YTCrawl

---

A lightweight CLI tool to extract YouTube comment data without the overhead. Built for speed, complex pattern recognition.

Features

• Fetch paginated comments from a YouTube video
• Support for standard URLs, short links, and YouTube Shorts
• Extract:
  • author name
  • text of the comment
  • reply_count
  • likes count
  • pfp_url (author profile picture)
• Modular architecture for easier maintenance
• Export comments to JSON

Requirements

• Python 3.x
• requests

Install dependencies:

pip install requests

Usage

Run the script by passing a YouTube URL or Video ID as a positional argument:

python main.py https://www.youtube.com/watch?v=dQw4w9WgXcQ

CLI Arguments

Argument	Description	Default
input	YouTube video URL, Shorts URL, or ID	(Required)
--limit	Number of comment pages to fetch	5
--output	Target JSON filename	comments.json
-v, --verbose	Print author pfp_url in the console	False
--minimal	Remove type key from the saved JSON file	False

How it works

• helpers.py: Logic for extract_video_id and JSON I/O operations.
• scraper.py: Manages requests.Session and YTScraper orchestration.
• parser.py: Uses CommentParser to scrape youtubei/v1/next JSON payloads.
• generator.py: Generates the initial continuation token.

Example Output

Page 1...
[standard] User123: Nice video!
        └── Reply count(s): 5
        └── Like count(s): 42

TODO

• Add threaded replies support
• Add CSV export support
• Improve request handling and performance