tingwu_new

backend/ai_services/management/commands/test_tingwu_local.py

@@ -0,0 +1,102 @@
+import os
+import sys
+import django
+import json
+import logging
+from django.conf import settings
+
+# 设置 Django 环境
+# 添加项目根目录到 sys.path
+sys.path.append('/Volumes/data/Quant-Speed/market_page/backend')
+os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'config.settings') # 修正为正确的 settings 模块路径
+django.setup()
+
+from ai_services.services import AliyunTingwuService
+from ai_services.models import TranscriptionTask
+
+# 配置日志
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+def test_tingwu_transcription():
+    file_url = "https://tangledup-ai-staging.oss-cn-shanghai.aliyuncs.com/Video/%E6%95%99%E5%AD%A6.mp4"
+    
+    print(f"Testing transcription for: {file_url}")
+    
+    service = AliyunTingwuService()
+    
+    # 1. 创建任务
+    try:
+        print("Creating task...")
+        response = service.create_transcription_task(file_url)
+        print(f"Create task response: {json.dumps(response, indent=2, ensure_ascii=False)}")
+        
+        if 'Data' in response and isinstance(response['Data'], dict):
+            task_id = response['Data'].get('TaskId')
+        else:
+            task_id = response.get('TaskId')
+            
+        if not task_id:
+            print("Failed to get TaskId")
+            return
+            
+        print(f"Task created with ID: {task_id}")
+        
+        # 2. 轮询查询任务状态
+        import time
+        max_retries = 60 # 5 minutes
+        for i in range(max_retries):
+            print(f"Checking status (attempt {i+1}/{max_retries})...")
+            result = service.get_task_info(task_id)
+            
+            # 解析结果
+            if isinstance(result, str):
+                try:
+                    result = json.loads(result)
+                except:
+                    pass
+            
+            if isinstance(result, dict):
+                data_obj = result.get('Data', result)
+            else:
+                data_obj = result
+                
+            task_status = data_obj.get('TaskStatus')
+            if not task_status:
+                task_status = data_obj.get('Status')
+                
+            print(f"Current status: {task_status}")
+            
+            if task_status in ['COMPLETE', 'COMPLETED', 'SUCCEEDED']:
+                print("Task succeeded!")
+                print(f"Full Result: {json.dumps(data_obj, indent=2, ensure_ascii=False)}")
+                
+                # 尝试解析 Transcription
+                task_result = data_obj.get('Result', {})
+                transcription_data = task_result.get('Transcription', {})
+                
+                if isinstance(transcription_data, str) and transcription_data.startswith('http'):
+                    import requests
+                    print(f"Downloading transcription from {transcription_data}")
+                    t_resp = requests.get(transcription_data)
+                    if t_resp.status_code == 200:
+                        content = t_resp.json()
+                        print(f"Downloaded content structure keys: {content.keys()}")
+                        # print(f"Content sample: {json.dumps(content, indent=2, ensure_ascii=False)[:500]}...")
+                    else:
+                        print(f"Failed to download: {t_resp.status_code}")
+                
+                break
+            elif task_status == 'FAILED':
+                print(f"Task failed: {data_obj}")
+                break
+                
+            time.sleep(5)
+            
+    except Exception as e:
+        print(f"Error: {e}")
+        import traceback
+        traceback.print_exc()
+
+if __name__ == "__main__":
+    test_tingwu_transcription()

backend/ai_services/migrations/0003_transcriptiontask_auto_chapters_data_and_more.py

@@ -0,0 +1,28 @@
+# Generated by Django 6.0.1 on 2026-03-11 12:30
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('ai_services', '0002_transcriptiontask_evaluation_transcriptiontask_score'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='transcriptiontask',
+            name='auto_chapters_data',
+            field=models.JSONField(blank=True, help_text='阿里云返回的AutoChapters完整JSON', null=True, verbose_name='章节原始数据'),
+        ),
+        migrations.AddField(
+            model_name='transcriptiontask',
+            name='summary_data',
+            field=models.JSONField(blank=True, help_text='阿里云返回的Summarization完整JSON', null=True, verbose_name='总结原始数据'),
+        ),
+        migrations.AddField(
+            model_name='transcriptiontask',
+            name='transcription_data',
+            field=models.JSONField(blank=True, help_text='阿里云返回的Transcription完整JSON', null=True, verbose_name='转写原始数据'),
+        ),
+    ]

backend/ai_services/models.py

@@ -18,6 +18,11 @@ class TranscriptionTask(models.Model):
         choices=Status.choices,
         default=Status.PENDING
     )
+    # 存储阿里云听悟返回的原始 JSON 结构
+    transcription_data = models.JSONField(verbose_name=_('转写原始数据'), blank=True, null=True, help_text=_('阿里云返回的Transcription完整JSON'))
+    summary_data = models.JSONField(verbose_name=_('总结原始数据'), blank=True, null=True, help_text=_('阿里云返回的Summarization完整JSON'))
+    auto_chapters_data = models.JSONField(verbose_name=_('章节原始数据'), blank=True, null=True, help_text=_('阿里云返回的AutoChapters完整JSON'))
+    
     transcription = models.TextField(verbose_name=_('逐字稿'), blank=True, null=True)
     summary = models.TextField(verbose_name=_('AI总结'), blank=True, null=True)
     score = models.IntegerField(verbose_name=_('AI评分'), blank=True, null=True, help_text=_('基于转写内容的评分'))

backend/ai_services/serializers.py

@@ -4,8 +4,8 @@ from .models import TranscriptionTask
 class TranscriptionTaskSerializer(serializers.ModelSerializer):
     class Meta:
         model = TranscriptionTask
-        fields = ['id', 'file_url', 'task_id', 'status', 'transcription', 'summary', 'error_message', 'created_at', 'updated_at', 'score', 'evaluation']
+        fields = ['id', 'file_url', 'task_id', 'status', 'transcription', 'summary', 'error_message', 'created_at', 'updated_at', 'score', 'evaluation', 'transcription_data', 'summary_data', 'auto_chapters_data']
-        read_only_fields = ['id', 'file_url', 'task_id', 'status', 'transcription', 'summary', 'error_message', 'created_at', 'updated_at', 'score', 'evaluation']
+        read_only_fields = ['id', 'file_url', 'task_id', 'status', 'transcription', 'summary', 'error_message', 'created_at', 'updated_at', 'score', 'evaluation', 'transcription_data', 'summary_data', 'auto_chapters_data']
 
 class TranscriptionUploadSerializer(serializers.Serializer):
     file = serializers.FileField(help_text="上传的音频文件")

backend/ai_services/views.py

@@ -218,6 +218,8 @@ class TranscriptionTaskViewSet(viewsets.ModelViewSet):
                         if t_resp.status_code == 200:
                             transcription_data = t_resp.json()
                             logger.info(f"Downloaded transcription keys: {transcription_data.keys() if isinstance(transcription_data, dict) else 'Not a dict'}")
+                            # 保存原始数据
+                            task.transcription_data = transcription_data
                         else:
                             logger.warning(f"Failed to download transcription: {t_resp.status_code}")
                             transcription_data = {}
@@ -234,6 +236,8 @@ class TranscriptionTaskViewSet(viewsets.ModelViewSet):
                         if t_resp.status_code == 200:
                             transcription_data = t_resp.json()
                             logger.info(f"Downloaded transcription keys: {transcription_data.keys() if isinstance(transcription_data, dict) else 'Not a dict'}")
+                            # 保存原始数据
+                            task.transcription_data = transcription_data
                      except Exception as e:
                          logger.error(f"Error downloading transcription nested url: {e}")
                 
@@ -300,6 +304,8 @@ class TranscriptionTaskViewSet(viewsets.ModelViewSet):
                         s_resp = requests.get(summarization)
                         if s_resp.status_code == 200:
                             summarization = s_resp.json()
+                            # 保存原始数据
+                            task.summary_data = summarization
                         else:
                             logger.warning(f"Failed to download summarization: {s_resp.status_code}")
                             summarization = {}
@@ -317,6 +323,19 @@ class TranscriptionTaskViewSet(viewsets.ModelViewSet):
                 else:
                     # 尝试从章节摘要中提取
                     chapters = task_result.get('Chapters', [])
+                    # 处理 AutoChapters
+                    auto_chapters = task_result.get('AutoChapters', {})
+                    if isinstance(auto_chapters, str) and auto_chapters.startswith('http'):
+                         try:
+                            import requests
+                            logger.info(f"Downloading auto chapters from {auto_chapters}")
+                            ac_resp = requests.get(auto_chapters)
+                            if ac_resp.status_code == 200:
+                                auto_chapters = ac_resp.json()
+                                task.auto_chapters_data = auto_chapters
+                         except Exception as e:
+                             logger.error(f"Error downloading auto chapters: {e}")
+
                     summary_parts = []
                     for chapter in chapters:
                         if 'Headline' in chapter: