import os
import re

def fix_dangling_in_file(path):
    with open(path, 'r') as f:
        lines = f.readlines()
    
    new_lines = []
    
    # keys that take lists
    list_keys = [
        "slots", 
        "exact_mappings", "close_mappings", "broad_mappings", "related_mappings", "narrow_mappings",
        "examples", "comments", "see_also", "keywords", "structured_aliases",
        "subsets", "mixins", "apply_to", "union_of", "values", "equals_expression", "equals_string_in",
        "aliases", "local_names", "union_of", "defines"
    ]
    
    last_key_at_4 = None
    
    for i, line in enumerate(lines):
        stripped = line.strip()
        
        if not stripped or stripped.startswith('#'):
            new_lines.append(line)
            continue
            
        indent = len(line) - len(line.lstrip())
        
        # Check if line is a key at 4 spaces
        # Regex: start with 4 spaces, then key chars, then colon, then optional space/value
        if indent == 4 and ':' in stripped and not stripped.startswith('-'):
            key = stripped.split(':')[0].strip()
            last_key_at_4 = key
            new_lines.append(line)
            continue
            
        # Check if line is a list item at 4 spaces
        if indent == 4 and stripped.startswith('-'):
            if last_key_at_4 not in list_keys:
                print(f"Removing dangling list item in {path} at line {i+1} (after {last_key_at_4}): {stripped}")
                continue # Skip/Delete
        
        # Reset last_key_at_4 if indentation drops below 4
        if indent < 4:
            last_key_at_4 = None
            
        new_lines.append(line)
        
    with open(path, 'w') as f:
        f.writelines(new_lines)

def process_directory(directory):
    for root, dirs, files in os.walk(directory):
        for file in files:
            if file.endswith(".yaml"):
                fix_dangling_in_file(os.path.join(root, file))

process_directory("schemas/20251121/linkml/modules/classes")