Merge pull request #116 from DoiiarX/fix-news-id

修复news-id冲突问题

Merge pull request #116 from DoiiarX/fix-news-id
修复news-id冲突问题
Doiiars · GitHub
Commit f0772bfe6eaaeed2990a935784c1aa6d08f271c2 f0772bfe 2 parents 883a405c b774e5d1
Showing 1 changed file with 23 additions and 8 deletions
MindSpider/BroadTopicExtraction/database_manager.py
--- a/MindSpider/BroadTopicExtraction/database_manager.py
View file @f0772bf
+++ b/MindSpider/BroadTopicExtraction/database_manager.py
View file @f0772bf
@@ -25,6 +25,7 @@ except ImportError:
 from config import settings
+
 class DatabaseManager:
     """数据库管理器"""
@@ -46,14 +47,15 @@ class DatabaseManager:
         except ModuleNotFoundError as e:
             missing: str = str(e)
             if "psycopg" in missing:
-                logger.error("数据库连接失败: 未安装PostgreSQL驱动 psycopg。请安装: psycopg[binary]。参考指令：uv pip install psycopg[binary]")
+                logger.error(
+                    "数据库连接失败: 未安装PostgreSQL驱动 psycopg。请安装: psycopg[binary]。参考指令：uv pip install psycopg[binary]")
             elif "pymysql" in missing:
                 logger.error("数据库连接失败: 未安装MySQL驱动 pymysql。请安装: pymysql。参考指令：uv pip install pymysql")
             else:
                 logger.error(f"数据库连接失败(缺少驱动): {e}")
             raise
         except Exception as e:
-            logger.error(f"数据库连接失败: {e}")
+            logger.exception(f"数据库连接失败: {e}")
             raise
     def close(self):
@@ -97,7 +99,13 @@ class DatabaseManager:
             # 逐条插入，单条失败不影响后续（每条独立事务）
             for news_item in news_data:
                 try:
-                    news_id = f"{news_item.get('source', 'unknown')}_{news_item.get('id', news_item.get('rank', 0))}"
+                    # news_item.get('id') 已经是完整的 news_id（格式：source_item_id）
+                    # 为了支持同一条新闻在不同日期出现，将 crawl_date 加入到 news_id 中
+                    base_news_id = news_item.get(
+                        'id') or f"{news_item.get('source', 'unknown')}_rank_{news_item.get('rank', 0)}"
+                    # 将日期格式化为字符串并加入到 news_id 中，确保全局唯一性
+                    news_id = f"{base_news_id}_{crawl_date.strftime('%Y%m%d')}"
+
                     title_val = (news_item.get("title", "") or "")
                     if len(title_val) > 500:
                         title_val = title_val[:500]
@@ -124,7 +132,7 @@ class DatabaseManager:
                         )
                     saved_count += 1
                 except Exception as e:
-                    logger.warning(f"保存单条新闻失败: {e}")
+                    logger.exception(f"保存单条新闻失败: {e}")
                     continue
             logger.info(f"成功保存 {saved_count} 条新闻记录")
             return saved_count
@@ -174,17 +182,21 @@ class DatabaseManager:
         try:
             keywords_json = json.dumps(keywords, ensure_ascii=False)
+            # 为了支持外键引用，topic_id 需要全局唯一，所以将日期加入到 topic_id 中
+            topic_id = f"summary_{extract_date.strftime('%Y%m%d')}"
+
             with self.engine.begin() as conn:
                 check = conn.execute(
                     text("SELECT id FROM daily_topics WHERE extract_date = :d AND topic_id = :tid"),
-                    {"d": extract_date, "tid": "summary"},
+                    {"d": extract_date, "tid": topic_id},
                 ).first()
                 if check:
                     conn.execute(
                         text(
                             "UPDATE daily_topics SET keywords = :k, topic_description = :s, add_ts = :ts, last_modify_ts = :lmt, topic_name = :tn WHERE extract_date = :d AND topic_id = :tid"
                         ),
-                        {"k": keywords_json, "s": summary, "ts": current_timestamp, "lmt": current_timestamp, "d": extract_date, "tid": "summary", "tn": "每日新闻分析"},
+                        {"k": keywords_json, "s": summary, "ts": current_timestamp, "lmt": current_timestamp,
+                         "d": extract_date, "tid": topic_id, "tn": "每日新闻分析"},
                     )
                     logger.info(f"更新了 {extract_date} 的话题分析")
                 else:
@@ -192,7 +204,8 @@ class DatabaseManager:
                         text(
                             "INSERT INTO daily_topics (extract_date, topic_id, topic_name, keywords, topic_description, add_ts, last_modify_ts) VALUES (:d, :tid, :tn, :k, :s, :ts, :lmt)"
                         ),
-                        {"d": extract_date, "tid": "summary", "tn": "每日新闻分析", "k": keywords_json, "s": summary, "ts": current_timestamp, "lmt": current_timestamp},
+                        {"d": extract_date, "tid": topic_id, "tn": "每日新闻分析", "k": keywords_json, "s": summary,
+                         "ts": current_timestamp, "lmt": current_timestamp},
                     )
                     logger.info(f"保存了 {extract_date} 的话题分析")
             return True
@@ -215,7 +228,8 @@ class DatabaseManager:
         try:
             with self.engine.connect() as conn:
-                result = conn.execute(text("SELECT * FROM daily_topics WHERE extract_date = :d"), {"d": extract_date}).mappings().first()
+                result = conn.execute(text("SELECT * FROM daily_topics WHERE extract_date = :d"),
+                                      {"d": extract_date}).mappings().first()
                 if result:
                     result = dict(result)  # 转为可变dict以支持item赋值
                     result["keywords"] = json.loads(result["keywords"]) if result.get("keywords") else []
@@ -290,6 +304,7 @@ class DatabaseManager:
             logger.exception(f"获取统计摘要失败: {e}")
             return {"news_stats": [], "topics_stats": []}
+
 if __name__ == "__main__":
     # 测试数据库管理器
     with DatabaseManager() as db: