objdict-bf

objdict-bf is a Python module that provides a wrapper class to conveniently manipulate dictionaries or dict-based nested structures using attribute-like syntax. It is intended mostly to ease manipulation of serialized data, web requests and responses, configuration files, dynamic prototyping...

Features

• Attribute-style access to dictionary items (e.g., obj.key instead of obj['key']).
• Synchronization with the original dictionary if passed at instantiation.
• Utility methods for recursive conversion of nested structures to and from objdict and dict.
• Serialization and deserialization methods for both strings and files with json, toml, yaml and jsonpickle backend support.
• Advanced default value attribution features for missing keys.
• optional object-like behavior, by auto-passing the objdict instance as 'self' to callable attributes having 'self' in their signature.

Installation

pip install objdict-bf

Signature of the constructor

objdict(*args,_use_default=False,_default=None,_file=None,_backend=None,_auto_self=False,**kwargs)

Parameters:

• *args: either dicts, objdicts or iterables on key:value pairs. If the first arg is a dict, it will serve as the internal _data_dict of the objdict instance.
• _use_default: boolean, determines if a default value is attributed to missing keys
• _default: can be any value or callable. If it is callable with adequate signature, this callable will be used to handle default values generation.
• _file: reference to a file path for dumping (extension must match the backend used)
• _backend: either 'json', 'toml','yaml' or 'jsonpickle'. Determines the backend used for serialization/deserialization when dumping/loading (None defaults to 'json').
• _auto_self: boolean. Determines if the instance is auto-passed a 'self' to its callable attributes having 'self' in their signature (mocked object behavior).
• **kwargs: key value pairs passed as kwargs to update the objdict

Usage

from objdict_bf import objdict

# Create an objdict with some initial data
data = objdict(
    name='John',
    age=30,
    location='New York'
)

#Or synchronize with an existing dict
d={'name': 'John', 'age': 30, 'location': 'New York'}
data = objdict(d)

#Access data using attribute-style access
print(data.name)  # Output: John
print(data.age)   # Output: 30

#Modify data
data.age = 31

#Create a new key:value pair
data.job='developer'

#Changes are reflected on the original dict
print(d['age']) #Ouput: 31
print(d['job']) #Ouput: 'developer'

#Chaining attributes is supported for nested structures involving lists
d={
    'profile':{
        'name':'John',
        'hobbies':[
            {'type':'sport','title':'tennis'},
            {'type':'music','title':'guitar playing'}
        ]
    }
}
data = objdict(d)

print(data) #Output: the repr of the above dict
print(data.profile.hobbies[1].title) #Output: guitar playing

#Conversion of dict items to their objdict version is automatic upon access. 
#The created objdicts will inherit the parent objdict settings, namely: _backend,_use_default, _default, _auto_self).
#The objdict being essentially a wrapper interface on the initial dict,  
#this conversion is reflected in the initial dict content as well

print(isinstance(data.profile.hobbies[1],objdict)) #Output: True
print(isinstance(d['profile']['hobbies'][1],objdict)) #Output: True

#to_dict returns the underlying dict, converting recursively all objdicts found in the nested structure back to dicts
print(d is data.to_dict()) #Ouptut: True
print(isinstance(d['profile']['hobbies'][1], dict)) #Output: True 

#-----------------------------Serialization-------------------------------

# Serialize to JSON string
json_string = data.dumps()
print(json_string)
#or use another backend for serialization 
toml_string=data.dumps(_backend='toml')
print(toml_string)

#dump to a file
data.dump("my_json_file.json")
#or
data.dump("my_toml_file.toml",_backend='toml')

#make some more changes
data.email="dummy.email@gmail.com"

#the reference to the file and backend preference from the last dump is kept in the objdict instance so you don't have to pass them again
data.dump()

# Deserialize from a string (new instance keeping reference to the chosen backend)
data = objdict.loads(json_string)
#or
data = objdict.loads(toml_string,_backend='toml')


# Deserialize from a file (new instance keeping reference to the chosen file and backend)
data = objdict.load("my_json_file.json")
#or
data = objdict.load("my_toml_file.toml",_backend='toml')

#Mismatching file extension and backend will throw an exception 

#any class method creating a new instance can be passed parameters accepted in the objdict constructor to control the properties of the created instance:
data = objdict.loads(string,_backend='json',_use_default=True,_default=None,_auto_self=False)
data = objdict.load(file,_backend='toml',_use_default=False,_auto_self=True)

#update data
data.email="dummy.email@gmail.com"
data.user="dummy_username"

#dump changes to the file using the previously chosen backend 
data.dump()

#-------------------Working with default value generators-------------------


#Default value (None) when accessing a missing key
obj=objdict(_use_default=True)

#Will set the value to None and won't raise a KeyError
print(obj.a) #Output: None

#Or, choose a default value
obj=objdict(_use_default=True,_default=3)
#Missing key will be initialized to 3 and returned
print(obj.a) #Output: 3

#Or pass a default value generator depending on the key (must have 'key' in its signature)
default_gen=lambda key: f"Missing key: {key}" 
obj=objdict(_use_default=True,_default=default_gen)
print(obj.a) #Output: "Missing key: a"
print(obj.b) #Output: "Missing key: b"

#Or pass a default value generator whose output depends on the current state/content of the objdict
#Must have 'self' in its signature
#Will use 'self' as the keyword refering to the current objdict instance
def default_gen(self):
    if 'a' in self:
        return self.a.value
    else:
        return objdict(value=5)
        
obj=objdict(_use_default=True,_default=default_gen)
print(obj.a) #Output: {'value':5}
print(obj.b) #Output: 5

#Accepted signature of default value generators are () ; (self,); (key,) ; (self,key)
#This allows implementing context-aware and key-dependant logic for default value attribution. 
#Any other signature will be considered invalid and will fall back to assign the callable itself as the default value for all keys.

#Example: Using a default value generator to automatically create new child objdict instances inheriting the parent's settings when accessing missing keys
def child_instance(self):
    return objdict(_use_default=True,_default=child_instance,_backend=self._backend,_auto_self=self._auto_self)

obj=objdict(_use_default=True,_default=child_instance,_backend='toml',_auto_self=True)
obj.a.b.c=3
print(obj) #Output: {'a':{'b':{'c':3}}}
#child elements inherit the chosen parent properties
print(obj.a.b._backend) #Output: 'toml'
print(obj.a.b._auto_self) #Output: True

#The child_instance generator hard-coded above is already implemented as the objdict.child_instance static method which you may pass as _default parameter
obj=objdict(_use_default=True,_default=objdict.child_instance)
obj.a.b.c=3
print(obj) #Output: {'a':{'b':{'c':3}}}

#--------------------------------Mock objects-------------------------------

#Using the objdict as a mocked object with context aware methods thanks to the _auto_self parameter which automatically passes the objdict instance as 'self' to callable attributes having 'self' as first parameter in their signature.

obj=objdict(_auto_self=True)
obj.a=2

#create a function with 'self' as first parameter (any other name won't receive the instance)
def add_to_a(self,b):
    self.a+=b

#attach the function as attribute
obj.add_to_a=add_to_a
obj.add_to_a(3)
print(obj.a) #output 5

License

This project is licensed under the MIT License - see the LICENSE file for details.