Fix MongoDB's dumped JSONs which save indices information.

fix_mongodump.py

#!python3
import argparse
import json
import os
import shutil
from pathlib import Path

def convert(obj):
    if type(obj) == list:
        return [convert(item) for item in obj]
    elif type(obj) == dict:
        if '$numberInt' in obj:
            return int(obj['$numberInt'])
        elif '$numberDouble' in obj:
            return float(obj['$numberDouble'])
        else:
            return {key: convert(item) for key, item in obj.items()}
    else:
        return obj
    assert False


if __name__ == '__main__':
    parser = argparse.ArgumentParser(
        description="Fix MongoDB's dumped JSONs which save indices information.")
    parser.add_argument(
        'path', 
        help='Path of the target file or the folder keep the JSONs.')
    parser.add_argument(
        '-r', '--restore',
        action='store_true',
        help='Restore from the backup files')

    args = parser.parse_args()
    if args.restore:
        if args.path.endswith('.json.bak'):
            is_file = True
            file_path = Path(args.path[:-4])
            backup_path = Path(args.path)
        elif args.path.endswith('.json'):
            is_file = True
            file_path = Path(args.path)
            backup_path = Path(args.path + '.bak')
        else:
            is_file = False
            path = Path(args.path)
            if not path.is_dir():
                print('Cannot find the backup files.')
                exit(1)
        if is_file:
            if not backup_path.is_file():
                print('Cannot find the backup files.')
                exit(1)
            else:
                shutil.move(backup_path, file_path)
        else:
            for filename in os.listdir(path):
                if filename.endswith('.json.bak'):
                    file_path = Path(path, filename[:-4])
                    backup_path = Path(path, filename)
                    shutil.move(backup_path, file_path)
    else:  # convert JSONs
        path = Path(args.path)
        if path.is_file() and path.suffix == '.json':
            backup_path = Path(str(path) + '.bak')
            shutil.copy2(path, backup_path)
            with open(path) as fp:
                obj = json.load(fp)
            with open(path, 'w') as fp:
                json.dump(convert(obj), fp)    
        elif path.is_dir():
            for filename in os.listdir(path):
                if filename.endswith('.json'):
                    file_path = Path(path, filename)
                    backup_path = Path(path, filename + '.bak')
                    shutil.copy2(file_path, backup_path)
                    with open(file_path) as fp:
                        obj = json.load(fp)
                    with open(file_path, 'w') as fp:
                        json.dump(convert(obj), fp)
        else:
            print('Cannot find the JSON files.')
            exit(1)