dataset/problem_22/buggy.py

import re

def parse_and_validate_emails(email_list: list[str], domains_allowed: set) -> dict:
    """
    Parse a list of emails, validate them against allowed domains, and group them.
    
    email_list: list of strings (e.g., ["user@example.com", "admin@test.org"])
    domains_allowed: set of allowed domains (e.g., {"example.com", "test.org"})
    
    Returns a dict:
    - 'valid_emails': list of valid emails
    - 'invalid_emails': list of invalid emails (bad format or not in allowed domains)
    - 'domain_counts': dict mapping domain to count of valid emails
    """
    
    # BUG 1: regex is flawed, doesn't properly escape dot, allows invalid chars before @
    email_regex = re.compile(r"^[a-zA-Z0-9_]+@[a-zA-Z0-9]+\.[a-zA-Z]+$")
    
    result = {
        'valid_emails': [],
        'invalid_emails': [],
        # BUG 2: using list instead of dict for domain_counts
        'domain_counts': []
    }
    
    for email in email_list:
        # BUG 3: not trimming whitespace from email before validation
        if not email_regex.match(email):
            result['invalid_emails'].append(email)
            continue
            
        # BUG 4: split could fail or result in index error if @ is missing (though regex should catch it, regex is flawed)
        # BUG 5: split from the wrong direction if multiple @ exist
        parts = email.split('@')
        domain = parts[1]
        
        if domain not in domains_allowed:
            result['invalid_emails'].append(email)
        else:
            result['valid_emails'].append(email)
            
            # BUG 6: incorrect operation for list (trying to use it as dict)
            # This will raise TypeError since domain_counts is a list
            if domain in result['domain_counts']:
                result['domain_counts'][domain] += 1
            else:
                result['domain_counts'][domain] = 1
                
    return result