在python中将Word文件转换为json

Question

我有一些包含以下格式数据的 word（doc 和 docx）文件，我需要将它们转换为 JSON：

1.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890

2.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890

3.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890

在 python 中最简单的方法是什么？

Answer 1

有一个 docx module（尽管它明确不支持旧式 .doc 文件），您可以将其与 csv 阅读器结合使用以拆分列，然后从第一列中获取行索引。

from docx import Document
import json

document = Document('existing-document-file.docx')
lines = [para.text for para in document.paragraphs]
lines = [line.partition('.') for line in lines]
lines = [(int(row_num), row_text) for row_num, _, row_text in lines]
lines = [(n, [txt.partition(':') for txt in row_text.split(',')]) for n, row_text in lines]
lines = {n: {key.strip(): val.strip() for key, _, val in row} for n, row in lines}
json_result = json.dumps(lines)

通过您的示例输入，我使用此代码得到以下输出：

'{"1": {"Name": "ABC", "Place": "Maryland", "Country": "US", "PHONE NO.": "1234567890"},
"2": {"Name": "ABC", "Place": "Maryland", "Country": "US", "PHONE NO.": "1234567890"},
"3": {"Name": "ABC", "Place": "Maryland", "Country": "US", "PHONE NO.": "1234567890"}}'

Answer 2

要使用的库：

• 对于 docx 到 text 的转换使用 docx2text
• 对于json转换使用json库
• 要在字典中存储值，请使用defaultdict() from collections

步骤

1. 使用docx2text将文档转换为字符串
2. 将字符串转换为字符串列表，用新行\ncharacter 分割并删除垃圾空格''
3. 对于列表中的每个元素，由: , 拆分以进行操作。 通过:2删除数字拼接
4. 将列表li中的每个项目的每个键、值对存储在字典中
5. 将dict对象添加到json_li
6. 调用json.dumps(json_li)创建json字符串

代码

import docx2txt, json, collections
# step 1 get docx text
text = docx2txt.process("F:\workspace\StackOverFlow\guac.docx")
# convert to list
li = [x for x in text.split('\n')]
# remove ''s i.e Nones
li = list(filter(None, li))
print(li)
# json list
json_li = []
# convert and store all values
for x in li:
    x = x[2:] # remove 1. 2. 3. ...
    y = x.split(',')
    print(y)
    d = collections.defaultdict()
    for m in y:
        z = m.split(':')
        z1 = [x.strip() for x in z]
        d[z1[0]] = z1[1]
    json_li.append(d)
# JSON conversion
print(json.dumps(json_li, indent=4))

输出

['1.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890', '2.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890', '3.Name: ABC, Place: Maryland, Country: US, PHONE NO.:1234567890']
['Name: ABC', ' Place: Maryland', ' Country: US', ' PHONE NO.:1234567890']
['Name: ABC', ' Place: Maryland', ' Country: US', ' PHONE NO.:1234567890']
['Name: ABC', ' Place: Maryland', ' Country: US', ' PHONE NO.:1234567890']
[
    {
        "Name": "ABC",
        "Place": "Maryland",
        "Country": "US",
        "PHONE NO.": "1234567890"
    },
    {
        "Name": "ABC",
        "Place": "Maryland",
        "Country": "US",
        "PHONE NO.": "1234567890"
    },
    {
        "Name": "ABC",
        "Place": "Maryland",
        "Country": "US",
        "PHONE NO.": "1234567890"
    }
]

更新文档文件

如果您有 doc 文件，请使用

import textract
text = textract.process("path_to_file")

Answer 3

没有库/内置插件可以做到这一点。最简单的方法是将文件转换为 CSV（您自己删除所有逗号，然后用逗号替换空格，或者尽可能使用程序）

然后您可以使用 csv 包中的 DictReader 类将文件转换为字典，然后使用 json 模块将其转储为 json 字符串。

伪代码，例如转换为 csv 后。

import json

from csv import DictReader

COLUMN_NAMES = ['your', 'column', 'names,', '...'] 
    #Or the first row will be the column
    #(and the resulting key in the dictionary ) names

jsonCollection = {}
with open("your_csv_file.csv") as csvFile:
    #fieldnames is optional here
    reader = DictReader(csvFile, fieldnames=COLUMN_NAMES)
    for row in reader:
        for colName, rowVal in row.items():
            jsonCollection.setdefault(colName, []).append(rowVal)

json.dumps(jsonCollection) #should get you what you want