1

tools/indentify/video_identifier.py

@@ -413,8 +413,40 @@ class VideoIdentifier:
                     if hasattr(response, '_error') and response._error:
                         raise Exception(f"生成错误: {response._error}")
 
+                    # 安全获取文本：避免在无 Part 时访问 response.text 抛错
+                    def safe_extract_text(resp: Any) -> str:
+                        try:
+                            # 优先从 candidates 结构中提取
+                            candidates = getattr(resp, 'candidates', None)
+                            if candidates and len(candidates) > 0:
+                                first = candidates[0]
+                                # 记录 finish_reason 以便错误信息更清晰
+                                finish_reason = getattr(first, 'finish_reason', None)
+                                # parts 路径
+                                content = getattr(first, 'content', None)
+                                parts = getattr(content, 'parts', None) if content else None
+                                if parts and len(parts) > 0:
+                                    # 兼容 text 或直接包含的内容
+                                    # 常见 part 为 {text: str}
+                                    part0 = parts[0]
+                                    text = getattr(part0, 'text', None) if hasattr(part0, 'text') else part0.get('text') if isinstance(part0, dict) else None
+                                    if isinstance(text, str) and text.strip():
+                                        return text
+                                # 若无 parts 或为空，根据 finish_reason 返回清晰错误
+                                reason = str(finish_reason) if finish_reason is not None else 'unknown'
+                                raise ValueError(f"无有效输出内容，finish_reason={reason}")
+
+                            # 退化到 resp.text（可能抛错）
+                            if hasattr(resp, 'text') and isinstance(resp.text, str) and resp.text.strip():
+                                return resp.text
+
+                            raise ValueError("响应中没有可用的文本内容")
+                        except Exception as ex:
+                            raise ex
+
                     try:
-                        parsed = json.loads(response.text.strip())
+                        text_payload = safe_extract_text(response).strip()
+                        parsed = json.loads(text_payload)
                         if not isinstance(parsed, dict):
                             raise ValueError("响应格式错误：非字典结构")
                         
@@ -449,6 +481,18 @@ class VideoIdentifier:
                                 'ocr_content': '关键帧分析失败：JSON解析错误'
                             }]
                         }
+                    except Exception as e:
+                        # 捕获无 Part 或 finish_reason 封锁等导致的无法提取文本问题
+                        err_msg = str(e)
+                        return {
+                            'url': url, 'duration': duration,
+                            'asr_content': f'处理失败: {err_msg}',
+                            'iframe_details': [{
+                                'time_start': 0, 'time_end': 0,
+                                'content': f'处理失败: {err_msg}',
+                                'ocr_content': f'处理失败: {err_msg}'
+                            }]
+                        }
                         
                 except Exception as e:
                     return {