代码之家 › 专栏 › 技术社区 › David542

删除嵌套字典中的所有空键(通过yield)

python

David542 · 技术社区 · 6 年前

我有以下函数可以删除值为空或以下划线开头的键。它适用于非嵌套字典:

def _remove_missing_and_underscored_keys(d):
    if not d: return d
    for key in d.keys():
        if not d.get(key):
            del d[key]
        elif key.startswith('_'):
            del d[key]

    return d

d = {"Name": "David",
    "_Age": 50,
    "Numbers": [1,2,3,4,5],
    "Height": ""
}

>>> _remove_missing_and_underscored_keys(d)
{'Name': 'David', 'Numbers': [1, 2, 3, 4, 5]}

但是,我想创建上面的内容,以便它也可以删除嵌套项。我相信我需要使用一个yield语句来实现这一点,但是我没有幸运地正确实现它。下面是我想要它做什么的一个例子:

d = {
    "PlatformID": "B00EU7XL9Q",
    "Platform": "Amazon",
    "Type": "Collection",
    "Products": {
        "UK": {
            "URL": "http://www.amazon.co.uk/dp/B00EU7XL9Q",
            "Rating": None,
            "_IsAudited": True,
            "Offers": {
                "HDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                },
                "SDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                }
            }
        }
    }
}

>>> _remove_missing_and_underscored_keys(d)
{
    "PlatformID": "B00EU7XL9Q",
    "Platform": "Amazon",
    "Type": "Collection",
    "Products": {
        "UK": {
            "URL": "http://www.amazon.co.uk/dp/B00EU7XL9Q",
            "Offers": {
                "HDBUY": {
                    "Currency": "GBP",
                    "Cost": "14.99",
                },
                "SDBUY": {
                    "Currency": "GBP",
                    "Cost": "14.99",
                }
            }
        }
    }
}

换句话说,它将在dict的所有嵌套级别上执行上述操作。

4 回复 | 直到 6 年前

E.Serra 6 年前

def _remove_missing_and_underscored_keys(d):
    if not d: return d
    for key in d.keys():
        if not d.get(key):
            del d[key]
        elif key.startswith('_'):
            del d[key]
        elif type(d[key]) == dict:
            d[key] = _remove_missing_and_underscored_keys(d[key])

    return d

Skam 6 年前

我想你是想说,你需要用递归来解决这个问题。我认为使用发电机并不能完全解决你的问题。另一个警告是,不应该迭代正在更改的变量。这就是我创建副本的原因 _d 修改并返回它,并遍历原始结构。

import pprint

def _remove_missing_and_underscored_keys(d):
    if not d: return d
    _d = d.copy()
    for key in _d.keys():
        if not _d.get(key):
            del d[key]
        elif key.startswith('_'):
            del d[key]
        elif isinstance(_d[key], dict):
            _remove_missing_and_underscored_keys(_d[key])
    return _d

_d = {
    "PlatformID": "B00EU7XL9Q",
    "Platform": "Amazon",
    "Type": "Collection",
    "Products": {
        "UK": {
            "URL": "http://www.amazon.co.uk/dp/B00EU7XL9Q",
            "Rating": None,
            "_IsAudited": True,
            "Offers": {
                "HDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                },
                "SDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                }
            }
        }
    }
}
foo = _remove_missing_and_underscored_keys(_d)
pprint.pprint(foo)

输出:

 {'Platform': 'Amazon',
 'PlatformID': 'B00EU7XL9Q',
 'Products': {'UK': {'Offers': {'HDBUY': {'Cost': '14.99', 'Currency': 'GBP'},
                                'SDBUY': {'Cost': '14.99', 'Currency': 'GBP'}},
                     'URL': 'http://www.amazon.co.uk/dp/B00EU7XL9Q'}},
 'Type': 'Collection'}

Tushar 6 年前

只是递归。

再添加一个检查,看看主dict中的值是否是dict,并对其调用相同的函数。

# your code goes here
def _remove_missing_and_underscored_keys(d):
    if not d: return d
    for key in d.keys():
        if not d.get(key):
            del d[key]
        elif key.startswith('_'):
            del d[key]
        elif type(d[key]) is dict:
            #print("key '{}' stores a dict '{}', need to cleanup recursively".format(key, d[key]))
            d[key] = _remove_missing_and_underscored_keys(d[key])
            # Keep below check if you want to treat empty dict as `empty` as well
            if d[key] == None or d[key] == {}:
                del d[key]

    return d

d = {
    "PlatformID": "B00EU7XL9Q",
    "Platform": "Amazon",
    "Type": "Collection",
    "Products": {
        "UK": {
            "URL": "http://www.amazon.co.uk/dp/B00EU7XL9Q",
            "Rating": None,
            "_IsAudited": True,
            "Offers": {
                "HDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                },
                "SDBUY": {
                    "Currency": "GBP",
                    "FutureReleaseStartDate": None,
                    "Cost": "14.99",
                    "IsFutureRelease": False
                },
                "x" : {
                    "y":None
                }
            }
        }
    }
}

e = _remove_missing_and_underscored_keys(d)
print(e)

见效: https://ideone.com/5xDDZl

上面的代码还处理存储在任何键上的空dict,或者在递归地清除后变为空的任何dict。如果需要,您可以取消该支票。

Ajax1234 6 年前

您可以使用递归和字典理解:

d = {'PlatformID': 'B00EU7XL9Q', 'Platform': 'Amazon', 'Type': 'Collection', 'Products': {'UK': {'URL': 'http://www.amazon.co.uk/dp/B00EU7XL9Q', 'Rating': None, '_IsAudited': True, 'Offers': {'HDBUY': {'Currency': 'GBP', 'FutureReleaseStartDate': None, 'Cost': '14.99', 'IsFutureRelease': False}, 'SDBUY': {'Currency': 'GBP', 'FutureReleaseStartDate': None, 'Cost': '14.99', 'IsFutureRelease': False}}}}}
def _del(_d):
  return {a:_del(b) if isinstance(b, dict) else b for a, b in _d.items() if b and not a.startswith('_')}

import json
print(json.dumps(_del(d), indent=4))

输出:

{
 "PlatformID": "B00EU7XL9Q",
 "Platform": "Amazon",
 "Type": "Collection",
 "Products": {
    "UK": {
        "URL": "http://www.amazon.co.uk/dp/B00EU7XL9Q",
        "Offers": {
            "HDBUY": {
                "Currency": "GBP",
                "Cost": "14.99"
            },
            "SDBUY": {
                "Currency": "GBP",
                "Cost": "14.99"
            }
        }
     }
   }
}