URL Shortener Backend System (C++)

Overview

This project implements a backend URL shortening service in C++ that converts long URLs into short, unique identifiers and supports fast redirection back to the original URLs.

The system is designed for constant-time encode and decode operations, simulating the core logic used in real-world URL shorteners like Bitly.

---

Problem Statement

Design a backend system that generates short, unique URLs for given long URLs and efficiently resolves them back to the original URLs.

The system should:

• Ensure uniqueness
• Handle expiration of URLs (TTL)
• Support fast lookup operations

---

Features

• Encode long URLs into short Base62-encoded identifiers
• Decode short URLs back to original URLs in O(1) time
• Collision-safe unique ID generation
• Optional TTL (Time-To-Live) support for URL expiration
• Automatic cleanup of expired URLs during access

---

System Design & Approach

The system uses an incremental unique ID generator combined with Base62 encoding to produce compact, URL-safe short codes.

This approach is:

• Simple
• Collision-free
• Highly efficient for in-memory systems

---

Core Data Structures

Component	Purpose
unordered_map<long long, string>	Map ID → original URL
unordered_map<string, long long>	Map short URL → ID
unordered_map<long long, time_t>	Track expiration timestamps

Why this design?

• Constant-time encode and decode
• Minimal memory overhead
• Clean and scalable backend logic

---

Core Operations

encode(longUrl, ttl)

• Generates a unique numeric ID
• Converts ID to a Base62 short URL
• Stores bidirectional mappings
• Optionally assigns an expiration time

decode(shortUrl)

• Resolves short URL to original URL
• Validates expiration
• Returns error if URL is missing or expired

---

Time & Space Complexity

• Encode: O(1)
• Decode: O(1)
• Space: O(N) where N is the number of stored URLs

---

Example Usage

string shortUrl = encode("https://example.com", 5);
decode(shortUrl); // returns original URL

Limitations & Future Improvements

• In-memory only (no persistent storage)
• Not thread-safe
• No distributed support

---

Learning Outcomes

• Understanding backend ID generation strategies
• Practical use of hashing and encoding techniques
• Designing constant-time backend systems
• Evaluating scalability and system trade-offs

---