youtube_explode_dart 是一个 Dart 库,它复制了 YouTube Explode 项目的功能,使开发者能够在无需 YouTube API 密钥的情况下直接与 YouTube 内容交互。它通过解析 YouTube 的公开数据,支持获取视频详情、播放列表、搜索结果等。该库专为跨平台使用而设计,可在 Android、iOS、Web、macOS、Windows 和 Linux 上无缝运行。该包非常适合需要实时访问 YouTube 数据且设置简单的应用程序。
适合场景
获取视频元数据
搜索 YouTube 视频
提取播放列表内容
构建 YouTube 客户端
创建具有动态内容的视频播放器
主要特点
无需 API 密钥
跨平台支持
实时数据解析
轻便且快速
支持视频、播放列表和搜索操作
适合对象
基于YouTube的应用程序
媒体发现工具
教育平台
内容聚合服务
离线视频探索者
使用建议
依赖公开的 YouTube HTML 解析
可能因 YouTube 界面更改而失效
不适合高频率请求
有限的官方支持
由于直接抓取,可能存在速率限制
原始 README
以下为关联源码仓库的 README 原文快照,最新内容请访问源码仓库。
YoutubeExplodeDart
This is a port of the YoutubeExplode library from C#, most of the functions, doc comments, readme information, is taken from YoutubeExplode repository.
It used to build Youtube Downloader Flutter (A cross-platform application to download video streams from youtube using this library & flutter)
YoutubeExplode is a library that provides an interface to query metadata of YouTube videos, playlists and channels, as well as to resolve and download video streams and closed caption tracks. Behind a layer of abstraction, the library parses raw page content and uses reverse-engineered AJAX requests to retrieve information. As it doesn't use the official API, there's also no need for an API key and there are no usage quotas.
Features
Retrieve metadata on videos, playlists, channels, streams, and closed captions
The following example shows how you can extract various metadata from a YouTube video:
// You can provide either a video ID or URL as String or an instance of `VideoId`.var video = yt.videos.get('https://youtube.com/watch?v=Dpp1sIL1m5Q'); // Returns a Video instance.var title = video.title; // "Scamazon Prime"var author = video.author; // "Jim Browning"var duration = video.duration; // Instance of Duration - 0:19:48.00000
Downloading a video stream
Every YouTube video has a number of streams available. These streams may have different containers, video quality, bitrate, etc.
On top of that, depending on the content of the stream, the streams are further divided into 3 categories:
Muxed streams -- contain both video and audio (available only in 360p)
Audio-only streams -- contain only audio
Video-only streams -- contain only video
You can request the stream manifest to get available streams for a particular video:
var yt = YoutubeExplode();
var manifest = yt.videos.streams.getManifest('Dpp1sIL1m5Q');
print(manifest); // Prints the list of streams available for the video.// If you want to specify which youtube clients to use to get the manifest you can do so:var manifest = yt.videos.streams.getManifest(videoId, ytClients: [
YoutubeApiClient.safari,
YoutubeApiClient.androidVr
]); // The streams provided by both clients will be merged.
Once you get the manifest, you can filter through the streams and choose the one you're interested in downloading:
// highest bitrate audio-only streamvar streamInfo = streamManifest.audioOnly.withHigestBitrate();
// MP4 video-only streamvar streamInfo.videoOnly.where((e) => e.container == Container);
// Muxed (audio + video) stream with the highest video quality. NOTE: Don't rely on this, muxed streams are limited to 360p30.var streamInfo = streamManifest.muxed.withHigestVideoQuality();
// HLS (m3u8) streams are also supportedvar streamInfo = streamManifest.hls;
Finally, you can get the actual Stream object represented by the metadata:
// Get the actual byte streamvar stream = yt.video.streams.get(streamInfo);
// Open a file for writing.var file = File(filePath);
var fileStream = file.openWrite();
// Pipe all the content of the stream into the file.await stream.pipe(fileStream);
// Close the file.await fileStream.flush();
await fileStream.close();
While it may be tempting to just always use muxed streams, it's important to note that they are limited in quality. Muxed streams don't go beyond 360p30.
If you want to download the video in maximum quality, you need to download the audio-only and video-only streams separately and then mux them together on your own. There are tools like FFmpeg that let you do that.
Using a signature solver
Some youtube clients require you to complete a js challenge in order to download a video. This requires having a full js runtime like deno, nodejs or quickjs.
Currently the only implemented runtime is deno, see the following snippet for how to use it:
import'package:youtube_explode_dart/solvers.dart';
final solver = awaitDenoEJSSolver.init();
var yt = YoutubeExplode(jsSolver: solver);
Working with playlists
Among other things, YoutubeExplode also supports playlists:
var yt = YoutubeExplode();
// Get playlist metadata.var playlist = await yt.playlists.get('xxxxx');
var title = playlist.title;
var author = playlist.author;
awaitfor (var video in yt.playlists.getVideos(playlist.id)) {
var videoTitle = video.title;
var videoAuthor = video.author;
}
var playlistVideos = await yt.playlists.getVideos(playlist.id);
// Get first 20 playlist videos.var somePlaylistVideos = await yt.playlists.getVideos(playlist.id).take(20);
Extracting closed captions
Similarly, to streams, you can extract closed captions by getting the manifest and choosing the track you're interested in:
var yt = YoutubeExplode();
var trackManifest = await yt.videos.closedCaptions.getManifest('_QdPW8JrYzQ')
var trackInfo = manifest.getByLanguage('en'); // Get english caption.if (trackInfo != null)
{
// Get the actual closed caption track.var track = await youtube.videos.closedCaptions.get(trackInfo);
// Get the caption displayed at 1:01var caption = track.getByTime(Duration(seconds: 61));
var text = caption?.text; // "And the game was afoot."
}
Get a list of related videos
var video = yt.videos.get('https://youtube.com/watch?v=Dpp1sIL1m5Q');
var relatedVideos = await yt.videos.getRelatedVideos(video); // video must be a Video instance.print(relatedVideos); //prints the list of related videos// to get the next page of related videos
relatedVideos = await relatedVideos.nextPage();
If no related video is found getRelatedVideos or nextPage will return null.
Get comments
You can easily get the video comments of a given video, the return value of comments.getComments(video) is a list-like object which behaves exactly like a List but has an additional method nextPage() which is used in order to get the next comments, it returns null when there are no comments to be fetched anymore.
var comments = await yt.videos.comments.getComments(video);
var replies = await yt.videos.comments.getReplies(comment); // Fetch the comment replies
Cleanup
You need to close YoutubeExplode's http client, when done otherwise this could halt the dart process.
yt.close();
Troubleshooting
If you encounter any issues, please check that it has not been reported already in the issues section.
When reporting a new issue, make sure to follow the issue template and report the logs after having enabled the logging:
import'package:logging/logging.dart';
...
// Before any YoutubeExplode codeLogger.root.level = Level.FINER;
Logger.root.onRecord.listen((e) {
print(e);
if (e.error != null) {
print(e.error);
print(e.stackTrace);
}
});
To the very minimum the video id causing the issue (if applicable) and the code snippet that is causing the issue must be provided.