Initial commit
This commit is contained in:
Zhongwei Li
105
skills/plan-video/SKILL.md
Normal file
105
skills/plan-video/SKILL.md
Normal file
@@ -0,0 +1,105 @@
|
||||
---
|
||||
name: plan-video
|
||||
description: 视频处理任务规划工具。从用户输入中提取视频 URLs，生成唯一 VideoId，创建结构化的 todolist.md 追踪待生成文件。支持小红书、抖音、TikTok、B站、YouTube、快手等 11+ 平台。
|
||||
---
|
||||
|
||||
# Plan Video - 视频任务规划
|
||||
|
||||
## 概述
|
||||
|
||||
为批量视频处理创建结构化任务清单。提取 URLs → 生成 VideoIds → 创建 todolist.md 追踪文件。
|
||||
|
||||
## 使用时机
|
||||
|
||||
- 用户提供视频 URLs 需要批量处理
|
||||
- 在下载/处理前需要规划任务结构
|
||||
|
||||
## 工作流程
|
||||
|
||||
### 1. 提取视频 URLs
|
||||
|
||||
从用户输入（消息/文件）中提取所有视频链接。
|
||||
|
||||
```python
|
||||
from scripts.extract_video_id import extract_urls
|
||||
urls = extract_urls(user_input_text)
|
||||
```
|
||||
|
||||
**支持平台**：小红书、抖音、TikTok、B站、YouTube、快手等
|
||||
|
||||
### 2. 提取 VideoId
|
||||
|
||||
为每个 URL 提取唯一标识符。
|
||||
|
||||
```python
|
||||
from scripts.extract_video_id import extract_video_id
|
||||
video_id = extract_video_id(url)
|
||||
# 'http://xhslink.com/o/6VbNVltFQRX' → '6VbNVltFQRX'
|
||||
```
|
||||
|
||||
### 3. 生成 todolist.md
|
||||
|
||||
在输出目录创建任务清单，列出所有待生成文件。
|
||||
|
||||
**格式**：
|
||||
```markdown
|
||||
# Video Processing Tasks
|
||||
|
||||
## {VideoId} ({原始短链接URL})
|
||||
|
||||
- [ ] orgin/{VideoId}/{VideoId}.json # 元数据，通过 skill `parse-video` 获得
|
||||
- [ ] orgin/{VideoId}/{VideoId}.mp4 # 视频，通过 skill `parse-video` 获得
|
||||
- [ ] orgin/{VideoId}/{VideoId}_cover.jpg # 封面，通过 skill `rip-video` 获得
|
||||
- [ ] orgin/{VideoId}/{VideoId}.mp3 # 音频，通过 skill `rip-video` 获得
|
||||
- [ ] orgin/{VideoId}/{VideoId}.srt # 字幕，通过 skill `rip-video` 获得
|
||||
```
|
||||
|
||||
**示例**：
|
||||
```markdown
|
||||
## 6VbNVltFQRX (http://xhslink.com/o/6VbNVltFQRX)
|
||||
|
||||
- [ ] orgin/6VbNVltFQRX/6VbNVltFQRX.json
|
||||
- [ ] orgin/6VbNVltFQRX/6VbNVltFQRX.mp4
|
||||
- [ ] orgin/6VbNVltFQRX/6VbNVltFQRX_cover.jpg
|
||||
- [ ] orgin/6VbNVltFQRX/6VbNVltFQRX.mp3
|
||||
- [ ] orgin/6VbNVltFQRX/6VbNVltFQRX.srt
|
||||
```
|
||||
|
||||
### 4. 输出报告
|
||||
|
||||
```markdown
|
||||
============================================================
|
||||
视频处理计划已创建！
|
||||
============================================================
|
||||
找到视频: {total} 个
|
||||
输出目录: {outputDir}/orgin/
|
||||
任务列表: {outputDir}/todolist.md
|
||||
|
||||
可以继续使用 parse-video 和 rip-video 进行下载和处理。
|
||||
============================================================
|
||||
```
|
||||
|
||||
## 错误处理
|
||||
|
||||
- **未找到 URLs**：提示用户未检测到有效视频链接
|
||||
- **VideoId 提取失败**：跳过该 URL，在报告中说明
|
||||
- **输出目录不存在**：自动创建目录结构
|
||||
- **todolist.md 已存在**：询问用户是否覆盖或追加
|
||||
|
||||
## 资源文件
|
||||
|
||||
### scripts/extract_video_id.py
|
||||
|
||||
VideoId 提取工具。
|
||||
|
||||
**函数**：
|
||||
- `extract_video_id(url: str) -> str`：提取单个 URL 的 VideoId
|
||||
- `extract_urls(text: str) -> list[str]`：从文本提取所有视频 URLs
|
||||
|
||||
## 集成说明
|
||||
|
||||
此 skill 为后续处理准备基础：
|
||||
1. **parse-video**：使用 todolist.md 中的 VideoIds 和 URLs 解析元数据、下载视频
|
||||
2. **rip-video**：使用 todolist.md 中的 MP4 路径提取音频和字幕
|
||||
|
||||
todolist.md 是整个视频处理流程的中心追踪文档。
|
||||
140
skills/plan-video/scripts/extract_video_id.py
Normal file
140
skills/plan-video/scripts/extract_video_id.py
Normal file
@@ -0,0 +1,140 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
Extract VideoId from various video platform URLs.
|
||||
|
||||
Supports: 小红书, 抖音, TikTok, B站, YouTube, 快手等
|
||||
"""
|
||||
|
||||
import re
|
||||
from urllib.parse import urlparse, parse_qs
|
||||
|
||||
|
||||
def extract_video_id(url: str) -> str:
|
||||
"""
|
||||
Extract VideoId from video platform URL.
|
||||
|
||||
Args:
|
||||
url: Video URL from supported platforms
|
||||
|
||||
Returns:
|
||||
VideoId string
|
||||
|
||||
Raises:
|
||||
ValueError: If URL format is not recognized
|
||||
"""
|
||||
url = url.strip()
|
||||
|
||||
# 小红书短链接: http://xhslink.com/o/6VbNVltFQRX
|
||||
if 'xhslink.com' in url or 'xiaohongshu.com' in url:
|
||||
match = re.search(r'/o/([A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
|
||||
# 抖音: https://v.douyin.com/xxx/ or https://www.douyin.com/video/xxx
|
||||
if 'douyin.com' in url:
|
||||
# 短链接
|
||||
match = re.search(r'v\.douyin\.com/([A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
# 长链接
|
||||
match = re.search(r'/video/(\d+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
|
||||
# TikTok: https://www.tiktok.com/@user/video/1234567890
|
||||
if 'tiktok.com' in url:
|
||||
match = re.search(r'/video/(\d+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
# 短链接 https://vm.tiktok.com/xxx/
|
||||
match = re.search(r'vm\.tiktok\.com/([A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
|
||||
# B站: https://www.bilibili.com/video/BVxxx or https://b23.tv/xxx
|
||||
if 'bilibili.com' in url or 'b23.tv' in url:
|
||||
# BV号
|
||||
match = re.search(r'/(BV[A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
# av号
|
||||
match = re.search(r'/av(\d+)', url)
|
||||
if match:
|
||||
return f"av{match.group(1)}"
|
||||
# 短链接
|
||||
match = re.search(r'b23\.tv/([A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
|
||||
# YouTube: https://www.youtube.com/watch?v=xxx or https://youtu.be/xxx
|
||||
if 'youtube.com' in url or 'youtu.be' in url:
|
||||
# youtu.be 短链接
|
||||
match = re.search(r'youtu\.be/([A-Za-z0-9_-]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
# 标准链接
|
||||
parsed = urlparse(url)
|
||||
if parsed.query:
|
||||
params = parse_qs(parsed.query)
|
||||
if 'v' in params:
|
||||
return params['v'][0]
|
||||
|
||||
# 快手: https://www.kuaishou.com/short-video/xxx
|
||||
if 'kuaishou.com' in url:
|
||||
match = re.search(r'/short-video/(\d+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
match = re.search(r'\.com/([A-Za-z0-9]+)', url)
|
||||
if match:
|
||||
return match.group(1)
|
||||
|
||||
# 如果都不匹配，尝试提取 URL 最后的路径部分
|
||||
parsed = urlparse(url)
|
||||
path_parts = [p for p in parsed.path.split('/') if p]
|
||||
if path_parts:
|
||||
# 返回最后一个非空路径部分
|
||||
return path_parts[-1]
|
||||
|
||||
raise ValueError(f"无法从 URL 提取 VideoId: {url}")
|
||||
|
||||
|
||||
def extract_urls(text: str) -> list[str]:
|
||||
"""
|
||||
Extract video URLs from text.
|
||||
|
||||
Args:
|
||||
text: Text containing video URLs
|
||||
|
||||
Returns:
|
||||
List of unique video URLs
|
||||
"""
|
||||
pattern = r'https?://[^\s]+(?:douyin|tiktok|youtube|youtu\.be|xiaohongshu|xhslink|bilibili|b23\.tv|kuaishou)[^\s]*'
|
||||
urls = re.findall(pattern, text)
|
||||
return list(dict.fromkeys(urls)) # 保持顺序去重
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
# 测试用例
|
||||
test_urls = [
|
||||
'http://xhslink.com/o/6VbNVltFQRX',
|
||||
'http://xhslink.com/o/16X9vM9CTBo',
|
||||
'https://v.douyin.com/eFyQRjc/',
|
||||
'https://www.douyin.com/video/7123456789',
|
||||
'https://www.tiktok.com/@user/video/7123456789',
|
||||
'https://vm.tiktok.com/ZMeAbCdEf/',
|
||||
'https://www.bilibili.com/video/BV1xx411c7mD',
|
||||
'https://b23.tv/av12345678',
|
||||
'https://www.youtube.com/watch?v=dQw4w9WgXcQ',
|
||||
'https://youtu.be/dQw4w9WgXcQ',
|
||||
'https://www.kuaishou.com/short-video/1234567890',
|
||||
]
|
||||
|
||||
print("VideoId 提取测试:\n")
|
||||
for url in test_urls:
|
||||
try:
|
||||
video_id = extract_video_id(url)
|
||||
print(f"✅ {url}")
|
||||
print(f" → {video_id}\n")
|
||||
except ValueError as e:
|
||||
print(f"❌ {url}")
|
||||
print(f" → {e}\n")
|
||||
Reference in New Issue
Block a user