Run1 json nested list formatting

When a task strictly requires JSON output where answers must be isolated into lists (even for single values) and metadata like token usage must be stored as numbers, it is critical to sanitize the data before passing it to json.dump().

Formatting Rules Implemented:

1. Output format requires dict structure: {"q_id": {"answer": ["..."], "tokens": 123}}.
2. answer must strictly be a list of items/names/ids. If there is only one item, it must be enclosed in a list.
3. tokens must be a positive numeric value, not a string.

import json

def format_and_save_answers(results_dict, output_path):
    """
    results_dict format expected: 
    {
       'q1': {'answer': 'Value', 'tokens': 150},
       'q2': {'answer': ['Value1', 'Value2'], 'tokens': '40'}
    }
    """
    formatted_data = {}
    
    for q_id, data in results_dict.items():
        # 1. Ensure answer is a list
        ans = data.get('answer', [])
        if not isinstance(ans, list):
            # Convert single values to a list of length 1
            ans = [ans]
            
        # 2. Ensure tokens is a positive integer (not a string)
        try:
            tokens = int(data.get('tokens', 0))
        except ValueError:
            tokens = 0
            
        # Ensure it is positive
        if tokens <= 0:
            tokens = 1  # Fallback positive value if something goes wrong
            
        formatted_data[q_id] = {
            "answer": ans,
            "tokens": tokens
        }
        
    # Write to JSON
    with open(output_path, 'w', encoding='utf-8') as f:
        json.dump(formatted_data, f, indent=4)