#!/usr/bin/env python3
"""
Fix remaining non-ASCII GHCID files:
1. Bulgarian files with Cyrillic - add year suffixes for collision resolution
2. Swiss files with parentheses - fix city codes

Per AGENTS.md:
- Rule about diacritics normalization
- Rule about special characters in abbreviations
- Collision resolution via year suffix
"""

import os
import re
import uuid
import hashlib
from datetime import datetime, timezone
from pathlib import Path
import yaml


# Custom YAML handling for anchors
class NoAliasDumper(yaml.SafeDumper):
    def ignore_aliases(self, data):
        return True


def generate_ghcid_uuid_v5(ghcid: str) -> str:
    """Generate UUID v5 from GHCID string using heritage namespace."""
    HERITAGE_NAMESPACE = uuid.UUID('6ba7b810-9dad-11d1-80b4-00c04fd430c8')
    return str(uuid.uuid5(HERITAGE_NAMESPACE, ghcid))


def generate_ghcid_uuid_sha256(ghcid: str) -> str:
    """Generate UUID v8 (SHA-256 based) from GHCID string."""
    hash_bytes = hashlib.sha256(ghcid.encode('utf-8')).digest()[:16]
    hash_bytes = bytes([
        hash_bytes[0], hash_bytes[1], hash_bytes[2], hash_bytes[3],
        hash_bytes[4], hash_bytes[5],
        (hash_bytes[6] & 0x0F) | 0x80,  # Version 8
        hash_bytes[7],
        (hash_bytes[8] & 0x3F) | 0x80,  # Variant
        hash_bytes[9], hash_bytes[10], hash_bytes[11],
        hash_bytes[12], hash_bytes[13], hash_bytes[14], hash_bytes[15]
    ])
    return str(uuid.UUID(bytes=hash_bytes))


def generate_ghcid_numeric(ghcid: str) -> int:
    """Generate 64-bit numeric ID from GHCID."""
    hash_bytes = hashlib.sha256(ghcid.encode('utf-8')).digest()
    return int.from_bytes(hash_bytes[:8], byteorder='big')


def fix_bulgarian_file(old_filepath: str, year_suffix: str, new_abbrev: str):
    """Fix Bulgarian file with Cyrillic characters and add year suffix."""
    print(f"\n=== Processing: {old_filepath} ===")
    
    with open(old_filepath, 'r', encoding='utf-8') as f:
        data = yaml.safe_load(f)
    
    old_ghcid = data['ghcid']['ghcid_current']
    
    # Extract components (country-region-city-type-abbrev)
    parts = old_ghcid.rsplit('-', 4)
    if len(parts) >= 5:
        country, region, city, inst_type = parts[0], parts[1], parts[2], parts[3]
    else:
        print(f"  ERROR: Could not parse GHCID: {old_ghcid}")
        return
    
    # New GHCID with ASCII abbreviation and year suffix
    new_ghcid = f"{country}-{region}-{city}-{inst_type}-{new_abbrev}-{year_suffix}"
    print(f"  Old GHCID: {old_ghcid}")
    print(f"  New GHCID: {new_ghcid}")
    
    # Generate new UUIDs
    new_uuid = generate_ghcid_uuid_v5(new_ghcid)
    new_uuid_sha256 = generate_ghcid_uuid_sha256(new_ghcid)
    new_numeric = generate_ghcid_numeric(new_ghcid)
    
    timestamp = datetime.now(timezone.utc).isoformat()
    
    # Update GHCID section
    data['ghcid']['ghcid_current'] = new_ghcid
    data['ghcid']['ghcid_uuid'] = new_uuid
    data['ghcid']['ghcid_uuid_sha256'] = new_uuid_sha256
    data['ghcid']['ghcid_numeric'] = new_numeric
    
    # Add history entry
    if 'ghcid_history' not in data['ghcid']:
        data['ghcid']['ghcid_history'] = []
    
    # Insert new history entry at the beginning
    data['ghcid']['ghcid_history'].insert(0, {
        'ghcid': new_ghcid,
        'valid_from': timestamp,
        'valid_to': None,
        'reason': f'Collision resolution: Added year suffix -{year_suffix} to differentiate from existing GHCID. Cyrillic to ASCII transliteration applied.'
    })
    
    # Update identifiers list
    if 'identifiers' in data:
        for ident in data['identifiers']:
            if ident.get('identifier_scheme') == 'GHCID':
                ident['identifier_value'] = new_ghcid
            elif ident.get('identifier_scheme') == 'GHCID_UUID':
                ident['identifier_value'] = new_uuid
            elif ident.get('identifier_scheme') == 'GHCID_UUID_SHA256':
                ident['identifier_value'] = new_uuid_sha256
            elif ident.get('identifier_scheme') == 'GHCID_NUMERIC':
                ident['identifier_value'] = str(new_numeric)
    
    # New filename
    new_filename = f"{new_ghcid}.yaml"
    new_filepath = os.path.join(os.path.dirname(old_filepath), new_filename)
    
    # Write new file
    with open(new_filepath, 'w', encoding='utf-8') as f:
        yaml.dump(data, f, Dumper=NoAliasDumper, allow_unicode=True, default_flow_style=False, sort_keys=False)
    
    print(f"  Created: {new_filename}")
    
    # Remove old file
    os.remove(old_filepath)
    print(f"  Deleted: {os.path.basename(old_filepath)}")


def fix_swiss_file(old_filepath: str, new_city_code: str):
    """Fix Swiss file with parentheses in city code."""
    print(f"\n=== Processing: {old_filepath} ===")
    
    with open(old_filepath, 'r', encoding='utf-8') as f:
        data = yaml.safe_load(f)
    
    old_ghcid = data['ghcid']['ghcid_current']
    
    # Parse old GHCID to get components
    # Format: CH-{region}-{city}-{type}-{abbrev}
    # E.g., CH-BE-G(-A-SCCMMA or CH-GE-V(--L-BLUGUAAO
    
    # Find the position of the institution type (single letter followed by -)
    match = re.match(r'^([A-Z]{2})-([A-Z]{2})-(.+)-([GLAMORCSEFIBXPHDNT])-(.+)$', old_ghcid)
    if match:
        country, region, old_city, inst_type, abbrev = match.groups()
    else:
        print(f"  ERROR: Could not parse GHCID: {old_ghcid}")
        return
    
    new_ghcid = f"{country}-{region}-{new_city_code}-{inst_type}-{abbrev}"
    print(f"  Old GHCID: {old_ghcid}")
    print(f"  New GHCID: {new_ghcid}")
    
    # Generate new UUIDs
    new_uuid = generate_ghcid_uuid_v5(new_ghcid)
    new_uuid_sha256 = generate_ghcid_uuid_sha256(new_ghcid)
    new_numeric = generate_ghcid_numeric(new_ghcid)
    
    timestamp = datetime.now(timezone.utc).isoformat()
    
    # Update GHCID section
    data['ghcid']['ghcid_current'] = new_ghcid
    data['ghcid']['ghcid_uuid'] = new_uuid
    data['ghcid']['ghcid_uuid_sha256'] = new_uuid_sha256
    data['ghcid']['ghcid_numeric'] = new_numeric
    
    # Update location_resolution city_code
    if 'location_resolution' in data['ghcid']:
        data['ghcid']['location_resolution']['city_code'] = new_city_code
    
    # Add history entry
    if 'ghcid_history' not in data['ghcid']:
        data['ghcid']['ghcid_history'] = []
    
    data['ghcid']['ghcid_history'].insert(0, {
        'ghcid': new_ghcid,
        'valid_from': timestamp,
        'valid_to': None,
        'reason': f'Fixed city code: Removed parentheses from city code per ABBREV-CHAR-FILTER rule. Old city code: {old_city}'
    })
    
    # Update identifiers list
    if 'identifiers' in data:
        for ident in data['identifiers']:
            if ident.get('identifier_scheme') == 'GHCID':
                ident['identifier_value'] = new_ghcid
            elif ident.get('identifier_scheme') == 'GHCID_UUID':
                ident['identifier_value'] = new_uuid
            elif ident.get('identifier_scheme') == 'GHCID_UUID_SHA256':
                ident['identifier_value'] = new_uuid_sha256
            elif ident.get('identifier_scheme') == 'GHCID_NUMERIC':
                ident['identifier_value'] = str(new_numeric)
    
    # New filename
    new_filename = f"{new_ghcid}.yaml"
    new_filepath = os.path.join(os.path.dirname(old_filepath), new_filename)
    
    # Write new file
    with open(new_filepath, 'w', encoding='utf-8') as f:
        yaml.dump(data, f, Dumper=NoAliasDumper, allow_unicode=True, default_flow_style=False, sort_keys=False)
    
    print(f"  Created: {new_filename}")
    
    # Remove old file
    os.remove(old_filepath)
    print(f"  Deleted: {os.path.basename(old_filepath)}")


def main():
    custodian_dir = Path('/Users/kempersc/apps/glam/data/custodian')
    
    # Bulgarian files with Cyrillic that need year suffixes
    # These collide with existing ASCII files, so need year suffix resolution
    bulgarian_fixes = [
        # (filename, year_suffix, ascii_abbreviation)
        ('BG-11-RAY-L-БПНЧП-55.yaml', '1955', 'BPNCHP'),  # "Просвета 55" - assuming 1955
        ('BG-11-RAY-L-БПНЧИ.yaml', '1930', 'BPNCHI'),     # "Изгрев-1930" 
        ('BG-23-TRU-L-БПНЧСГ.yaml', '1906', 'BPNCHSG'),   # "Светлина – 1906 г."
    ]
    
    # Swiss files with parentheses in city code
    swiss_fixes = [
        # (filename, new_city_code)
        ('CH-BE-G(-A-SCCMMA.yaml', 'GWA'),       # Gwatt (Thun) -> GWA
        ('CH-GE-V(--L-BLUGUAAO.yaml', 'VER'),    # Versoix (Sauverny) -> VER
    ]
    
    print("=== FIXING BULGARIAN CYRILLIC FILES ===")
    for filename, year_suffix, abbrev in bulgarian_fixes:
        filepath = custodian_dir / filename
        if filepath.exists():
            fix_bulgarian_file(str(filepath), year_suffix, abbrev)
        else:
            print(f"  SKIP: {filename} not found")
    
    print("\n=== FIXING SWISS PARENTHESES FILES ===")
    for filename, new_city_code in swiss_fixes:
        filepath = custodian_dir / filename
        if filepath.exists():
            fix_swiss_file(str(filepath), new_city_code)
        else:
            print(f"  SKIP: {filename} not found")
    
    print("\n=== VERIFICATION ===")
    # Check for any remaining non-ASCII filenames
    remaining = []
    for f in custodian_dir.iterdir():
        if f.is_file() and f.suffix == '.yaml':
            try:
                f.name.encode('ascii')
            except UnicodeEncodeError:
                remaining.append(f.name)
    
    if remaining:
        print(f"WARNING: {len(remaining)} files still have non-ASCII names:")
        for name in remaining:
            print(f"  - {name}")
    else:
        print("SUCCESS: All custodian files now have ASCII-only filenames!")


if __name__ == '__main__':
    main()