This repository contains my Tiny Search Engine, a project implemented in C that contains three unique subsystems:
crawler, which takes a root URL and recursively crawls through the linked webpagesindexer, which takes a crawler-produced directory and indexes every wordquerier, which, using the directories and indexes, takes an inputted query and returns the most relevant webpages, ranked according to relevance.
All specific functionality and limitations of this design are well-documented in README, IMPLEMENTATION, and DESIGN markdown files.
To compile, run make. To clean the subdirectories and the object files, run make clean.