Tilde Search Engine

Search engine for tilde-based websites

Discovery

Responsible for:

• Discovering + temp storing users
• Discovering + temp storing public websites

Crawler

Not creepy at all. Responsible for:

• Downloading and creating per-word document-frequency dictionary for tf-idf
• Storing which websites have been tagged with timestamp and hash of content
• Pulling keywords and tagging websites into general tag dictionary

Content explanantion

• tokenize_corpus and Porter files - are responsible for cleaning corpus data into stemmed tokens. Needs stopwords.txt file in same dir
• data file - interfaces with numerous text and json files for easy data management
• parse_url file - handles html, including requests and parsing text and metadata
• init_freq_dir file - creates and/or updates document frequency dictionary
• crawl file - goes thru urls and gathers tags + metadata for dictionaries

This document last updated: Jul 20 2020