#!/usr/bin/env python3
"""
Naming Convention Validator for Firefrost Operations Manual

Validates file and directory naming follows FFG standards:
- lowercase-with-hyphens for files and directories
- UPPERCASE for root-level critical docs
- Proper date formats (YYYY-MM-DD)
- No spaces, underscores (except photos), or special chars

Usage:
    python3 validate-naming.py [--fix] [--verbose]
"""

import os
import re
import sys
from pathlib import Path
from typing import List, Tuple

class NamingValidator:
    def __init__(self, repo_root: str, fix: bool = False, verbose: bool = False):
        self.repo_root = Path(repo_root)
        self.fix = fix
        self.verbose = verbose
        self.violations: List[Tuple[str, str]] = []
        
        # Naming rules
        self.valid_root_uppercase = [
            'README.md',
            'DOCUMENT-INDEX.md',
            'SESSION-HANDOFF-PROTOCOL.md',
            'DERP.md',
            'LICENSE',
            '.gitignore'
        ]
        
    def is_valid_lowercase_hyphen(self, name: str) -> bool:
        """Check if name follows lowercase-with-hyphens convention"""
        # Allow: lowercase letters, numbers, hyphens, dots (for extensions)
        pattern = r'^[a-z0-9\-\.]+$'
        return bool(re.match(pattern, name))
    
    def is_valid_date_format(self, name: str) -> bool:
        """Check if name contains properly formatted dates"""
        # Look for YYYY-MM-DD patterns
        date_pattern = r'\d{4}-\d{2}-\d{2}'
        dates = re.findall(date_pattern, name)
        for date in dates:
            year, month, day = date.split('-')
            # Basic validation
            if not (1900 <= int(year) <= 2100):
                return False
            if not (1 <= int(month) <= 12):
                return False
            if not (1 <= int(day) <= 31):
                return False
        return True
    
    def check_path(self, path: Path):
        """Check a single file or directory path"""
        relative_path = path.relative_to(self.repo_root)
        parts = relative_path.parts
        
        # Check each part of the path
        for i, part in enumerate(parts):
            is_root_level = (i == 0)
            is_file = (i == len(parts) - 1 and path.is_file())
            
            # Skip hidden files/dirs
            if part.startswith('.'):
                continue
            
            # Root level uppercase exceptions
            if is_root_level and part in self.valid_root_uppercase:
                continue
            
            # Photos directory exception (allows underscores)
            if 'photos' in parts:
                # Photos can use YYYY-MM-DD_subject_description_NN.jpg format
                if is_file and part.endswith(('.jpg', '.jpeg', '.png', '.gif')):
                    photo_pattern = r'^\d{4}-\d{2}-\d{2}_[a-z0-9\-_]+\d*\.(jpg|jpeg|png|gif)$'
                    if re.match(photo_pattern, part):
                        continue
            
            # Check for violations
            if ' ' in part:
                self.violations.append((str(relative_path), "Contains spaces"))
            elif part != part.lower() and not (is_root_level and part in self.valid_root_uppercase):
                self.violations.append((str(relative_path), "Not lowercase"))
            elif '_' in part and 'photos' not in parts:
                self.violations.append((str(relative_path), "Contains underscore (use hyphens)"))
            elif not self.is_valid_lowercase_hyphen(part) and 'photos' not in parts:
                if self.verbose:
                    print(f"  ⚠️  {relative_path}: {part}")
            
            # Check date formats
            if not self.is_valid_date_format(part):
                self.violations.append((str(relative_path), "Invalid date format"))
    
    def run(self):
        """Run naming validator on all files and directories"""
        print("🔍 Scanning repository structure...")
        
        for root, dirs, files in os.walk(self.repo_root):
            # Skip hidden directories
            dirs[:] = [d for d in dirs if not d.startswith('.')]
            
            root_path = Path(root)
            
            # Check directories
            for dir_name in dirs:
                self.check_path(root_path / dir_name)
            
            # Check files
            for file_name in files:
                if not file_name.startswith('.'):
                    self.check_path(root_path / file_name)
        
        self.report()
    
    def report(self):
        """Generate report of naming violations"""
        print("\n" + "="*80)
        print("📊 NAMING CONVENTION REPORT")
        print("="*80)
        
        print(f"\nViolations found: {len(self.violations)}\n")
        
        if self.violations:
            print("❌ NAMING VIOLATIONS:")
            print("-" * 80)
            for path, reason in self.violations:
                print(f"\n📄 {path}")
                print(f"   Issue: {reason}")
            print("\n" + "="*80)
            print("\n💡 Fix suggestions:")
            print("  - Replace spaces with hyphens: 'My File.md' → 'my-file.md'")
            print("  - Use lowercase: 'MyFile.md' → 'my-file.md'")
            print("  - Replace underscores: 'my_file.md' → 'my-file.md'")
            print("  - Date format: YYYY-MM-DD")
            print("\n" + "="*80)
            sys.exit(1)
        else:
            print("✅ All naming conventions followed!")
            print("="*80)
            sys.exit(0)

if __name__ == "__main__":
    import argparse
    
    parser = argparse.ArgumentParser(description="Validate naming conventions")
    parser.add_argument("--fix", action="store_true", help="Attempt to fix violations")
    parser.add_argument("--verbose", "-v", action="store_true", help="Verbose output")
    parser.add_argument("--repo", default=".", help="Repository root path")
    
    args = parser.parse_args()
    
    validator = NamingValidator(
        repo_root=args.repo,
        fix=args.fix,
        verbose=args.verbose
    )
    
    validator.run()