Python flatten dictionaries

android Android

web Web

5-Mar-24 (Created: 5-Mar-24) | More in 'Python'

Python flatten dictionaries

satya - 3/5/2024, 8:16:20 AM

Why do I need this?

For providing universal configuration files
To have a model that is independent of representation
All data is fundamentally hierarchical
Can switch between toml, yaml, xml, or properties

satya - 3/5/2024, 8:16:56 AM

All one need to know may be here: SOF

satya - 3/5/2024, 8:17:25 AM

what you have here

Lot of sample code
Other suggestions
etc.

satya - 3/5/2024, 8:17:44 AM

Google question that is good for this

satya - 3/5/2024, 8:17:52 AM

is there a flatten method to flatten hierarchical dictionaries?

Search for: is there a flatten method to flatten hierarchical dictionaries?

satya - 3/5/2024, 8:23:52 AM

github flatten-dict homepage

Search for: github flatten-dict homepage

satya - 3/5/2024, 8:30:46 AM

flatten-dict homepage

satya - 3/5/2024, 9:49:14 AM

Example using flatten-dict


from flatten_dict import flatten

nested_dict = {
    'a': 1,
    'b': {
        'c': 2,
        'd': {
            'e': 3
        }
    }
}

flat = flatten(nested_dict, reducer='dot')
print(flat)

satya - 3/5/2024, 10:23:18 AM

Nature of hierarchy in toml


[database]
type = "sqlite"

[database.settings]
path = "/path/to/database/file.db"

[database.settings.options]
timeout = 30
cache_size = 5000

satya - 3/5/2024, 10:23:47 AM

Notice how "."s are used in section headers

satya - 3/5/2024, 10:26:03 AM

Some toml rules

1. "." are how the hierarchy is established in toml
2. they are allowed in section headers
3. Through those section headers form a hierarchy
4. Generally the "." are not allowed in keys
5. If you want, you have to put them in double quotes
6. The section headers are broken down into hierarchical dictionaries in memory
7. But the double quote keys with "."s are retained as they are

satya - 3/5/2024, 10:26:37 AM

So the above will look like this in python, the sections expanded to nested dictionaries


{
    "database": {
        "type": "sqlite",
        "settings": {
            "path": "/path/to/database/file.db",
            "options": {
                "timeout": 30,
                "cache_size": 5000
            }
        }
    }
}