final_path)),
                         mime_type=f"image/{os.path.splitext(final_filename)[1].lstrip('.') or 'jpeg'}",
@@ -533,9 +543,9 @@ async def universal_save(
                     db.flush()
 
                     # 替换内容中的图片 URL
-                    original_url = str(img.url)
                     new_url = f"/uploads/images/{final_filename}"
-                    content = content.replace(original_url, new_url)
+                    content = content.replace(img_url, new_url)
+                    images_saved += 1
 
                     downloaded_images.append({
                         "id": db_image.id,
@@ -543,25 +553,19 @@ async def universal_save(
                         "thumb": thumb_path_db,
                     })
             except Exception as e:
+                images_failed += 1
                 print(f"[Universal Crawl] Download image error: {e}")
                 continue
 
-        # 清理 excerpt
-        if excerpt:
-            excerpt = re.sub(r'<[^>]+>', '', excerpt)
-            excerpt = excerpt.replace('\n', ' ').strip()[:200]
-        else:
-            excerpt = content.replace('\n', ' ').strip()[:200]
+        # 生成摘要
+        summary = excerpt if excerpt else content[:300] if len(content) > 300 else content
+        summary = re.sub(r'[#*`\[\]\(\)!]', '', summary).strip()[:300]
 
         # 创建文章
         now = datetime.now()
         slug = re.sub(r'[^\w\u4e00-\u9fff-]', '-', title)[:50].strip('-')
         slug = f"{slug}-{now.strftime('%Y%m%d%H%M%S')}-{uuid.uuid4().hex[:4]}"
 
-        # 生成摘要
-        summary = content[:300] if len(content) > 300 else content
-        summary = re.sub(r'[#*`\[\]\(\)!]', '', summary)
-
         # 查找或创建分类
         final_category_id = category_id
         if not final_category_id:
@@ -576,28 +580,42 @@ async def universal_save(
             title=title,
             slug=slug,
             content=content,
-            excerpt=excerpt,
+            content_type="markdown",
             summary=summary,
             category_id=final_category_id,
-            author_id=1,
-            status="draft",
+            is_published=is_published,
+            is_draft=is_draft,
             view_count=0,
-            is_original=False,
-            source_url=source_url,
             created_at=now,
             updated_at=now,
-            published_at=None,
         )
         db.add(post)
         db.flush()
 
+        # 处理标签
+        if request.tags:
+            for tag_name in request.tags:
+                tag_name = tag_name.strip()
+                if not tag_name:
+                    continue
+                tag = db.query(Tag).filter(Tag.name == tag_name).first()
+                if not tag:
+                    tag_slug = re.sub(r'[^\w\u4e00-\u9fff-]', '-', tag_name).strip('-')
+                    tag = Tag(name=tag_name, slug=tag_slug)
+                    db.add(tag)
+                    db.flush()
+                post_tag = PostTag(post_id=post.id, tag_id=tag.id)
+                db.add(post_tag)
+
         db.commit()
 
         return {
-            "status": "success",
+            "success": True,
             "post_id": post.id,
+            "slug": slug,
             "message": "文章保存成功",
-            "images": downloaded_images,
+            "images_saved": images_saved,
+            "images_failed": images_failed,
         }
 
     except Exception as e:

personal-blog/frontend/src/pages/UniversalCrawlEdit.jsx

@@ -6,7 +6,6 @@ import { authApi } from '../api/auth'
 import { Card, CardHeader, CardTitle, CardDescription, CardBody, CardFooter } from '../components-preline/Card'
 import { Button } from '../components-preline/Button'
 import { Input, Textarea, Label, FormError } from '../components-preline/Input'
-import { postsApi } from '../api/posts'
 import { categoriesApi } from '../api/categories'
 import { tagsApi } from '../api/tags'
 import { API_BASE_URL } from '../api/config'
@@ -200,25 +199,55 @@ export function UniversalCrawlEdit() {
   // 保存文章
   const handleSubmit = async (e, isDraft = false) => {
     e.preventDefault()
-    
+
     if (!form.title.trim() || !form.content.trim()) {
       alert('请填写标题和内容')
       return
     }
-    
+
     setSaving(true)
     try {
-      const data = {
-        ...form,
-        category_id: form.category_id || null,
-        is_draft: isDraft,
-        is_published: !isDraft
+      const token = localStorage.getItem('token')
+      // 将 tag_ids 转换为 tag 名称数组
+      const tagNames = form.tag_ids.map(tid => {
+        const tag = tags.find(t => t.id === Number(tid))
+        return tag ? tag.name : ''
+      }).filter(Boolean)
+
+      const response = await fetch(`${API_BASE_URL}/universal-crawl/save`, {
+        method: 'POST',
+        headers: {
+          'Content-Type': 'application/json',
+          ...(token ? { 'Authorization': `Bearer ${token}` } : {})
+        },
+        body: JSON.stringify({
+          title: form.title,
+          content: form.content,
+          content_html: '',
+          images: crawledImages,
+          summary: form.summary,
+          category_id: form.category_id || null,
+          is_draft: isDraft,
+          is_published: !isDraft,
+          tags: tagNames,
+          source_url: crawlUrl
+        }),
+        credentials: 'include'
+      })
+
+      const data = await response.json()
+      if (!data.success) {
+        throw new Error(data.detail || data.message || '保存失败')
       }
-      
-      await postsApi.createPost(data)
+
+      // 提示图片下载情况
+      if (data.images_failed > 0) {
+        alert(`文章已保存成功！\n图片：${data.images_saved} 张下载成功，${data.images_failed} 张下载失败（保留原始链接）。`)
+      }
+
       navigate('/admin/posts')
     } catch (error) {
-      alert('保存失败: ' + (error.response?.data?.detail || '未知错误'))
+      alert('保存失败: ' + (error.message || error.response?.data?.detail || '未知错误'))
     } finally {
       setSaving(false)
     }

skills/baoyu-imagine/prompts/gao-shanwen-cover-system.md

@@ -0,0 +1,9 @@
+你是一位专业的信息图设计师。请根据用户提供的内容生成一张人物主题信息图。
+
+设计原则：
+- 整体色调：深蓝、藏青、金色渐变，体现庄重、专业、肃穆的氛围
+- 主体：一位经济学家形象的剪影/侧影，与数据图表、趋势线等宏观分析元素融合
+- 文字：可包含标题"高善文"以及关键词如"宏观研究""资产重估""因果推断"
+- 风格：严肃、专业、有纪念意义
+- 布局：人物肖像位于左侧或中央，右侧/周围环绕经济数据图表元素
+- 避免过于具体的数字和统计数据，使用抽象概念表达

skills/baoyu-post-to-wechat/SKILL.md

@@ -273,6 +273,7 @@ python3 baoyu_with_image_gen.py markdown /home/ubuntu/test_ai_images.md --genera
 5. **凭证安全**：API凭证已加密存储
 6. **AI图片生成**：使用豆包Seedream API，需要网络连接，生成时间约30-60秒/张
 7. **Frontmatter格式**：使用标准YAML格式，注意缩进
+8. **API 48001 限制**：微信API仅支持创建草稿，不支持自动发布。所有公众号文章需在微信后台手动发布。博客同步不受此限制。
 
 ---