Refactor and Modernization of esoreader.py

[JaskRendix]

PR improves readability, type safety and usability of the esoreader module while preserving backward compatibility.

Key changes include:

• replaced legacy class definitions with modern Python syntax using type hints (Optional, dict, tuple, list)
• introduced to_series() method in the EsoFile class to allow direct extraction of a single variable as a pandas Series
• updated internal logic to handle malformed lines more gracefully during parsing of the data dictionary and data section
• improved variable filtering in find_variable() to ensure case-insensitive matching and optional key-based filtering
• enhanced to_frame() method with optional column naming based on either key or variable name
• added __repr__() methods to DataDictionary and EsoFile classes for easier debugging and logging
• replaced manual dictionary initialization with dictionary comprehensions for cleaner code
• consolidated imports and removed unused legacy comments and docstrings where appropriate
• maintained the original read() function for compatibility, while encouraging use of the newer read_from_path() interface
• added a test suite using both unittest and pytest, all tests passed before and after the refactor, confirming functional parity. The new to_series() method was excluded from earlier tests but has been validated post-refactor. Note: pytest is included in the separate PR via the pyproject.toml