min-endow-scraper

min-endow-scraper is a Python tool for extracting, filtering, and summarizing endowment-related quotes and statistics from institutional websites. It automates the process of crawling, parsing, and aggregating relevant financial quotes, producing clean markdown outputs for further analysis.

Features

• Automated Web Scraping: Extracts candidate URLs from sitemaps and homepages.
• Quote Extraction: Identifies and scores financial quotes using customizable keyword and pattern matching.
• Output Aggregation: Produces per-entity markdown summaries and a combined report with statistics.
• Configurable Inputs: Easily specify target institutions and all important parameters of the system.
• No JSON Output: All results are in human-readable markdown format.

Usage

1. Install Requirements

   • Python 3+ is required.
   • Install dependencies:

     pip install -r requests
     pip install -r beautifulsoup4
     
     

2. Prepare Inputs

   • Place target URLs in URLs.txt (one per line).
   • names.txt contains personal names from a public-domain library (one per line).
   • parameters.txt contains sets of string texts and numeric parameters. The user can use the existing parameters or arbitrarily modify them.

3. Run the Scraper

   python main.py
   

4. View Results

   • Per-entity results: outputs/run_datetimestamp/<entity>/ (7 markdown files per entity)
   • Combined summary: AllBestQuotes.md

Output Structure

• outputs/<entity>/BestQuotes.md — Top 5 quotes for each entity
• outputs/<entity>/CandidateURLs.md, RedFlagURLs.md, VisitedURLs.md, CandidateQuotes.md, RejectedQuotes.md, OpenQuotes.md — Supporting data
• AllBestQuotes.md — Combined summary with statistics and all best quotes

Customization

• Extend or modify output formatting in output.py.

License

This project is released under the MIT License.

Disclaimer

This tool is intended for research and educational purposes.