Fix: add paper role to auto tagger

Author: actions-user

scripts/auto_tagger.py

@@ -8,88 +8,175 @@
 API_URL = "https://api.deepseek.com/chat/completions"
 POSTS_DIR = "_posts"
 
+
 def get_existing_tags():
+    """读取已有的全部标签"""
     all_tags = set()
-    if not os.path.exists(POSTS_DIR): return []
+    if not os.path.exists(POSTS_DIR):
+        return []
+
     for filename in os.listdir(POSTS_DIR):
         if filename.endswith(".md"):
             filepath = os.path.join(POSTS_DIR, filename)
             try:
-                with open(filepath, 'r', encoding='utf-8') as f:
+                with open(filepath, "r", encoding="utf-8") as f:
                     content = f.read()
-                parts = content.split('---', 2)
+                
+                parts = content.split("---", 2)
                 if len(parts) >= 3:
                     front_matter = yaml.safe_load(parts[1])
+                    if not front_matter:
+                        continue
+                        
                     tags = front_matter.get("tags")
                     if isinstance(tags, list):
-                        for t in tags: all_tags.add(t)
-            except: continue
+                        for t in tags:
+                            all_tags.add(t)
+            except Exception:
+                continue
+
     return sorted(list(all_tags))
 
+
 def get_tags_from_ai(title, content, category, existing_tags):
-    if not API_KEY: return []
-    headers = {"Content-Type": "application/json", "Authorization": f"Bearer {API_KEY}"}
-    
+    """调用大模型为文章自动生成标签"""
+    if not API_KEY:
+        print("未检测到 DEEPSEEK_API_KEY，跳过 AI 打标签。")
+        return []
+
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": f"Bearer {API_KEY}"
+    }
+
+    role_map = {
+        "study": "资深技术专家",
+        "anime": "资深二次元漫评人",
+        "music": "专业乐评人",
+        "paint": "美术鉴赏家",
+        "game": "骨灰级游戏玩家",
+        "snap": "专业摄影师",
+        "asmr": "ASMR深度体验者",
+        "emo": "情感作家",
+        "paper": "资深学术研究员"
+    }
+    role = role_map.get(category, "专业博客编辑")
+
     guidance = ""
-    if category == "study": guidance = "这是一篇学习笔记。请侧重提取技术领域词。"
-    elif category == "anime": guidance = "这是一篇动漫相关博文。请务必提取作品名称作为首个标签、要根据该动漫的真实内容来提取其他标签。"
-    elif category == "music": guidance = "这是一篇音乐鉴赏。请提取社团/作者名、曲风等。"
-    elif category == "paint": guidance = "这是一篇绘画分享。请提取其中的人物、风格等。"
-    elif category == "game": guidance = "这是一篇游戏记录。请务必提取游戏名称作为首个标签、要根据游戏的真实内容来提取其他标签。"
-    elif category == "snap": guidance = "这是一篇摄影作品。请提取镜头焦段、拍摄地点等。"
-    elif category == "asmr": guidance = "这是一篇助眠相关内容。请提取作者名等。"
-    elif category == "emo": guidance = "这是一篇心情随笔。请提取情感意象或核心感悟。"
+    if category == "study":
+        guidance = "这是一篇学习笔记。请侧重提取核心技术栈、框架、编程语言等专业词汇。"
+    elif category == "anime":
+        guidance = "这是一篇动漫相关博文。请务必提取作品名称作为首个标签，并提取核心角色、制作公司、类型题材（如机战、日常）等。"
+    elif category == "music":
+        guidance = "这是一篇音乐鉴赏。请提取歌手/社团/作者名、曲风、专辑名等。"
+    elif category == "paint":
+        guidance = "这是一篇绘画分享。请提取其中的人物名字、画师名、艺术风格、绘制工具等。"
+    elif category == "game":
+        guidance = "这是一篇游戏记录。请务必提取游戏名称作为首个标签，并提取游戏类型、核心机制或开发商等。"
+    elif category == "snap":
+        guidance = "这是一篇摄影作品。请提取相机型号、镜头焦段、拍摄地点、摄影风格等。"
+    elif category == "asmr":
+        guidance = "这是一篇助眠相关内容。请提取音声作者名、触发音类型（如耳语、心跳、底噪）、设备等。"
+    elif category == "emo":
+        guidance = "这是一篇心情随笔。请提取抽象的情感意象或核心感悟名词。"
+    elif category == "paper":
+        guidance = "这是一篇学术论文阅读笔记或研究总结。请务必提取论文的研究领域、核心算法、关键模型（如Transformer、CNN）、数据集名称等学术专业词汇。"
 
-    prompt = f"角色：专业技术博客编辑。任务：为文章生成 3-5 个精准标签。\n分类背景：{guidance}\n要求：1. 可以复用已有标签：{', '.join(existing_tags)}\n2. 格式：JSON 数组。\n标题：{title}\n摘要：{content[:1000]}"
+    prompt = f"""角色：{role}。
+        任务：请仔细阅读以下文章内容，为其生成 3-5 个最核心、最准确的标签。
+        分类指导：{guidance}
+        严格要求：
+            1. 尽可能复用现有标签库中的标签：{', '.join(existing_tags)}
+            2. 标签必须是具体的专有名词（如人名、技术名、作品名、算法名），绝对禁止使用宽泛的形容词（如“好看的”、“好听的”）或长句！
+            3. 必须以 JSON 对象格式返回，键名固定为 "tags"，值为字符串数组。
+
+        文章标题：{title}
+        内容摘要：{content[:3000]}"""
 
     payload = {
         "model": "deepseek-chat",
         "messages": [
-            {"role": "system", "content": "You are an intelligent tag generator. Output JSON arrays only."},
-            {"role": "user", "content": prompt}
+            {
+                "role": "system",
+                "content": 'You are a precise and professional tag extractor. You must reply ONLY with a valid JSON object containing a "tags" string array. Do not include markdown formatting like ```json.'
+            },
+            {
+                "role": "user",
+                "content": prompt
+            }
         ],
-        "response_format": {"type": "json_object"}
+        "response_format": {"type": "json_object"},
+        "temperature": 0.1
     }
 
     try:
         response = requests.post(API_URL, headers=headers, json=payload, timeout=30)
+        response.raise_for_status()
+        
         data = response.json()
         res = data['choices'][0]['message']['content']
         parsed = json.loads(res)
-        return parsed["tags"] if isinstance(parsed, dict) and "tags" in parsed else parsed
-    except: return []
+        
+        tags = parsed.get("tags")
+        if isinstance(tags, list):
+            return tags
+        elif isinstance(parsed, list):
+            return parsed
+        return []
+    except Exception as e:
+        print(f"Error calling AI: {e}")
+        return []
+
 
 def process_posts():
-    if not os.path.exists(POSTS_DIR): return
+    """遍历文章目录并为缺失标签的文章补充标签"""
+    if not os.path.exists(POSTS_DIR):
+        print(f"文章目录 {POSTS_DIR} 不存在。")
+        return
+
     existing_tags = get_existing_tags()
 
     for filename in os.listdir(POSTS_DIR):
-        if not filename.endswith(".md") or filename == "BLOG_TEMPLATE.md": continue
+        if not filename.endswith(".md") or filename == "BLOG_TEMPLATE.md":
+            continue
+
         filepath = os.path.join(POSTS_DIR, filename)
-        with open(filepath, 'r', encoding='utf-8') as f:
+        
+        with open(filepath, "r", encoding="utf-8") as f:
             full_content = f.read()
 
-        parts = full_content.split('---', 2)
-        if len(parts) < 3: continue
+        parts = full_content.split("---", 2)
+        if len(parts) < 3:
+            continue
 
         try:
             front_matter = yaml.safe_load(parts[1])
             post_body = parts[2]
-        except: continue
+        except Exception:
+            continue
 
-        if not front_matter.get("tags") or len(front_matter["tags"]) == 0:
+        if not front_matter or not front_matter.get("tags") or len(front_matter["tags"]) == 0:
+            if not front_matter:
+                front_matter = {}
+                
             title = front_matter.get("title", "")
-            category = (front_matter.get("categories") or [""])[0].lower()
+            categories = front_matter.get("categories", [])
+            category = categories[0].lower() if categories else ""
 
             print(f"Intelligently tagging: {title}...")
             new_tags = get_tags_from_ai(title, post_body.strip(), category, existing_tags)
+            
             if new_tags:
                 front_matter["tags"] = new_tags
-                # 重新写入文件，保持日期原封不动
-                new_content = f"---\n{yaml.dump(front_matter, allow_unicode=True, sort_keys=False).strip()}\n---\n{post_body}"
-                with open(filepath, 'w', encoding='utf-8') as f:
+                
+                
+                yaml_content = yaml.dump(front_matter, allow_unicode=True, sort_keys=False).strip()
+                new_content = f"---\n{yaml_content}\n---\n{post_body}"
+                
+                with open(filepath, "w", encoding="utf-8") as f:
                     f.write(new_content)
                 print(f"Updated tags for: {filename}")
 
+
 if __name__ == "__main__":
-    process_posts()
+    process_posts()