'commit'

Controller/DouYinController.py

@@ -8,6 +8,7 @@ from datetime import datetime
 from typing import List, Optional
 
 from fastapi import APIRouter, HTTPException, BackgroundTasks
+from fastapi.responses import StreamingResponse
 from pydantic import BaseModel
 import pymysql
 
@@ -38,6 +39,9 @@ def get_db_connection():
 class ParseRequest(BaseModel):
     text: str
 
+class SummaryRequest(BaseModel):
+    ids: List[str] = []
+
 def update_status(id, status, error_msg=None):
     try:
         conn = get_db_connection()
@@ -258,6 +262,64 @@ def delete_record(id: str):
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))
 
+@router.post("/api/douyin/summary")
+async def generate_summary(request: SummaryRequest):
+    try:
+        # Fetch transcripts
+        conn = get_db_connection()
+        cursor = conn.cursor()
+        
+        if request.ids:
+            # Secure way to handle list in SQL
+            format_strings = ','.join(['%s'] * len(request.ids))
+            sql = f"SELECT video_name, transcript FROM t_douyin_record WHERE id IN ({format_strings}) AND status='COMPLETED'"
+            cursor.execute(sql, tuple(request.ids))
+        else:
+            # Default to latest 20
+            cursor.execute("SELECT video_name, transcript FROM t_douyin_record WHERE status='COMPLETED' ORDER BY create_time DESC LIMIT 20")
+        
+        records = cursor.fetchall()
+        conn.close()
+        
+        if not records:
+            # If no records, just return a simple message stream
+            async def empty_stream():
+                yield "未找到可总结的已完成记录，请先解析视频。"
+            return StreamingResponse(empty_stream(), media_type="text/event-stream")
+            
+        # Prepare text
+        full_text = ""
+        for r in records:
+            if r['transcript']:
+                full_text += f"【标题：{r['video_name']}】\n内容：{r['transcript']}\n\n"
+                
+        if not full_text:
+             async def empty_text_stream():
+                yield "记录中没有有效的文案内容。"
+             return StreamingResponse(empty_text_stream(), media_type="text/event-stream")
+
+        # Prompt
+        prompt = f"""
+        请对以下充电行业相关的视频内容进行知识精华提取。
+        要求：
+        1. 忽略无关闲聊和口语化表达；
+        2. 按条目列出核心知识点，不要长篇大论；
+        3. 保持简洁专业，只保留干货；
+        4. 返回格式为Markdown列表。
+        
+        内容如下：
+        {full_text[:15000]}
+        """
+        
+        # Limit context to avoid errors, 15000 chars is roughly safe for most models, 
+        # but if using a small model, might need less. Assuming robust model.
+        
+        return StreamingResponse(get_llm_response(prompt), media_type="text/event-stream")
+        
+    except Exception as e:
+        logger.error(f"Summary generation failed: {e}")
+        raise HTTPException(status_code=500, detail=str(e))
+
 async def recover_pending_tasks():
     """
     Check for tasks stuck in PENDING or PROCESSING state (due to server restart)