How to unpack a dictionary of list (of dictionaries!) and return as grouped tuples?

I have a data structure consisting of mixed dictionaries and lists. I am trying to unpack this so as to get tuples of the keys and all sub-values for each key.

I am working with list comprehensions, but just not getting it to work. Where am I going wrong?

I saw many other answers about unpacking a list-of-lists (e.g. 1,2), but could not find an example where a single key unpacks against multiple sub-values.

desired output --> [('A',1,2),('B',3,4)]
actual output --> [('A',1), ('A',2), ('B',3), ('B',4)]

code:

dict_of_lists = {'A':[{'x':1},{'x':2}], 'B':[{'x':3},{'x':4}] }
print [(key,subdict[subkey],) for key in dict_of_lists.keys() for subdict in dict_of_lists[key] for subkey in subdict.keys()]

Solution

When list comprehensions become

long
unclear/hard to read
and most importantly, do not work

ditch them and go with the manual for loop(s) every time:

Python 2.x

def unpack(d):
    for k, v in d.iteritems():
        tmp = []
        for subdict in v:
            for _, val in subdict.iteritems():
                tmp.append(val)
        yield (k, tmp[0], tmp[1])


print list(unpack({'A':[{'x':1},{'x':2}], 'B':[{'x':3},{'x':4}] }))

Python 3.x

def unpack(d):
        for k, v in d.items():
            tmp = []
            for subdict in v:
                for _, val in subdict.items():
                    tmp.append(val)
            yield (k, *tmp) # stared expression used to unpack iterables were
                            # not created yet in Python 2.x                

print(list(unpack({'A':[{'x':1},{'x':2}], 'B':[{'x':3},{'x':4}] })))

Output:

[('A', 1, 2), ('B', 3, 4)]