feat(frontend): 新增多版本笔记功能，并做了向下兼容。

- 新增关于页面组件，介绍项目背景、功能和使用方法
- 重构笔记生成逻辑，支持多版本笔记
- 新增笔记版本选择、复制和导出功能
-优化笔记界面布局和交互
- 调整部分组件样式，提升用户体验

backend/app/downloaders/bilibili_downloader.py

@@ -7,6 +7,7 @@ import yt_dlp
 from app.downloaders.base import Downloader, DownloadQuality, QUALITY_MAP
 from app.models.notes_model import AudioDownloadResult
 from app.utils.path_helper import get_data_dir
+from app.utils.url_parser import extract_video_id
 
 
 class BilibiliDownloader(Downloader, ABC):
@@ -69,10 +70,19 @@ class BilibiliDownloader(Downloader, ABC):
         """
         下载视频，返回视频文件路径
         """
+
         if output_dir is None:
             output_dir = get_data_dir()
-
         os.makedirs(output_dir, exist_ok=True)
+        print("video_url",video_url)
+        video_id=extract_video_id(video_url, "bilibili")
+        video_path = os.path.join(output_dir, f"{video_id}.mp4")
+        if os.path.exists(video_path):
+            return video_path
+
+        # 检查是否已经存在
+
+
         output_path = os.path.join(output_dir, "%(id)s.%(ext)s")
 
         ydl_opts = {

backend/app/downloaders/douyin_downloader.py

@@ -249,13 +249,21 @@ class DouyinDownloader(Downloader):
         )
 
     def download_video(self, video_url: str, output_dir: Union[str, None] = None) -> str:
+
         try:
+
             if output_dir is None:
                 output_dir = get_data_dir()
             if not output_dir:
                 output_dir = self.cache_data
             os.makedirs(output_dir, exist_ok=True)
 
+            video_id = self.extract_video_id(video_url)
+            video_path = os.path.join(output_dir, f"{video_id}.mp4")
+            if os.path.exists(video_path):
+                return video_path
+
+
             output_path = os.path.join(output_dir, "%(id)s.%(ext)s")
 
             video_data = self.fetch_video_info(video_url)

backend/app/downloaders/youtube_downloader.py

@@ -7,6 +7,7 @@ import yt_dlp
 from app.downloaders.base import Downloader, DownloadQuality
 from app.models.notes_model import AudioDownloadResult
 from app.utils.path_helper import get_data_dir
+from app.utils.url_parser import extract_video_id
 
 
 class YoutubeDownloader(Downloader, ABC):
@@ -67,12 +68,15 @@ class YoutubeDownloader(Downloader, ABC):
         """
         if output_dir is None:
             output_dir = get_data_dir()
-
+        video_id = extract_video_id(video_url, "youtube")
+        video_path = os.path.join(output_dir, f"{video_id}.mp4")
+        if os.path.exists(video_path):
+            return video_path
         os.makedirs(output_dir, exist_ok=True)
         output_path = os.path.join(output_dir, "%(id)s.%(ext)s")
 
         ydl_opts = {
-            'format': 'worst[ext=mp4]/worst',
+            'format': 'bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]',
             'outtmpl': output_path,
             'noplaylist': True,
             'quiet': False,

backend/app/gpt/prompt.py

@@ -35,11 +35,10 @@ BASE_PROMPT = '''
 根据上面的分段转录内容，生成结构化的笔记，遵循以下原则：
 
 1. **完整信息**：记录尽可能多的相关细节，确保内容全面。
-2. **清晰结构**：用合适的标题级别（`##`，`###`）整理内容，概述每个部分的要点。主标题用`#`来标识(如果额外重要的任务有格式需求可以不遵守)
-3. **去除无关内容**：省略广告、填充词、问候语和不相关的言论。
-4. **保留关键细节**：保留重要事实、示例、结论和建议。(如果额外重要的任务有格式需求可以不遵守)
-5. **可读布局**：必要时使用项目符号，并保持段落简短，增强可读性。(如果额外重要的任务有格式需求可以不遵守)
-6. 视频中提及的数学公式必须保留，并以 LaTeX 语法形式呈现，适合 Markdown 渲染。
+2. **去除无关内容**：省略广告、填充词、问候语和不相关的言论。
+3. **保留关键细节**：保留重要事实、示例、结论和建议。(如果额外重要的任务有格式需求可以不遵守)
+4. **可读布局**：必要时使用项目符号，并保持段落简短，增强可读性。(如果额外重要的任务有格式需求可以不遵守)
+5. 视频中提及的数学公式必须保留，并以 LaTeX 语法形式呈现，适合 Markdown 渲染。
 
 
 请始终遵循此规则。

backend/app/gpt/prompt_builder.py

@@ -58,7 +58,7 @@ def get_format_function(format_type):
 def get_style_format(style):
     style_map = {
         'minimal': '1. **精简信息**: 仅记录最重要的内容，简洁明了。',
-        'detailed': '2. **详细记录**: 包含完整的时间戳和每个部分的详细讨论。',
+        'detailed': '2. **详细记录**: 包含完整的内容和每个部分的详细讨论。需要尽可能多的记录视频内容，最好详细的笔记',
         'academic': '3. **学术风格**: 适合学术报告，正式且结构化。',
         'xiaohongshu': '''4. **小红书风格**: 
 ### 擅长使用下面的爆款关键词：

backend/app/routers/note.py

@@ -217,17 +217,25 @@ def get_task_status(task_id: str):
 @router.get("/image_proxy")
 async def image_proxy(request: Request, url: str):
     headers = {
-        "Referer": "https://www.bilibili.com/",  # 模拟B站来源
+        "Referer": "https://www.bilibili.com/",
         "User-Agent": request.headers.get("User-Agent", ""),
     }
 
     try:
         async with httpx.AsyncClient(timeout=10.0) as client:
             resp = await client.get(url, headers=headers)
+
             if resp.status_code != 200:
                 raise HTTPException(status_code=resp.status_code, detail="图片获取失败")
 
             content_type = resp.headers.get("Content-Type", "image/jpeg")
-            return StreamingResponse(resp.aiter_bytes(), media_type=content_type)
+            return StreamingResponse(
+                resp.aiter_bytes(),
+                media_type=content_type,
+                headers={
+                    "Cache-Control": "public, max-age=86400",  # ✅ 缓存一天
+                    "Content-Type": content_type,
+                }
+            )
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))

backend/app/routers/provider.py

@@ -47,26 +47,26 @@ def add_provider(data: ProviderRequest):
 @router.get("/get_all_providers")
 def get_all_providers():
     try:
-        res = ProviderService.get_all_providers()
+        res = ProviderService.get_all_providers_safe()
         return R.success(data=res)
     except Exception as e:
         return R.error(msg=e)
 
-@router.get("/get_provider_by_id/{id}")
-def get_provider_by_id(id: str):
-    try:
-        res = ProviderService.get_provider_by_id(id)
-        return R.success(data=res)
-    except Exception as e:
-        return R.error(msg=e)
-
-@router.get("/get_provider_by_name/{name}")
-def get_provider_by_name(name: str):
-    try:
-        res = ProviderService.get_provider_by_name(name)
-        return R.success(data=res)
-    except Exception as e:
-        return R.error(msg=e)
+# @router.get("/get_provider_by_id/{id}")
+# def get_provider_by_id(id: str):
+#     try:
+#         res = ProviderService.get_provider_by_id(id)
+#         return R.success(data=res)
+#     except Exception as e:
+#         return R.error(msg=e)
+#
+# @router.get("/get_provider_by_name/{name}")
+# def get_provider_by_name(name: str):
+#     try:
+#         res = ProviderService.get_provider_by_name(name)
+#         return R.success(data=res)
+#     except Exception as e:
+#         return R.error(msg=e)
 
 
 @router.post("/update_provider")

backend/app/services/model.py

@@ -45,7 +45,16 @@ class ModelService:
         except Exception as e:
             print(f"获取所有模型失败: {e}")
             return []
-
+    @staticmethod
+    def get_all_models_safe(verbose: bool = False):
+        try:
+            raw_models = get_all_models()
+            if verbose:
+                print(f"所有模型列表: {raw_models}")
+            return ModelService._format_models(raw_models)
+        except Exception as e:
+            print(f"获取所有模型失败: {e}")
+            return []
     @staticmethod
     def _format_models(raw_models: list) -> list:
         """

backend/app/services/note.py

@@ -140,7 +140,6 @@ class NoteGenerator:
                 replacement = f"![]({image_url})"
                 new_markdown = new_markdown.replace(marker, replacement, 1)
 
-
             return new_markdown
         except Exception as e:
             logger.error(f"截图生成失败：{e}")
@@ -201,16 +200,23 @@ class NoteGenerator:
             # -------- 1. 下载音频 --------
             try:
                 self.update_task_status(task_id, TaskStatus.DOWNLOADING)
+
+                # 加载音频缓存（如果存在）
+                audio = None
                 if os.path.exists(audio_cache_path):
                     logger.info(f"检测到已有音频缓存，直接读取，task_id={task_id}")
                     with open(audio_cache_path, "r", encoding="utf-8") as f:
                         audio_data = json.load(f)
                     audio = AudioDownloadResult(**audio_data)
-                else:
-                    if 'screenshot' in _format or video_understanding:
+
+                # 需要视频的情况（截图 or 视频理解）
+                need_video = 'screenshot' in _format or video_understanding
+                if need_video:
+                    try:
                         video_path = downloader.download_video(video_url)
                         self.video_path = video_path
                         logger.info(f"成功下载视频文件: {video_path}")
+
                         video_img_urls = VideoReader(
                             video_path=video_path,
                             grid_size=tuple(grid_size),
@@ -219,13 +225,17 @@ class NoteGenerator:
                             unit_height=720,
                             save_quality=90,
                         ).run()
+                    except Exception as e:
+                        logger.error(f"❌ 下载视频失败，task_id={task_id}，错误信息：{e}")
+                        self.update_task_status(task_id, TaskStatus.FAILED, message=f"下载音频失败：{e}")
 
-                    screenshot = 'screenshot' in _format
-                    audio: AudioDownloadResult = downloader.download(
+                # 没有音频缓存就下载音频（可能同时也带上视频）
+                if audio is None:
+                    audio = downloader.download(
                         video_url=video_url,
                         quality=quality,
                         output_dir=path,
-                        need_video=screenshot
+                        need_video='screenshot' in _format,  # 注意这里只为了截图需要
                     )
                     with open(audio_cache_path, "w", encoding="utf-8") as f:
                         json.dump(asdict(audio), f, ensure_ascii=False, indent=2)
@@ -266,27 +276,27 @@ class NoteGenerator:
             # -------- 3. 总结内容 --------
             try:
                 self.update_task_status(task_id, TaskStatus.SUMMARIZING)
-                if os.path.exists(markdown_cache_path):
-                    logger.info(f"检测到已有总结缓存，直接读取，task_id={task_id}")
-                    with open(markdown_cache_path, "r", encoding="utf-8") as f:
-                        markdown = f.read()
-                else:
-                    source = GPTSource(
-                        title=audio.title,
-                        segment=transcript.segments,
-                        tags=audio.raw_info.get('tags'),
-                        screenshot=screenshot,
-                        video_img_urls=video_img_urls,
-                        link=link,
-                        _format=_format,
-                        style=style,
-                        extras=extras
-                    )
+                # if os.path.exists(markdown_cache_path):
+                #     logger.info(f"检测到已有总结缓存，直接读取，task_id={task_id}")
+                #     with open(markdown_cache_path, "r", encoding="utf-8") as f:
+                #         markdown = f.read()
+                # else:
+                source = GPTSource(
+                    title=audio.title,
+                    segment=transcript.segments,
+                    tags=audio.raw_info.get('tags'),
+                    screenshot=screenshot,
+                    video_img_urls=video_img_urls,
+                    link=link,
+                    _format=_format,
+                    style=style,
+                    extras=extras
+                )
 
-                    markdown: str = gpt.summarize(source)
-                    with open(markdown_cache_path, "w", encoding="utf-8") as f:
-                        f.write(markdown)
-                    logger.info(f"GPT总结并缓存成功，task_id={task_id}")
+                markdown: str = gpt.summarize(source)
+                with open(markdown_cache_path, "w", encoding="utf-8") as f:
+                    f.write(markdown)
+                logger.info(f"GPT总结并缓存成功，task_id={task_id}")
             except Exception as e:
                 logger.error(f"❌ 总结内容失败，task_id={task_id}，错误信息：{e}")
                 self.update_task_status(task_id, TaskStatus.FAILED, message=f"总结内容失败：{e}")
@@ -313,10 +323,11 @@ class NoteGenerator:
             # -------- 6. 完成 --------
             self.update_task_status(task_id, TaskStatus.SUCCESS)
             logger.info(f"✅ 笔记生成成功，task_id={task_id}")
-            if platform != 'local':
-                transcription_finished.send({
-                    "file_path": audio.file_path,
-                })
+            # TODO :改为前端一键清除缓存
+            # if platform != 'local':
+            #     transcription_finished.send({
+            #         "file_path": audio.file_path,
+            #     })
             return NoteResult(
                 markdown=markdown,
                 transcript=transcript,

backend/app/services/provider.py

@@ -14,6 +14,7 @@ from app.models.model_config import ModelConfig
 
 
 class ProviderService:
+
     @staticmethod
     def serialize_provider(row: tuple) -> dict:
         if not row:
@@ -28,7 +29,25 @@ class ProviderService:
             "enabled": row[6],
             "created_at": row[7],
         }
-
+    @staticmethod
+    def serialize_provider_safe(row: tuple) -> dict:
+        if not row:
+            return None
+        return {
+            "id": row[0],
+            "name": row[1],
+            "logo": row[2],
+            "type": row[3],
+            "api_key": ProviderService.mask_key(row[4]),
+            "base_url": row[5],
+            "enabled": row[6],
+            "created_at": row[7],
+        }
+    @staticmethod
+    def mask_key(key: str) -> str:
+        if not key or len(key) < 8:
+            return '*' * len(key)
+        return key[:4] + '*' * (len(key) - 8) + key[-4:]
     @staticmethod
     def add_provider( name: str, api_key: str, base_url: str, logo: str, type_: str, enabled: int = 1):
         try:
@@ -42,7 +61,10 @@ class ProviderService:
     def get_all_providers():
         rows = get_all_providers()
         return [ProviderService.serialize_provider(row) for row in rows] if rows else []
-
+    @staticmethod
+    def get_all_providers_safe():
+        rows = get_all_providers()
+        return [ProviderService.serialize_provider(row) for row in rows] if rows else []
     @staticmethod
     def get_provider_by_name(name: str):
         row = get_provider_by_name(name)