迭代嵌套列表，元组和字典

Question

I've another question to the thematic of Iterate over nested lists and dictionaries .

I need some extended functionality to the topic of the link above. The iterable element now also contains tuples . Also integers in tuples need to be converted to a hex string. Therefor I tried with following code, to convert the tuples to lists.

for path, value in objwalk(element):
    if isinstance(value, tuple):
        parent = element
        for step in path[:-1]:
            parent = parent[step]
        parent[path[-1]] = list(value)

But my problem is, that tuples in tuples are not converted. How can I convert "sub-tuples" to lists in an elegant way?

Best regards wewa

PS: I created a new topic, because the other-one is fixed for me.

Answer 1

In this case, it would be easier to handle tuples directly in the objwalk structure traverser. Here is a modified version that converts tuples to lists before traversing over them to find nested elements:

def objwalk(obj, path=(), memo=None):
    if memo is None:
        memo = set()
    iterator = None
    if isinstance(obj, dict):
        iterator = iteritems
    elif isinstance(obj, (list, set)) and not isinstance(obj, string_types):
        iterator = enumerate
    if iterator:
        if id(obj) not in memo:
            memo.add(id(obj))
            for path_component, value in iterator(obj):
                if isinstance(value, tuple):
                    obj[path_component] = value = list(value)
                for result in objwalk(value, path + (path_component,), memo):
                    yield result
            memo.remove(id(obj))
    else:
        yield path, obj

Using a slightly modified example from your previous question, and the same hex solution I gave you in that question:

>>> element = {'Request': (16, 2), 'Params': ('Typetext', [16, 2], 2), 'Service': 'Servicetext', 'Responses': ({'State': 'Positive', 'PDU': [80, 2, 0]}, {})}
>>> for path, value in objwalk(element):
...     if isinstance(value, int):
...         parent = element
...         for step in path[:-1]:
...             parent = parent[step]
...         parent[path[-1]] = hex(value)
... 
>>> element
{'Params': ['Typetext', ['0x10', '0x2'], '0x2'], 'Request': ['0x10', '0x2'], 'Responses': [{'State': 'Positive', 'PDU': ['0x50', '0x2', '0x0']}, {}], 'Service': 'Servicetext'}

Answer 2

If the overhead of creating new objects is not an issue, I think it's pretty clear to go with:

def transform(obj):
    _type = type(obj)
    if _type == tuple: _type = list
    rslt = _type()
    if isinstance(obj, dict):
        for k, v in obj.iteritems():
            rslt[k] = transform(v)
    elif isinstance(obj, (list, tuple)):
        for x in obj:
            rslt.append(transform(x))
    elif isinstance(obj, set):
        for x in obj:
            rslt.add(transform(x))
    elif isinstance(obj, (int, long)):
        rslt = hex(obj)
    else:
        rslt = obj

    return rslt

element = transform(element)