Added unit tests

[AkshatSharma25]

Closes #2
The description of the test cases added is as follows:

File 1: test_utils.py

Before data goes anywhere near an AI, it needs a cleanup. These 13 tests make sure our functions aren’t slacking.

Text Cleaning

We don’t want wasted tokens or messy formatting confusing the model.

• Extra Newlines – Collapses 4+ line breaks into a neat two. No endless scrolling.
• Space Crusher – Turns Hello World into Hello World.
• Tab-to-Space – Replaces \t with normal spaces so searches behave.
• The Trimmer – Cuts sneaky leading/trailing whitespace from copy-paste jobs.
• “Don’t Panic” Test – Handles empty strings without crashing.

Filename Regex

Makes sure we correctly split CSV rows into filename and content.

• Standard Case – Parses document.txt,Content cleanly.
• Special Characters – Supports dashes and underscores in filenames.
• The Bouncer – Rejects malformed lines instead of breaking the system.

Hash Functions

Prevents us from paying to process the same text twice.

• Consistency – Same input, same hash. Every time.
• SHA256 Check – Verifies the hash is the correct 64 characters.
• Case Insensitivity – Treats “Hello” and “hello” as identical to avoid duplicates.

---

File 2: test_basic.py

These 9 tests make sure our data structures and settings won’t fall apart when we hit “Go.”

Data Structures

• Doc Format – Every document must have text and file. No exceptions.
• Metadata Integrity – Chunks must know their source and index.
• JSON Round-trip – Convert to JSON and back without losing anything.

Config Guards

• Overlap Rule – chunk_overlap must be smaller than chunk_size.
• API Safety – temperature stays between 0–2, and max_tokens can’t be negative.

Prompt Templates

• Placeholder Check – Ensures {context} and {question} still exist.
• Formatting Test – Runs a mock prompt to confirm it renders exactly right.

---

Clean data. Solid structure. Safe configs. No surprises.