SearchEngine

Mini Search Engine with Stack

A command-line search engine over local text files, built as a 2nd semester DSA lab project (BS AI, NFC IET Multan). It uses an inverted index (hash map) for keyword search, two stacks for browser-style undo/redo of queries, and falls back to OpenAI GPT-3.5-turbo when no local document matches.

How it works

• On startup, every .txt file in documents/ is indexed into an inverted index mapping word -> {document: frequency} (lowercased, punctuation stripped).
• A query matches documents containing all of its words; results are ranked by the summed frequency of the query words.
• Each query is pushed onto a history stack. back pops it to return to the previous query (undo), next redoes an undone one, and show prints the history stack. A new search clears the redo stack.
• If nothing matches locally, the query is sent to GPT-3.5-turbo. The answer is saved as gpt_docs/gpt_N.txt, added to the index, and shown as a result.
• After each search you can type a result number to print the file's contents, or continue to keep searching. quit exits.

Files

• main.py - entry point
• search.py - search loop, undo/redo, GPT fallback
• index.py - inverted index (add, clean, search, rank)
• stack.py - list-based stack (push, pop, peek, show)
• documents/ - sample text files to index
• gpt_docs/ - saved GPT answers (gpt_1.txt, gpt_2.txt, ...)

Running

Requires Python 3.12+ and an OpenAI API key.

cd SearchEngine
pip install -r requirements.txt
echo "OPENAI_API_KEY=sk-..." > .env
python main.py

The key is required even for local-only searches, since it is loaded at startup.

Credits

Course: Data Structures & Algorithms, 2nd Semester, BS Artificial Intelligence, NFC IET Multan.

Team: Muawiya Amir (team leader), M. Umar.