Class to access nested dics via attributes shows different values via print and __dict__

Question

I am working on a project where I need objects with a few attributes (e.g. date, name) and then a collection of values (normally pd.DataFrames with different aggregates, but I will use integers as an example here). Instead of doing multiple nested dicts and typing many extra characters for the keys like data["df"]["yearly"], it would be easier to access the values with data.df.yearly. (is there a better structure?)

Thus I use a class which sets the keys as attributes. But I get different outputs depending on how I access that data.

Here is the code:

class Map(dict):
    """
    https://stackoverflow.com/questions/2352181/how-to-use-a-dot-to-access-members-of-dictionary
    Example:
    m = Map({'first_name': 'Eduardo'}, last_name='Pool', age=24, sports=['Soccer'])
    """
    def __init__(self, *args, **kwargs):
        super(Map, self).__init__(*args, **kwargs)
        for arg in args:
            if isinstance(arg, dict):
                for k, v in arg.items():
                    self[k] = v

        if kwargs:
            for k, v in kwargs.items():
                self[k] = v

    def __getattr__(self, attr):
        super( Map, self ).__getattr__( attr )
        return self.get(attr)

    def __setattr__(self, key, value):
        self.__setitem__(key, value)

    def __setitem__( self, key, value ):
        super( Map, self ).__setitem__( key, value )
        if isinstance( value, dict ):
            self.__dict__.update( {key: Map(value) } )         
        else:
            self.__dict__.update( {key: value} )

    def __delattr__(self, item):
        self.__delitem__(item)

    def __delitem__(self, key):
        super(Map, self).__delitem__(key)
        del self.__dict__[key]   

Now when I initialize an instance and assign some data to it in the following way, I will have two different outputs depending on if I write data.df.yearly or data.df["yearly"].

Does someone see the mistake (or is there a suggestion for a better data structure?).

Create an instance of Map()

data = Map()
data.start_date = "2020-01-01"
data.finish_date = '2021-01-01'
data.name = "Type A"
data.df = {"yearly" : 0, "monthly" : 0, "weekly" : 0}

Change the values of data.df

for time in data.df.keys():
    data.df[time] = 123

Expected output:

data
> {'start_date': '2020-01-01',
>  'finish_date': '2021-01-01',
>  'name': 'Type A',
>  'df': {'yearly': 123, 'monthly': 123, 'weekly': 123}}

Output (same with print(data)):

data
> {'start_date': '2020-01-01',
>  'finish_date': '2021-01-01',
>  'name': 'Type A',
>  'df': {'yearly': 0, 'monthly': 0, 'weekly': 0}}

Output via dict matches epected output:

data.__dict__
> {'start_date': '2020-01-01',
>  'finish_date': '2021-01-01',
>  'name': 'Type A',
>  'df': {'yearly': 123, 'monthly': 123, 'weekly': 123}}

Answer 1

Your __setitem__ method is calling the superclass method with the original value, but if the value if a dictionary it converts it to a Map when putting it into self.__dict__. So these two locations contain different objects, and nothing keeps them in sync.

You should convert the value to a Map before storing it in both places.

    def __setitem__( self, key, value ):
        if isinstance(value, dict):
            value = Map(value)
        super( Map, self ).__setitem__( key, value )
        self.__dict__.update({key: value})