I have a function where one of the parameters is a dictionary, but it is optional.
I need to assign a value in this dictionary, but take into account the case where this parameter is not filled in.
The following solution uses two lines for this:
def uma_funcao(um_dict=None):
um_dict = um_dict or {}
um_dict['uma_chave'] = True
I believe there is a more Pythonic way of doing this, in one line. Is it possible?
No, I believe that this form you presented is already minimal. I can not talk about Python 3, as I have no experience.
(Answers in the negative are complicated, but I have no evidence to support it except the absence of evidence to the contrary ...)
Update: As 3 people have already posted the same incorrect answer, I'll point it out here to avoid future identical responses.
>>> def uma_funcao(um_dict={}):
... um_dict['uma_chave'] = True
... return um_dict
...
>>> x = uma_funcao()
>>> x['teste'] = True
>>> y = uma_funcao()
>>> y
{'uma_chave': True, 'teste': True}
As you can see, you should not use a changeable value as a default parameter for a function. The expression {} creates a before object to use as a parameter, so any call of uma_funcao without passing parameters will use the same < dict . Since this is seldom the desired behavior, there is no option but to use None followed by a test, just like the original question code.
Another way I thought of it, very common in languages like C, was to merge the assignment with use into a single expression:
(um_dict = um_dict or {})['uma_chave'] = True
However this is not allowed by the Python syntax (at least 2, if something like this has been added in future versions, it is not known to me).
Is not this good here?
def uma_funcao(um_dict = {}):
um_dict['uma_chave'] = True
return um_dict
I do not understand why you assign None instead of assigning {} directly. If there is any reason for this, please go into more detail.
One way to get around the problem presented by @mgibsonbr is to return a copy of the dictionary instead of it:
def uma_funcao(um_dict={}):
um_dict['uma_chave'] = True
return um_dict.copy()
Note that if um_dict is passed, it will continue to be modified:
>>> def uma_funcao(um_dict={}):
... um_dict['uma_chave'] = True
... return um_dict.copy()
...
>>> x = uma_funcao()
>>> x['teste'] = True
>>> y = uma_funcao()
>>> y
{'uma_chave': True}
>>> d = {'uma_chave': False}
>>> z = uma_funcao(d)
>>> z
{'uma_chave': True}
>>> d
{'uma_chave': True}
It is possible to do everything in a single line, even though it is far from being considered something pythonic:
def uma_funcao(um_dict={}):
return um_dict.update({'uma_chave': True}) or um_dict.copy()