Storage dict features.

[natankeddem]

Utilizing standard dict for highly nested arbitrary data can really get really complicated. Has there been any thought to integrating any functionality similar to libraries like https://github.com/mewwts/addict ?

To be more specific I don't mean the attribute access just the defaulting/arbitrary feature. Maybe utilizing 'defaultdict'?

[falkoschindler]

I don't quite understand the advantage. If, e.g., app.storage.general was an addict, you could write

x = app.storage.general.count

instead of

x = app.storage.general.get('count', 0)

But is it so much better? Personally I'm not 100% sure if I like the implicitness of addict. You never know if something is really the value or just the default value because a key is missing.

[natankeddem]

For me the attribute by name feature is interesting but not that useful. What I do find useful is the ability to make deeply nested dicts without extra verbosity. It might not be that great of a feature but I prefer the look of the addict syntax.

normal_dict = {'a': {'b': {'c': {'d': {'e': 'f'}}}}}
addict['a']['b']['c']['d']['e'] = 'f'