运行成功，但脚本问题还很大

app/services/llm.py

@@ -451,7 +451,7 @@ def gemini_video2json(video_origin_name: str, video_origin_path: str, video_plot
 """ % (language, video_plot)
 
     logger.debug(f"视频名称: {video_origin_name}")
-    try:
+    # try:
     gemini_video_file = gemini.upload_file(video_origin_path)
     logger.debug(f"上传视频至 Google cloud 成功: {gemini_video_file.name}")
     while gemini_video_file.state.name == "PROCESSING":
@@ -461,9 +461,9 @@ def gemini_video2json(video_origin_name: str, video_origin_path: str, video_plot
         logger.debug(f"视频当前状态(ACTIVE才可用): {gemini_video_file.state.name}")
     if gemini_video_file.state.name == "FAILED":
         raise ValueError(gemini_video_file.state.name)
-    except Exception as err:
+    # except Exception as err:
-        logger.error(f"上传视频至 Google cloud 失败, 请检查 VPN 配置和 APIKey 是否正确 \n{traceback.format_exc()}")
+    #     logger.error(f"上传视频至 Google cloud 失败, 请检查 VPN 配置和 APIKey 是否正确 \n{traceback.format_exc()}")
-        raise TimeoutError(f"上传视频至 Google cloud 失败, 请检查 VPN 配置和 APIKey 是否正确; {err}")
+    #     raise TimeoutError(f"上传视频至 Google cloud 失败, 请检查 VPN 配置和 APIKey 是否正确; {err}")
 
     streams = model.generate_content([prompt, gemini_video_file], stream=True)
     response = []
@@ -490,7 +490,7 @@ if __name__ == "__main__":
     # sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
     #
     video_subject = "卖菜大妈竟是皇嫂"
-    video_path = "/NarratoAI/resource/videos/demoyasuo.mp4"
+    video_path = "../../resource/videos/demoyasuo.mp4"
 
     video_plot = ''' '''
     language = "zh-CN"

app/services/task.py

@@ -440,8 +440,9 @@ def start_subclip(task_id, params: VideoClipParams, subclip_path_videos):
 
         logger.info(f"\n\n## 6. 最后一步: {index} => {final_video_path}")
         # 把所有东西合到在一起
-        video.generate_video(video_path=combined_video_path,
+        video.generate_video_v2(
-                             audio_path=audio_file,
+            video_path=combined_video_path,
+            audio_paths=audio_files,
             subtitle_path=subtitle_path,
             output_file=final_video_path,
             params=params,

app/services/video.py

@@ -1,4 +1,5 @@
 import re
+import os
 import glob
 import random
 from typing import List
@@ -369,10 +370,17 @@ def generate_video_v2(
         return _clip
 
     video_clip = VideoFileClip(video_path)
+    original_audio = video_clip.audio  # 保存原始视频的音轨
+    video_duration = video_clip.duration
 
     # 处理多个音频文件
     audio_clips = []
     for audio_path in audio_paths:
+        # 确保每个音频文件路径是正确的
+        if not os.path.exists(audio_path):
+            logger.warning(f"音频文件不存在: {audio_path}")
+            continue
+
         # 从文件名中提取时间信息
         match = re.search(r'audio_(\d{2}-\d{2}-\d{2}-\d{2})\.mp3', os.path.basename(audio_path))
         if match:
@@ -382,28 +390,53 @@ def generate_video_v2(
             end_time = sum(int(x) * 60 ** i for i, x in enumerate(reversed(end)))
 
             audio_clip = AudioFileClip(audio_path).volumex(params.voice_volume)
-            audio_clip = audio_clip.set_start(start_time).set_end(end_time)
+            
+            # 确保结束时间不超过音频实际长度
+            actual_end_time = min(end_time - start_time, audio_clip.duration)
+            
+            audio_clip = audio_clip.subclip(0, actual_end_time)
+            audio_clip = audio_clip.set_start(start_time).set_end(start_time + actual_end_time)
             audio_clips.append(audio_clip)
         else:
             logger.warning(f"无法从文件名解析时间信息: {audio_path}")
 
-    # 合并所有音频剪辑
+    # 合并所有音频剪辑，包括原始音轨
     if audio_clips:
+        audio_clips.insert(0, original_audio)  # 将原始音轨添加到音频剪辑列表的开头
         audio_clip = CompositeAudioClip(audio_clips)
     else:
-        logger.warning("没有有效的音频文件")
+        logger.warning("没有有效的音频文件，使用原始音轨")
-        audio_clip = AudioClip(lambda t: 0, duration=video_clip.duration)
+        audio_clip = original_audio
 
-    # 字幕处理部分保持不变
+    # 字幕处理部分
     if subtitle_path and os.path.exists(subtitle_path):
         sub = SubtitlesClip(subtitles=subtitle_path, encoding="utf-8")
         text_clips = []
+        
         for item in sub.subtitles:
             clip = create_text_clip(subtitle_item=item)
+            
+            # 确保字幕的开始时间不早于视频开始
+            start_time = max(clip.start, 0)
+            
+            # 如果字幕的开始时间晚于视频结束时间，则跳过此字幕
+            if start_time >= video_duration:
+                continue
+            
+            # 调整字幕的结束时间，但不要超过视频长度
+            end_time = min(clip.end, video_duration)
+            
+            # 调整字幕的时间范围
+            clip = clip.set_start(start_time).set_end(end_time)
+            
             text_clips.append(clip)
+        
+        logger.info(f"处理了 {len(text_clips)} 段字幕")
+        
+        # 创建一个新的视频剪辑，包含所有字幕
         video_clip = CompositeVideoClip([video_clip, *text_clips])
 
-    # 背景音乐处理部分保持不变
+    # 背景音乐处理部分
     bgm_file = get_bgm_file(bgm_type=params.bgm_type, bgm_file=params.bgm_file)
     if bgm_file:
         try:
@@ -573,39 +606,43 @@ def combine_clip_videos(combined_video_path: str,
 
 
 if __name__ == "__main__":
-    combined_video_path = "../../storage/tasks/12312312/com123.mp4"
+    # combined_video_path = "../../storage/tasks/12312312/com123.mp4"
-
+    #
-    video_paths = ['../../storage/cache_videos/vid-00_00-00_03.mp4',
+    # video_paths = ['../../storage/cache_videos/vid-00_00-00_03.mp4',
-                   '../../storage/cache_videos/vid-00_03-00_07.mp4',
+    #                '../../storage/cache_videos/vid-00_03-00_07.mp4',
-                   '../../storage/cache_videos/vid-00_12-00_17.mp4',
+    #                '../../storage/cache_videos/vid-00_12-00_17.mp4',
-                   '../../storage/cache_videos/vid-00_26-00_31.mp4']
+    #                '../../storage/cache_videos/vid-00_26-00_31.mp4']
-    video_ost_list = [False, True, False, True]
+    # video_ost_list = [False, True, False, True]
-    list_script = [
+    # list_script = [
-        {
+    #     {
-            "picture": "夜晚，一个小孩在树林里奔跑，后面有人拿着火把在追赶",
+    #         "picture": "夜晚，一个小孩在树林里奔跑，后面有人拿着火把在追赶",
-            "timestamp": "00:00-00:03",
+    #         "timestamp": "00:00-00:03",
-            "narration": "夜黑风高的树林，一个小孩在拼命奔跑，后面的人穷追不舍！",
+    #         "narration": "夜黑风高的树林，一个小孩在拼命奔跑，后面的人穷追不舍！",
-            "OST": False
+    #         "OST": False,
-        },
+    #         "new_timestamp": "00:00-00:03"
-        {
+    #     },
-            "picture": "追赶的人命令抓住小孩",
+    #     {
-            "timestamp": "00:03-00:07",
+    #         "picture": "追赶的人命令抓住小孩",
-            "narration": "原声播放1",
+    #         "timestamp": "00:03-00:07",
-            "OST": True
+    #         "narration": "原声播放1",
-        },
+    #         "OST": True,
-        {
+    #         "new_timestamp": "00:03-00:07"
-            "picture": "小孩躲在草丛里，黑衣人用脚踢了踢他",
+    #     },
-            "timestamp": "00:12-00:17",
+    #     {
-            "narration": "小孩脱下外套，跑进树林, 一路奔跑，直到第二天清晨",
+    #         "picture": "小孩躲在草丛里，黑衣人用脚踢了踢他",
-            "OST": False
+    #         "timestamp": "00:12-00:17",
-        },
+    #         "narration": "小孩脱下外套，跑进树林, 一路奔跑，直到第二天清晨",
-        {
+    #         "OST": False,
-            "picture": "小孩跑到车前，慌慌张张地对女人说有人要杀他",
+    #         "new_timestamp": "00:07-00:12"
-            "timestamp": "00:26-00:31",
+    #     },
-            "narration": "原声播放2",
+    #     {
-            "OST": True
+    #         "picture": "小孩跑到车前，慌慌张张地对女人说有人要杀他",
-        }
+    #         "timestamp": "00:26-00:31",
-    ]
+    #         "narration": "原声播放2",
+    #         "OST": True,
+    #         "new_timestamp": "00:12-00:17"
+    #     }
+    # ]
     # combine_clip_videos(combined_video_path=combined_video_path, video_paths=video_paths, video_ost_list=video_ost_list, list_script=list_script)
 
     cfg = VideoClipParams()
@@ -633,14 +670,18 @@ if __name__ == "__main__":
     #                params=cfg
     #                )
 
-    video_path = "../../storage/tasks/12312312/com123.mp4"
+    video_path = "../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/combined-1.mp4"
 
-    audio_paths = ['../../storage/tasks/12312312/audio_00-00-00-03.mp3',
+    audio_paths = ['../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/audio_00-00-00-07.mp3',
-                   '../../storage/tasks/12312312/audio_00-12-00-17.mp3']
+                   '../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/audio_00-14-00-17.mp3',
+                   '../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/audio_00-17-00-22.mp3',
+                   '../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/audio_00-34-00-45.mp3',
+                   '../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/audio_00-59-01-09.mp3',
+                   ]
 
-    subtitle_path = "../../storage/tasks/12312312/subtitle_multiple.srt"
+    subtitle_path = "../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa\subtitle.srt"
 
-    output_file = "../../storage/tasks/12312312/out123.mp4"
+    output_file = "../../storage/tasks/7f5ae494-abce-43cf-8f4f-4be43320eafa/final-123.mp4"
 
     generate_video_v2(video_path=video_path,
                        audio_paths=audio_paths,

app/services/voice.py

@@ -1213,7 +1213,7 @@ def create_subtitle_from_multiple(text: str, sub_maker_list: List[SubMaker], lis
             if script_item['OST']:
                 continue
 
-            start_time, end_time = script_item['timestamp'].split('-')
+            start_time, end_time = script_item['new_timestamp'].split('-')
             if sub_maker_index >= len(sub_maker_list):
                 logger.error(f"Sub maker list index out of range: {sub_maker_index}")
                 break
@@ -1317,7 +1317,7 @@ def tts_multiple(task_id: str, list_script: list, voice_name: str, voice_rate: f
 
     for item in list_script:
         if not item['OST']:
-            timestamp = item['timestamp'].replace(':', '-')
+            timestamp = item['new_timestamp'].replace(':', '-')
             audio_file = os.path.join(output_dir, f"audio_{timestamp}.mp3")
             
             # 检查文件是否已存在，如存在且不强制重新生成，则跳过

app/utils/utils.py

@@ -7,7 +7,7 @@ from loguru import logger
 import json
 from uuid import uuid4
 import urllib3
-from datetime import datetime
+from datetime import datetime, timedelta
 
 from app.models import const
 
@@ -326,3 +326,42 @@ def calculate_total_duration(scenes):
         total_seconds += duration.total_seconds()
     
     return total_seconds
+
+
+def add_new_timestamps(scenes):
+    """
+    新增新视频的时间戳，并为"原生播放"的narration添加唯一标识符
+    Args:
+        scenes: 场景列表
+
+    Returns:
+        更新后的场景列表
+    """
+    current_time = timedelta()
+    updated_scenes = []
+
+    for scene in scenes:
+        new_scene = scene.copy()  # 创建场景的副本，以保留原始数据
+        start, end = new_scene['timestamp'].split('-')
+        start_time = datetime.strptime(start, '%M:%S')
+        end_time = datetime.strptime(end, '%M:%S')
+        duration = end_time - start_time
+
+        new_start = current_time
+        current_time += duration
+        new_end = current_time
+
+        # 将 timedelta 转换为分钟和秒
+        new_start_str = f"{int(new_start.total_seconds() // 60):02d}:{int(new_start.total_seconds() % 60):02d}"
+        new_end_str = f"{int(new_end.total_seconds() // 60):02d}:{int(new_end.total_seconds() % 60):02d}"
+
+        new_scene['new_timestamp'] = f"{new_start_str}-{new_end_str}"
+
+        # 为"原生播放"的narration添加唯一标识符
+        if new_scene.get('narration') == "原声播放":
+            unique_id = str(uuid4())[:8]  # 使用UUID的前8个字符作为唯一标识符
+            new_scene['narration'] = f"原声播放_{unique_id}"
+
+        updated_scenes.append(new_scene)
+
+    return updated_scenes

webui/Main.py

@@ -395,7 +395,7 @@ with left_panel:
                     # 去掉json的头尾标识
                     input_json = input_json.strip('```json').strip('```')
                     try:
-                        data = json.loads(input_json)
+                        data = utils.add_new_timestamps(json.loads(input_json))
                     except Exception as err:
                         raise ValueError(
                             f"视频脚本格式错误，请检查脚本是否符合 JSON 格式；{err} \n\n{traceback.format_exc()}")