要从文本创建一个csv文件并包含列，可以使用csv.DictWriter创建列

import csv

# Set fieldnames for the CSV file
fieldnames = ['Name', 'Error Text', 'Correct Text', 'Type of Error']

# convert data to dictionaries 
rows = []

with open('yourfile.txt', 'r') as infile:
    lines = infile.readlines()
    for i in range(0, len(lines), 4):  # Process four lines at a time
        name = lines[i].split(":")[1].strip()
        error_text = lines[i + 1].split(":")[1].strip()
        correct_text = lines[i + 2].split(":")[1].strip()
        type_of_error = lines[i + 3].split(":")[1].strip()
        rows.append({'Name': name, 'Error Text': error_text, 'Correct Text': correct_text, 'Type of Error': type_of_error})

# Write CSV file
with open('yourfileoutput.csv', 'w', newline='') as outfile:
    writer = csv.DictWriter(outfile, fieldnames=fieldnames)
    writer.writeheader()
    writer.writerows(rows)

字符串

逐行读取输入文件。在结肠处裂开。检查关键字是否按顺序排列。构建输入数据的字典。
打开一个输出文件（CSV）并转储列标题和字典内容，如下所示：

from itertools import cycle

COLUMNS = ['Name', 'Error Text', 'Correct Text', 'Type of Error']
TXTFILE = '/Volumes/G-Drive/input.txt'
CSVFILE = '/Volumes/G-Drive/output.csv'
SEP = '|'

d = {}

column = cycle(COLUMNS)

with open(TXTFILE) as indata:
    for line in indata:
        k, v = line.split(':')
        if k == next(column):
            d.setdefault(k, []).append(v.strip())
        else:
            raise ValueError(f'Unexpected keyword "{k}"')

with open(CSVFILE, 'w') as outdata:
    print(SEP.join(COLUMNS), file=outdata)
    for data in zip(*d.values()):
        print(SEP.join(data), file=outdata)

字符串

假设输入文件为：

Name: name_1
Error Text: some_text_1_1
Correct Text: some_text_1_2
Type of Error: some_text_1_3
Name: name_2
Error Text: some_text_2_1
Correct Text: some_text_2_2
Type of Error: some_text_2_3
Name: name_3
Error Text: some_text_3_1
Correct Text: some_text_3_2
Type of Error: some_text_3_3

字符串
只需使用pure pandas，read_csv和': '作为分隔符，然后使用assign作为重复数据删除列，该列基于使用每个“Name”重新启动的组（在cumsum的帮助下），最后使用pivot，并导出to_csv：

df = (pd.read_csv(filename, sep=r':\s+', engine='python', header=None)
        .assign(idx=lambda d: d[0].eq('Name').cumsum())
        .pivot(index='idx', columns=0, values=1)
     )
df.to_csv('outfile.csv', index=False)

型
输出文件：

Correct Text,Error Text,Name,Type of Error
some_text_1_2,some_text_1_1,name_1,some_text_1_3
some_text_2_2,some_text_2_1,name_2,some_text_2_3
some_text_3_2,some_text_3_1,name_3,some_text_3_3

型
要保持列的原始顺序，请执行以下操作：

df = (pd.read_csv(filename, sep=r':\s+', engine='python', header=None)
        .assign(idx=lambda d: d[0].eq('Name').cumsum())
        .pipe(lambda d: d.pivot(index='idx', columns=0, values=1)[d[0].unique()])
     )
df.to_csv('outfile.csv', index=False)

型
输出量：

Name,Error Text,Correct Text,Type of Error
name_1,some_text_1_1,some_text_1_2,some_text_1_3
name_2,some_text_2_1,some_text_2_2,some_text_2_3
name_3,some_text_3_1,some_text_3_2,some_text_3_3

型