将嵌套字典转换为表/父子结构，Python 3.6

Posted 2023-03-11

【中文标题】将嵌套字典转换为表/父子结构，Python 3.6

【英文标题】：Convert Nested Dictionary into Table/Parent Child Structure, Python 3.6

【发布时间】：2020-04-03 19:13:24

【问题描述】：

想从下面的代码转换嵌套字典。

import requests
from bs4 import BeautifulSoup

url = 'https://www.bundesbank.de/en/statistics/time-series-databases/time-series-databases/743796/openAll?treeAnchor=BANKEN&statisticType=BBK_ITS'
result = requests.get(url)
soup = BeautifulSoup(result.text, 'html.parser')

def get_child_nodes(parent_node):
    node_name = parent_node.a.get_text(strip=True)

    result = "name": node_name, "children": []

    children_list = parent_node.find('ul', recursive=False)
    if not children_list:
    return result

    for child_node in children_list('li', recursive=False):
    result["children"].append(get_child_nodes(child_node))

    return result

Data_Dict = get_child_nodes(soup.find("div", class_="statisticTree"))

是否可以如图所示导出父-子？

以上代码来自@alecxe 的回答：Fetch complete List of Items using BeautifulSoup, Python 3.6

我试过了，但它太复杂了，无法理解，请帮忙。

字典：http://s000.tinyupload.com/index.php?file_id=97731876598977568058

样本字典数据：

"name": "Banks", "children": ["name": "Banks", "children": ["name": "Balance sheet items", "children": 
["name": "Minimum reserves", "children": ["name": "Reserve maintenance in the euro area", "children": [], "name": "Reserve maintenance in Germany", "children": []], 

"name": "Bank Lending Survey (BLS) - Results for Germany", "children": ["name": "Lending", "children": ["name": "Enterprises", "children": ["name": "Changes over the past three months", "children": ["name": "Credit standards and explanatory factors", "children": ["name": "Overall", "children": [], "name": "Loans to small and medium-sized enterprises", "children": [], "name": "Loans to large enterprises", "children": [], "name": "Short-term loans", "children": [], "name": "Long-term loans", "children": []], "name": "Terms and conditions and explanatory factors", "children": ["name": "Overall", "children": ["name": "Overall terms and conditions and explanatory factors", "children": [], "name": "Margins on average loans and explanatory factors", "children": [], "name": "Margins on riskier loans and explanatory factors", "children": [], "name": "Non-interest rate charges", "children": [], "name": "Size of the loan or credit line", "children": [], "name": "Collateral requirements", "children": [], "name": "Loan covenants", "children": [], "name": "Maturity", "children": []], "name": "Loans to small and medium-sized enterprises", "children": [], "name": "Loans to large enterprises", "children": []], "name": "Share of enterprise rejected loan applications", "children": []], "name": "Expected changes over the next three months", "children": ["name": "Credit standards", "children": []]], "name": "Households", "children": ["name": "Changes over the past three months", "children": ["name": "Credit standards and explanatory factors", "children": ["name": "Loans for house purchase", "children": [], "name": "Consumer credit and other lending", "children": []], 

【问题讨论】：

由于转换只是字典，与BeautifulSoup无关，请提供一些可以复制/粘贴到Python中的示例数据，去掉不需要的部分代码;将其设为minimal reproducible example。

@kaya3，更新样本数据，请检查

【参考方案1】：

您可以使用递归函数来处理这个问题。

def get_pairs(data, parent=''):
    rv = [(data['name'], parent)]
    for d in data['children']:    
        rv.extend(get_pairs(d, parent=data['name']))
    return rv

Data_Dict = get_child_nodes(soup.find("div", class_="statisticTree"))

pairs = get_pairs(Data_Dict)

然后，您可以选择创建 DataFrame，或立即导出到 csv，如您的示例输出所示。要创建 DataFrame，我们可以这样做：

df = pd.DataFrame(get_pairs(Data_Dict), columns=['Name', 'Parent'])

给予：

                                             Name               Parent
0                                           Banks                     
1                                           Banks                Banks
2                             Balance sheet items                Banks
3                                Minimum reserves  Balance sheet items
4            Reserve maintenance in the euro area     Minimum reserves
                                          ...                  ...
3890  Number of transactions per type of terminal  Payments statistics
3891   Value of transactions per type of terminal  Payments statistics
3892                   Number of OTC transactions  Payments statistics
3893                    Value of OTC transactions  Payments statistics
3894                        Issuance of banknotes  Payments statistics

[3895 rows x 2 columns]

或者要输出到 csv，我们可以使用 csv 内置库：

import csv

with open('out.csv', 'w', newline='') as f:
    writer = csv.writer(f, delimiter=',')
    writer.writerow(('Name', 'Parent'))
    for pair in pairs:
        writer.writerow(pair)

输出：