Ndtyjky

list1 = [{'key1': 'item1'}, {'key2': 'item2'}]

list2 = [{'key1': 'item1'}, {'key2': 'item2'}, {'key3': 'item3'}]

Is there a way to get the difference between those two lists?

Basically, I need a scaleable way to get the differences between 2 lists that contain dictionaries. So I'm trying to compare those lists, and just get a return of {'key3': 'item3'}

You could use a list comprehension:

list1 = [{'key1': 'item1'}, {'key2': 'item2'}]

list2 = [{'key1': 'item1'}, {'key2': 'item2'}, {'key3': 'item3'}]



print([x for x in list2 if x not in list1])

Which will give [{'key3': 'item3'}]

You can use set() with a comprehension like this example:

def get_diff(elm1, elm2):

    a = set((m, n) for k in elm1 for m, n in k.items())

    b = set((m, n) for k in elm2 for m, n in k.items())

    if len(b) > len(a):

        return dict(b - a)

    return dict(a - b)





list1 = [{'key1': 'item1'}, {'key2': 'item2'}]

list2 = [{'key1': 'item1'}, {'key2': 'item2'}, {'key3': 'item3'}]

get_diff(list1, list2)

Output:

{'key3': 'item3'}

in_list1_not_in_list2 = [i for i in list1 if i not in list2]

in_list2_not_in_list1 = [i for i in list2 if i not in list1]

Since dictionaries are not hashable there isn't an easy way to hash, but since we have one key and one val per dictionary, we can build our own key! So you can do something like this:

list1_set = set()



for dictionary in list1:

    key = dictionary.keys()[0]

    vals = dictionary.values()[0]

    custom_key = '{}|{}'.format(key,vals)

    list1_set.add(custom_key)



differences = 

for dictionary in list2:

    key = dictionary.keys()[0]

    vals = dictionary.values()[0]

    custom_key = '{}|{}'.format(key,vals)



    if custom_key not in list1_set:

        differences.append(dictionary)



print differences

output:

[{'key3': 'item3'}]

Not this solution is much more scalable then simply iterating through the first list because of the constant lookup ability.

You can also try using set.symmetric_difference() to get the difference between the sets both ways:

list1 = [{'key1': 'item1'}, {'key2': 'item2'}]

list2 = [{'key1': 'item1'}, {'key2': 'item2'}, {'key3': 'item3'}]



set1 = set(tuple(x.items())[0] for x in list1)

set2 = set(tuple(x.items())[0] for x in list2)



print([dict(list(set1.symmetric_difference(set2)))])

# [{'key3': 'item3'}]



print([dict(list(set2.symmetric_difference(set1)))])

# [{'key3': 'item3'}]

Another way would be to use itertools.filterfalse():

from itertools import filterfalse



diff1 = list(filterfalse(lambda d: d in list2, list1))

diff2 = list(filterfalse(lambda d: d in list1, list2))



print(diff1 + diff2)

# [{'key3': 'item3'}]

You can inform the dictionary how to hash itself and then you could use sets

import json



class HashableDict(dict):   

    def __hash__(self):

        # convert the dictionary to something hashable - in this case a str

        return hash(json.dumps(self)) 

then you can do

hashable_list1 = map(HashableDict, list1)

hashable_list2 = map(HashableDict, list2)

set(hashable_list2).difference(hashable_list1)

difference gives you the elements in lists2 that are not in list1.

If you wanted all the difference, so all the items that are not in both lists, do:

set(hashable_list2).symmetric_difference(hashable_list1)

Note this will not work for all dictionaries (e.g., dictionaries containing objects the json.dumps cannot work with) unless you handle those explicitly too with a custom JSONEncoder