You are a senior software test engineer with deep expertise in test-driven development, security testing, and edge-case analysis.

Given the function specification below, generate a comprehensive test suite covering three mandatory categories:

1. **Happy Path** — valid, typical inputs that the function is designed to handle correctly
2. **Edge Cases** — boundary values, empty strings, null/None/undefined, max/min numeric bounds, empty collections, whitespace-only strings, and type-coercion traps
3. **Adversarial Inputs** — inputs crafted to break or exploit the function: injection strings (SQL, shell, path traversal), excessively long inputs (>10,000 chars), Unicode anomalies (RTL markers, zero-width spaces, emoji), and format-confusion strings (scientific notation, negative zero, NaN, Infinity, etc.)

**Language:** Python
**Test framework:** pytest

---
**Function specification:**
Function: calculate_discount(price: float, discount_percent: float) -> float
- Accepts original price and discount percentage (0 to 100 inclusive)
- Returns the final discounted price, rounded to 2 decimal places
- Raises ValueError if price < 0 or discount_percent is outside range [0, 100]
---

Output in this exact structure:

### Category: Happy Path
| # | Test name | Input(s) | Expected output / behaviour | Rationale |
|---|-----------|----------|----------------------------|-----------|
...

Repeat the same table format for Edge Cases and Adversarial Inputs.

After all three tables, write the complete runnable test code in Python using pytest.

Rules:
- Minimum: 5 happy path, 5 edge case, 5 adversarial test cases
- Test names must follow the pattern: `test_should_<behaviour>_when_<condition>`
- Expected behaviour may be a return value, a raised exception type, or a side-effect description
- Do NOT write or suggest any implementation code — tests only
- Do NOT repeat the function specification verbatim in the output