test_skill

Web crawler using BFS and anti-scraping to extract and save structured BBC and general news content in Markdown with multi-site and dedup support.

BBC Crawler MaxClaw

Description

A powerful, universal web crawler optimized for BBC News but capable of crawling other sites. It integrates advanced scraping technologies including Crawl4AI and Playwright to handle dynamic content and anti-bot protections.

Features

• Multi-Method Extraction:
  • crawl4ai: Primary method using AsyncWebCrawler for high performance and accuracy.
  • playwright: Full browser rendering fallback for complex dynamic pages.
  • requests: Fast fallback for static content.
  • auto: Automatically detects the best method (Prioritizes Crawl4AI).
• Hierarchical Storage: Saves content in a structured format: YYYY-MM-DD/Category/Title.md.
• Local Image Archiving: Downloads images locally, names them by MD5 hash, and updates Markdown references.
• Content Filtering: Intelligently extracts main article content and relevant images using CSS selectors.

Requirements

• Python 3.9+
• See requirements.txt for Python packages.

Installation

# 1. Install dependencies
# Note: install.py supports passing arguments to pip, e.g., --break-system-packages
python install.py

# Example for environments requiring --break-system-packages:
python install.py --break-system-packages

Usage

Basic Usage

python universal_crawler_v2.py --url https://www.bbc.co.uk/news --max-pages 50

Advanced Usage

# Force Crawl4AI
python universal_crawler_v2.py --url https://www.bbc.co.uk/news --method crawl4ai

# Force Playwright
python universal_crawler_v2.py --url https://www.bbc.co.uk/news --method playwright

# Control depth and delay
python universal_crawler_v2.py --url https://www.bbc.co.uk/news --depth 3 --delay 2.5

# Specify output directory
python universal_crawler_v2.py --url https://www.bbc.co.uk/news --output ./my_data

Troubleshooting

• Import Errors: If you see "No module named 'crawl4ai'" or similar, run python install.py again.
• Empty Responses: Ensure you have the latest version of the crawler. Some sites may block specific IPs or user agents; try increasing delay or switching methods.