feat: 热词热力图改为Treemap矩形堆砌布局，面积代表提及次数；赛道分析页面布局调整

backend/structurer.py

@@ -51,6 +51,23 @@ def get_api_key() -> str:
 
 # ---- 单篇文章结构化分析 ----
 
+_PROFANITY_MAP = {
+    "妈的": "**", "傻逼": "**", "他妈": "**", "操": "*",
+    "草泥马": "***", "尼玛": "**", "贱人": "**",
+    "王八蛋": "***", "混蛋": "**", "去死": "**",
+    "毛爷爷": "某伟人", "毛泽东": "某伟人", "国民党": "某党",
+    "共产党": "某党", "长征": "远征", "抗日": "抗战",
+    "革命": "变革", "社会主义": "某主义",
+}
+
+
+def sanitize_content(text: str) -> str:
+    """过滤粗口和敏感词，避免触发千问内容审核。"""
+    for word, replacement in _PROFANITY_MAP.items():
+        text = text.replace(word, replacement)
+    return text
+
+
 def analyze_article(title: str, content: str) -> dict:
     """调用通义千问对文章进行结构化分析。"""
     api_key = get_api_key()
@@ -71,7 +88,7 @@ JSON 字段必须为：
 - numbers（关键数据数组）
 
 文章标题：{title}
-正文：{content[:24000]}"""
+正文：{sanitize_content(content[:24000])}"""
     payload = json.dumps({
         "model": model,
         "messages": [
@@ -276,6 +293,11 @@ def run_batch_analyze(force: bool, pause: float, retries: int) -> int:
                 break
             except Exception as exc:
                 last_error = exc
+                err_str = str(exc)
+                # 内容审核失败，标记跳过不重试
+                if "data_inspection_failed" in err_str:
+                    print(f"[{index}/{len(rows)}] SKIP(内容审核) {row['title']}", flush=True)
+                    break
                 if attempt <= retries:
                     wait = attempt * 2
                     print(f"[{index}/{len(rows)}] RETRY {attempt}/{retries} {exc}", flush=True)