dsh-vidi 0.0.0-stage → 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +64 -2
- package/README.zh-CN.md +65 -0
- package/cordis.patch.yml +4 -0
- package/lib/client.js +36967 -0
- package/lib/index.mjs +4542 -0
- package/package.json +62 -4
package/README.md
CHANGED
|
@@ -1,3 +1,65 @@
|
|
|
1
|
-
#
|
|
1
|
+
# dsh-vidi
|
|
2
2
|
|
|
3
|
-
|
|
3
|
+
English | [中文](README.zh-CN.md)
|
|
4
|
+
|
|
5
|
+
A video tool for DeepSeek Harness: watch live streams, videos and podcasts with subtitles in any language, and let DeepSeek take frames, cut clips, download and summarise for you.
|
|
6
|
+
|
|
7
|
+
## What it does
|
|
8
|
+
|
|
9
|
+
- **Opens anything**: YouTube, Bilibili and other sites yt-dlp supports, Xiaoyuzhou podcast episodes, m3u8 / rtmp / srt streams, local files (drop them in), or a stream pushed from OBS.
|
|
10
|
+
- **Subtitles start by themselves** when a video opens; pause and resume any time. One or two languages (a main line and a second line); the site's own subtitles can be shown too, or used as a reference.
|
|
11
|
+
- **Two paces**: real-time shows lines as they are heard; polished arrives a little later with misheard words fixed from context.
|
|
12
|
+
- **The subtitle model looks at the video**: polishing uses the page's title, description and comments, and — when the chosen model reads images — a frame of that stretch. It also keeps a glossary of names and terms for the video.
|
|
13
|
+
- **Chat with DeepSeek beside the video**: grab this moment, cut this part, summarise, download the original — all in one conversation, with files saved in the video's folder.
|
|
14
|
+
- **Export**: SRT / VTT subtitles, transcripts by line or paragraph, clips with subtitle tracks.
|
|
15
|
+
- English and Chinese interface; works on phones.
|
|
16
|
+
|
|
17
|
+
## Install
|
|
18
|
+
|
|
19
|
+
In DSH, open **Plugins → Install third-party plugin** and enter `dsh-vidi`. A **Video** button appears in the sidebar; pasting a video link in a chat also offers to open it.
|
|
20
|
+
|
|
21
|
+
Needs DSH 0.2.2 or later.
|
|
22
|
+
|
|
23
|
+
### Dependencies
|
|
24
|
+
|
|
25
|
+
| Tool | For | Install (macOS) |
|
|
26
|
+
|---|---|---|
|
|
27
|
+
| ffmpeg | playing, recording, frames, clips | `brew install ffmpeg` |
|
|
28
|
+
| yt-dlp | site videos | `brew install yt-dlp` |
|
|
29
|
+
| whisper.cpp (optional) | transcription on this computer | `brew install whisper-cpp`, then put `ggml-large-v3-turbo-q5_0.bin` in `~/.dsh/live/models/` |
|
|
30
|
+
|
|
31
|
+
Without whisper.cpp, pick an online service as the transcription model in settings.
|
|
32
|
+
|
|
33
|
+
## Two models
|
|
34
|
+
|
|
35
|
+
Set separately under the gear:
|
|
36
|
+
|
|
37
|
+
- **Transcription**: speech to text. Whisper on this computer by default; or OpenAI, Google Gemini, Groq, SiliconFlow, Zhipu GLM.
|
|
38
|
+
- **Subtitle model**: polishing, translation, glossary. DeepSeek by default; the list shows every model configured in DSH and marks the ones that read images.
|
|
39
|
+
|
|
40
|
+
Online services use the keys DSH already has (the ones entered on its Models page work as they are); a key entered on the video page is checked, then saved in DSH.
|
|
41
|
+
|
|
42
|
+
| Transcription service | Key names read |
|
|
43
|
+
|---|---|
|
|
44
|
+
| OpenAI | `OPENAI_API_KEY` |
|
|
45
|
+
| Google Gemini | `GEMINI_API_KEY` or `GOOGLE_API_KEY` |
|
|
46
|
+
| Groq | `GROQ_API_KEY` |
|
|
47
|
+
| SiliconFlow | `SILICONFLOW_API_KEY` |
|
|
48
|
+
| Zhipu GLM | `GLM_API_KEY`, `ZHIPUAI_API_KEY` or `ZAI_API_KEY` |
|
|
49
|
+
|
|
50
|
+
## Streaming from OBS
|
|
51
|
+
|
|
52
|
+
Open the video page: the empty page shows the server address and stream key. In OBS choose the "Custom" service and paste them. Default ports: RTMP 1935, SRT 9000 (changeable in the plugin settings).
|
|
53
|
+
|
|
54
|
+
## Where things are saved
|
|
55
|
+
|
|
56
|
+
Recordings, subtitles, frames and clips live in `~/.dsh/live/jobs/<id>/` (changeable in the plugin settings).
|
|
57
|
+
|
|
58
|
+
## Known limits
|
|
59
|
+
|
|
60
|
+
- Xiaoyuzhou video episodes only offer a one-minute preview on the web (the full video needs the app and a login), so they play as the full audio with the cover.
|
|
61
|
+
- Site videos that need a login use a browser's saved login (Settings → Site login).
|
|
62
|
+
|
|
63
|
+
## License
|
|
64
|
+
|
|
65
|
+
MIT
|
package/README.zh-CN.md
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
1
|
+
# dsh-vidi
|
|
2
|
+
|
|
3
|
+
[English](README.md) | 中文
|
|
4
|
+
|
|
5
|
+
DeepSeek Harness 里的视频工具：看直播、视频和播客，同时出任意语言的字幕，再让 DeepSeek 帮你截图、剪片段、下载、总结。
|
|
6
|
+
|
|
7
|
+
## 能做什么
|
|
8
|
+
|
|
9
|
+
- **打开任何来源**：YouTube、B 站等 yt-dlp 支持的网站，小宇宙节目，m3u8 / rtmp / srt 直播流，本地文件（直接拖进来），也可以用 OBS 推流过来。
|
|
10
|
+
- **自动出字幕**：视频一打开就开始，可以随时暂停、继续。单语、双语都行，主字幕和副字幕各选一种语言；网站自带的字幕也能直接显示，或者作为参考。
|
|
11
|
+
- **两种节奏**：实时模式边播边出；精修模式晚一点出，但会结合上下文把听错的词改对。
|
|
12
|
+
- **字幕模型会看画面**：精修时结合视频页面的标题、简介、评论，所选模型能看图时还会带上那一段的画面，顺便整理这期视频的词表（人名、术语）。
|
|
13
|
+
- **在视频旁边和 DeepSeek 对话**：截取此刻、剪下这段、总结要点、下载原片，都在同一个会话里完成，文件存在视频目录下。
|
|
14
|
+
- **导出**：SRT / VTT 字幕，逐句或按段落的文字稿，带字幕轨的片段。
|
|
15
|
+
- 中英文界面，手机上也能用。
|
|
16
|
+
|
|
17
|
+
## 安装
|
|
18
|
+
|
|
19
|
+
在 DSH 里打开 **插件 → 安装第三方插件**，填 `dsh-vidi`。装好后侧栏会出现 **视频** 按钮；在对话里贴视频链接，输入框上方也会出现「在视频里打开」。
|
|
20
|
+
|
|
21
|
+
需要 DSH 0.2.2 或更新的版本。
|
|
22
|
+
|
|
23
|
+
### 依赖
|
|
24
|
+
|
|
25
|
+
| 工具 | 用途 | 安装（macOS） |
|
|
26
|
+
|---|---|---|
|
|
27
|
+
| ffmpeg | 播放、录制、截图、剪辑 | `brew install ffmpeg` |
|
|
28
|
+
| yt-dlp | 打开网站视频 | `brew install yt-dlp` |
|
|
29
|
+
| whisper.cpp（可选） | 在本机转写 | `brew install whisper-cpp`，再下载模型 `ggml-large-v3-turbo-q5_0.bin` 到 `~/.dsh/live/models/` |
|
|
30
|
+
|
|
31
|
+
不装 whisper.cpp 也能用：在设置里把转写模型换成在线服务即可。
|
|
32
|
+
|
|
33
|
+
## 两个模型
|
|
34
|
+
|
|
35
|
+
设置（齿轮）里分开配置：
|
|
36
|
+
|
|
37
|
+
- **转写模型**：把语音转成文字。默认用本机 Whisper；也可以选 OpenAI、Google Gemini、Groq、硅基流动、智谱 GLM。
|
|
38
|
+
- **字幕模型**：精修、翻译、整理词表。默认 DeepSeek，列表里是你在 DSH 里配置的所有模型，能看图的会标出来。
|
|
39
|
+
|
|
40
|
+
在线服务的密钥和 DSH 共用同一份（DSH 的模型设置里填过的直接能用）；在视频页里填的新密钥会先验证，再存进 DSH。
|
|
41
|
+
|
|
42
|
+
| 转写服务 | 读取的密钥名 |
|
|
43
|
+
|---|---|
|
|
44
|
+
| OpenAI | `OPENAI_API_KEY` |
|
|
45
|
+
| Google Gemini | `GEMINI_API_KEY` 或 `GOOGLE_API_KEY` |
|
|
46
|
+
| Groq | `GROQ_API_KEY` |
|
|
47
|
+
| 硅基流动 | `SILICONFLOW_API_KEY` |
|
|
48
|
+
| 智谱 GLM | `GLM_API_KEY`、`ZHIPUAI_API_KEY` 或 `ZAI_API_KEY` |
|
|
49
|
+
|
|
50
|
+
## 用 OBS 推流
|
|
51
|
+
|
|
52
|
+
打开视频页，空白处会显示推流地址和串流密钥。OBS 里服务选「自定义」，填进去就行。默认端口 RTMP 1935、SRT 9000，可在插件设置里改。
|
|
53
|
+
|
|
54
|
+
## 数据放在哪
|
|
55
|
+
|
|
56
|
+
录像、字幕、截图、片段都在 `~/.dsh/live/jobs/<编号>/`，可在插件设置里改目录。
|
|
57
|
+
|
|
58
|
+
## 已知限制
|
|
59
|
+
|
|
60
|
+
- 小宇宙的视频节目在网页上只有 60 秒预览，完整视频要在 App 里登录才能看，所以按完整音频加封面播放。
|
|
61
|
+
- 需要登录才能看的网站视频，会借用浏览器里已登录的状态（设置 → 网站登录）。
|
|
62
|
+
|
|
63
|
+
## 许可
|
|
64
|
+
|
|
65
|
+
MIT
|
package/cordis.patch.yml
ADDED