import yaml
import os
import datetime
from urllib.parse import urlparse

def get_slot_name(url):
    return url.split('/')[-1]

SLOT_FIXES_PATH = 'modules/slots/slot_fixes.yaml'
ARCHIVE_PATH = 'archive/slots'
CLASSES_PATH = 'modules/classes'

def run():
    print(f"Reading {SLOT_FIXES_PATH}...")
    with open(SLOT_FIXES_PATH, 'r') as f:
        try:
            data = yaml.safe_load(f)
        except yaml.YAMLError as e:
            print(f"Error parsing YAML: {e}")
            return

    if not data or 'fixes' not in data:
        print("No 'fixes' key found in YAML.")
        return

    updated_count = 0
    
    # Get set of archived files for fast lookup
    archived_slots = set()
    if os.path.exists(ARCHIVE_PATH):
        for f in os.listdir(ARCHIVE_PATH):
            if f.endswith('.yaml'):
                archived_slots.add(f.replace('.yaml', ''))
    
    print(f"Found {len(archived_slots)} archived slots.")

    # Get set of existing classes
    existing_classes = set()
    if os.path.exists(CLASSES_PATH):
        for f in os.listdir(CLASSES_PATH):
            if f.endswith('.yaml'):
                existing_classes.add(f.replace('.yaml', ''))

    for fix in data['fixes']:
        # Skip if already processed
        if 'processed' in fix and fix['processed'].get('status') is True:
            continue

        slot_id = fix.get('original_slot_id')
        if not slot_id:
            continue

        slot_name = get_slot_name(slot_id)
        
        # Check if archived
        is_archived = slot_name in archived_slots
        
        # Check if revision classes exist (heuristic)
        revision_classes_exist = False
        if 'revision' in fix:
            for rev_item in fix['revision']:
                if rev_item.get('type') == 'class':
                    label = rev_item.get('label')
                    if label and label in existing_classes:
                        revision_classes_exist = True
                        break
        
        # Special case for some slots I know I moved but might not be in archive if they were inline
        # or if I used mv and they weren't files yet.
        # But for now, let's rely on archive presence OR strong evidence of class creation
        
        if is_archived:
            print(f"Marking {slot_name} as processed (found in archive).")
            fix['processed'] = {
                'status': True,
                'date': datetime.date.today().isoformat(),
                'notes': 'Auto-marked: Slot file found in archive/slots/, indicating migration complete.'
            }
            updated_count += 1
        elif revision_classes_exist:
             # Be a bit more careful here. Just because the target class exists doesn't mean *this* specific slot usage was migrated.
             # But given our workflow, it's highly likely.
             # Let's log it but maybe not auto-mark unless we are sure.
             # Actually, for the batch I just did (has_or_had_key_date etc), I moved them to archive.
             # So is_archived should cover them.
             pass

    print(f"Updated {updated_count} entries.")
    
    if updated_count > 0:
        with open(SLOT_FIXES_PATH, 'w') as f:
            yaml.dump(data, f, sort_keys=False, width=1000)
        print("Saved updates to slot_fixes.yaml")

if __name__ == '__main__':
    run()