在 Python 中将 CSV 转换为重组后的 JSON

Question

我有一个包含以下内容的 CSV 文件：

来源：data.opennepal.net

District,Zone,Geographical Region,Development Region,Causalities,In Number
Sindhupalchok,Bagmati,Mountain,Central,Total No. of Houses,66688
Sindhupalchok,Bagmati,Mountain,Central,Total Population,287798
Sindhupalchok,Bagmati,Mountain,Central,Dead Male,1497
Sindhupalchok,Bagmati,Mountain,Central,Dead Female,1943
Kathmandu,Bagmati,Hill,Central,Total No. of Houses,436344
Kathmandu,Bagmati,Hill,Central,Total Population,1744240
Kathmandu,Bagmati,Hill,Central,Dead Male,621
Kathmandu,Bagmati,Hill,Central,Dead Female,600

我的目标是从中生成一个这样的 JSON 对象：

{
  "district":{
    "Sindhupalchok":{
      "Causalities":{
        "Total No. of Houses":66688,
        "Total Population":287798,
        "Dead Male":1497,
        "Dead Female":1943
      },
      "geoInfo":{
        "Zone":"Bagmati",
        "geography":"Mountain",
        "Dev Region":"Central"
      }
    },
    "Kathmandu":{
      "Causalities":{
        "Total No. of Houses":436344,
        "Total Population":1744240,
        "Dead Male":621,
        "Dead Female":600
      },
      "geoInfo":{
        "Zone":"Bagmati",
        "geography":"Hill",
        "Dev Region":"Central"
      }
    }
  }
}

我尝试使用 csv.DictReader(csvfile, fieldnames)，但它会在 JSON 中生成冗余节点，难以解析且不必要地冗长。

我正在使用 python 2.x 这是我迄今为止的尝试：

>>> csvData = open('data.csv','rb')
>>> fieldnames = ("district", "zone", "geographicalRegion", "developmentRegion", "causalities", "injuredNumber")
>>> reader = csv.DictReader(csvData, fieldnames)
>>> rawJson = json.dumps([ row for row in reader ])

rawJson 不是我一直在寻找的那个。它只是将字段名映射到各个数据集。

所以问题是：如何在没有冗余节点的情况下创建这个 JSON 对象？

Answer 1

正如 glibdud 在 cmets 中提到的，您需要更手动地循环数据，以便创建所需的 JSON 结构。

我们将CSV数据的每一行读取为dict，并检查我们是否遇到了新的区，如果是，我们为它创建一个新的datadict，并插入一个geoInfo dict 变成 data。然后我们可以从该线路和该地区的后续线路收集伤亡数据。一旦我们收集了所有数据，我们就可以将 data dict 插入主 all_data dict。

为了测试代码，我将您的 .csv 数据放入名为“qdata.csv”的文件中

import csv
import json

filename = 'qdata.csv'
fieldnames = ('district', 'Zone', 'geography', 
    'Dev Region', 'casualties', 'injured')

geo_keys = ('Zone', 'geography', 'Dev Region')

all_data = {}
with open(filename, 'rb') as csvfile:
    reader = csv.DictReader(csvfile, fieldnames)
    # skip header
    next(reader)
    current_district = None
    for row in reader:
        district = row['district']
        if district != current_district:
            current_district = district
            data = all_data[district] = {}
            casualties = data['Casualties'] = {}
            data['geoInfo'] = dict((k, row[k]) for k in geo_keys)
        casualties[row['casualties']] = row['injured']

print json.dumps(all_data, indent=4, sort_keys=True)    

输出

{
    "Kathmandu": {
        "Casualties": {
            "Dead Female": "600", 
            "Dead Male": "621", 
            "Total No. of Houses": "436344", 
            "Total Population": "1744240"
        }, 
        "geoInfo": {
            "Dev Region": "Central", 
            "Zone": "Bagmati", 
            "geography": "Hill"
        }
    }, 
    "Sindhupalchok": {
        "Casualties": {
            "Dead Female": "1943", 
            "Dead Male": "1497", 
            "Total No. of Houses": "66688", 
            "Total Population": "287798"
        }, 
        "geoInfo": {
            "Dev Region": "Central", 
            "Zone": "Bagmati", 
            "geography": "Mountain"
        }
    }
}

这个输出不是完全你在你的问题中得到的，但我认为你应该能够从这里得到它。 :)