crawl-youtube-channel 0.0.1a2

A package to crawl a youtube channel

Crawl YouTube Channel

This Python package provides tools to crawl and extract data from YouTube channels.

Features

• Crawl an entire YouTube channel for video information.
• Extract metadata, comments, transcripts, audio, and video for each video.
• Provides a base class to easily implement your own video processing and storage logic.
• Includes a Sqlite3YouTubeVideoProcessor for storing data in a local SQLite database.
• Provides data classes for easy access to crawled data.

Prerequisites

• Python 3.10+
• Google Cloud YouTube API Key

Installation

1. Install the package:

   pip install crawl-youtube-channel
   

2. Set up your environment:

   Create a .env file in your project root and add your Google Cloud YouTube API key:

   GOOGLE_CLOUD_YOUTUBE_API_KEY=your_api_key
   

Usage

To use the crawler, you need to implement the YouTubeVideoProcessorBase abstract class. This class defines how to check for existing videos and how to process new ones.

Here is a basic skeleton for a custom processor:

import asyncio
from crawl_youtube_channel import YouTubeVideoProcessorBase, YouTubeVideo

class MyVideoProcessor(YouTubeVideoProcessorBase):
    async def check_video(self, video_id: str) -> bool:
        # Implement logic to check if the video has already been processed.
        # Return True if it exists, False otherwise.
        ...

    async def process_video(self, v: YouTubeVideo) -> None:
        # Implement logic to save or process the video data.
        # For example, save it to a database, a file, or another service.
        ...

async def main():
    # Initialize your custom processor
    processor = MyVideoProcessor()

    # Start crawling the channel
    await processor.process_channel(channel_url='https://www.youtube.com/@YourFavoriteChannel/videos')

if __name__ == '__main__':
    asyncio.run(main())

For a concrete implementation example, see the Sqlite3YouTubeVideoProcessor class in the source code, which stores video data in a SQLite database.

Data Models

The following data classes are used to structure the crawled data:

• YouTubeVideo: The main container for all video-related data.
• YouTubeThumbnail: Basic information about a video thumbnail.
• YouTubeData: Contains detailed information about a video, including:
  • Meta: Video metadata (title, description, tags, etc.).
  • Comment: A YouTube comment, including replies.
  • Transcript: The video's transcript.
  • audio: The audio file in M4A format (as bytes).
  • video: The video file in MP4 format (as bytes).

License

This project is licensed under the MIT License - see the LICENSE file for details.