NXC Spider2CSV

spider2csv.py

#!/usr/bin/env python3 
'''
Converts nxc file share spider_plus json output to csv, for better grepping.
'''

import json
import csv
import os
import argparse
import sys
import glob

def get_server_name(filename):
    # Extract server name from the json filename
    # Assuming filename format like 'jup-fs.jupiter.lab.json'
    try:
        base_name = os.path.basename(filename)
        # Remove .json extension and split by dots
        name_parts = base_name.replace('.json', '').split('.')
        # The server name is the first part before the first dot
        server_name = name_parts[0]
        return server_name
    except IndexError:
        return "unknown"

def get_json_files_from_directory(directory_path):
    """Get all JSON files from the specified directory (non-recursive)"""
    if not os.path.isdir(directory_path):
        print(f"Error: Directory '{directory_path}' not found", file=sys.stderr)
        sys.exit(1)
    
    json_files = glob.glob(os.path.join(directory_path, "*.json"))
    if not json_files:
        print(f"Error: No JSON files found in directory '{directory_path}'", file=sys.stderr)
        sys.exit(1)
    
    return json_files

def parse_json_to_smb(input_file, output_file):
    # Read JSON file
    try:
        with open(input_file, 'r') as f:
            data = json.load(f)
    except FileNotFoundError:
        print(f"Error: Input file '{input_file}' not found", file=sys.stderr)
        sys.exit(1)
    except json.JSONDecodeError:
        print(f"Error: Invalid JSON in file '{input_file}'", file=sys.stderr)
        sys.exit(1)
    
    # Get server name
    server_name = get_server_name(input_file)
    
    # Prepare CSV output
    try:
        with open(output_file, 'w', newline='') as f:
            writer = csv.writer(f)
            # Write header
            writer.writerow(['SMB_Path', 'Size', 'Last_Modified'])
            
            # Process each share and its contents
            for share, contents in data.items():
                for filepath, details in contents.items():
                    # Construct SMB path
                    smb_path = f"\\\\{server_name}\\{share}\\{filepath}"
                    # Replace forward slashes with backslashes
                    smb_path = smb_path.replace('/', '\\')
                    
                    # Write row with path and details
                    writer.writerow([
                        smb_path,
                        details.get('size', ''),
                        details.get('mtime_epoch', '')
                    ])
    except PermissionError:
        print(f"Error: Permission denied when writing to '{output_file}'", file=sys.stderr)
        sys.exit(1)

def main():
    parser = argparse.ArgumentParser(
        description='Convert JSON file system structure to SMB share paths CSV',
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog='''
Example:
    %(prog)s -i /path/to/input.json -o output.csv
    %(prog)s -d /path/to/json/files/ -o output.csv
    
The input JSON should contain file system structure information.
The output CSV will contain SMB paths and file details.''')
    
    parser.add_argument('-i', '--input', 
                        help='Input JSON file path')
    parser.add_argument('-d', '--directory',
                        help='Directory containing JSON files to process')
    parser.add_argument('-o', '--output',
                        required=True,
                        help='Output CSV file path')
    parser.add_argument('-v', '--verbose',
                        action='store_true',
                        help='Enable verbose output')

    args = parser.parse_args()

    # Validate that either input file or directory is provided
    if not args.input and not args.directory:
        parser.error("Either --input or --directory must be specified")
    
    if args.input and args.directory:
        parser.error("Cannot specify both --input and --directory")

    # Determine files to process
    if args.directory:
        json_files = get_json_files_from_directory(args.directory)
        if args.verbose:
            print(f"Found {len(json_files)} JSON files in directory {args.directory}")
    else:
        json_files = [args.input]

    if args.verbose:
        print(f"Output will be written to {args.output}")

    # Process files and combine output
    try:
        with open(args.output, 'w', newline='') as f:
            writer = csv.writer(f)
            # Write header
            writer.writerow(['SMB_Path', 'Size', 'Last_Modified'])
            
            for json_file in json_files:
                if args.verbose:
                    print(f"Processing {json_file}")
                
                # Read JSON file
                try:
                    with open(json_file, 'r') as jf:
                        data = json.load(jf)
                except FileNotFoundError:
                    print(f"Error: Input file '{json_file}' not found", file=sys.stderr)
                    continue
                except json.JSONDecodeError:
                    print(f"Error: Invalid JSON in file '{json_file}'", file=sys.stderr)
                    continue
                
                # Get server name
                server_name = get_server_name(json_file)
                
                # Process each share and its contents
                for share, contents in data.items():
                    for filepath, details in contents.items():
                        # Construct SMB path
                        smb_path = f"\\\\{server_name}\\{share}\\{filepath}"
                        # Replace forward slashes with backslashes
                        smb_path = smb_path.replace('/', '\\')
                        
                        # Write row with path and details
                        writer.writerow([
                            smb_path,
                            details.get('size', ''),
                            details.get('mtime_epoch', '')
                        ])
    except PermissionError:
        print(f"Error: Permission denied when writing to '{args.output}'", file=sys.stderr)
        sys.exit(1)

    if args.verbose:
        print("Conversion completed successfully")

if __name__ == "__main__":
    main()