name: taoguba-crawler description: This skill should be used when the user asks to "crawl taoguba", "crawl tgb", "scrape taoguba articles", "run the crawler", "crawl bbs", "crawl home page", "generate article HTML", or needs to run the Taoguba (tgb.cn) web crawlers. version: 0.1.0 allowed-tools: Bash, Read

Taoguba Crawler

This skill runs the Taoguba (tgb.cn) article crawlers located in the project root.

Prerequisites

• Python 3 with requests, beautifulsoup4, python-dotenv installed
• A .env file in the project root containing COOKIE and optionally USER_AGENT

Available Crawlers

1. BBS Crawler (crawler_bbs.py)

Crawl the forum board at tgb.cn/bbs/1/1 using HTML scraping.

python crawler_bbs.py

• Extracts article list by parsing a.overhide.mw300 elements
• Gets each article's main post and author replies
• Downloads images and embeds them as base64 in HTML
• Outputs: output/bbs_YYYY-MM-DD.json and output/bbs_YYYY-MM-DD_HHMMSS.html

2. Home Crawler (crawler_home.py)

Crawl the homepage recommendations via JSON API (/newIndex/getZh).

python crawler_home.py

• Fetches articles from the JSON API (default 2 pages)
• Same content extraction and HTML generation as BBS crawler
• Outputs: output/home_YYYY-MM-DD.json and output/home_YYYY-MM-DD_HHMMSS.html

Common Workflow

To run both crawlers:

python crawler_bbs.py && python crawler_home.py

Key Implementation Details

• Authentication: Both scripts read COOKIE from .env via python-dotenv
• Rate limiting: 0.5-1s delay between requests to avoid being blocked
• Image handling: Images are downloaded and embedded as base64 in the HTML output
• Article content: Extracts main post (#first) and author replies (.comment-data with author badge)
• Output directory: All results saved to output/ folder

Scripts

The crawler scripts are bundled in scripts/:

• scripts/crawler_bbs.py - BBS forum crawler (HTML scraping)
• scripts/crawler_home.py - Homepage crawler (JSON API)

To run the bundled scripts directly:

python scripts/crawler_bbs.py
python scripts/crawler_home.py

Troubleshooting

• If no articles are returned, check that .env contains a valid COOKIE value
• If image downloads fail, the HTML will show error messages inline
• Network timeouts default to 10-15 seconds per request