在换行符之间创建空格\n

Question

背景

我有下面的df，它包含一个Text列，该列使用nltkSpaceTokenizer进行了标记以保留\n

import pandas as pd
text =[list(['\n[PROV', 'REPORT]\nPerson', 'Name:', '\n','John', 'Dear\nProgram', 'Date:', '1/11/2000', '10:42', 'AM\nMR']), 
           list(['\nToday', 'Name:', '\n','James', 'Jay\nProgram', 'Date:', '3/11/2000', '1:45', 'PM\nmissing']), 
           list(['\n[NEWS', 'REPORT]\nPerson', 'Name:', '\n','Jane', 'Doe\nProgram', 'Date:', '3/11/2000', '1:45', 'PM\nMR']), 
           list(['\n[PROV', 'REPORT]\nPerson', 'Name:', '\n','Amy', 'Army\nProgram', 'Date:', '10/1/2000', '11:45', 'AM\nMR'])]


df = pd.DataFrame({'Text' : text,
                          'ID': [1,2,3,4],
                           'P_ID': ['A','B','C','D'],

                         })

df
ID  P_ID    Text
0   1   A   [\n[PROV, REPORT]\nPerson, Name:, \n, John, Dear\nProgram, Date:, 1/11/2000, 10:42, AM\nMR]
1   2   B   [\nToday, Name:, \n, James, Jay\nProgram, Date:, 3/11/2000, 1:45, PM\nmissing]
2   3   C   [\n[NEWS, REPORT]\nPerson, Name:, \n, Jane, Doe\nProgram, Date:, 3/11/2000, 1:45, PM\nMR]
3   4   D   [\n[PROV, REPORT]\nPerson, Name:, \n, Amy, Army\nProgram, Date:, 10/1/2000, 11:45, AM\nMR]

使用以下代码

df['Text'].values

输出

提供以下输出

array([ list(['\n[PROV', 'REPORT]\nPerson', 'Name:', '\n', 'John', 'Dear\nProgram', 'Date:', '1/11/2000', '10:42', 'AM\nMR']),
       list(['\nToday', 'Name:', '\n', 'James', 'Jay\nProgram', 'Date:', '3/11/2000', '1:45', 'PM\nmissing']),
       list(['\n[NEWS', 'REPORT]\nPerson', 'Name:', '\n', 'Jane', 'Doe\nProgram', 'Date:', '3/11/2000', '1:45', 'PM\nMR']),
       list(['\n[PROV', 'REPORT]\nPerson', 'Name:', '\n', 'Amy', 'Army\nProgram', 'Date:', '10/1/2000', '11:45', 'AM\nMR'])], dtype=object)

目标

1）分开\n（使\n[PROV变成\n[PROV，REPORT]\nPerson变成REPORT]\nPerson，Doe\nProgram变成Doe\nProgram等）

2）新建列

尝试过

df['New_Text'] = df['Text'].replace(r'\n',  ' \n ', regex=True) 

所需输出

使用以下代码

df['New_Text'].values

我想要以下输出

array([list(['\n', '[PROV', 'REPORT]', '\n' ,'Person', 'Name:', '\n', 'John', 'Dear', '\n', 'Program', 'Date:', '1/11/2000', '10:42', 'AM', '\n', 'MR']),
       list(['\n', 'Today', 'Name:', '\n', 'James', 'Jay', '\n', 'Program', 'Date:', '3/11/2000', '1:45', 'PM','\n', 'missing']),
       list(['\n', '[NEWS', 'REPORT]','\n', 'Person', 'Name:', '\n', 'Jane', 'Doe', '\n', 'Program', 'Date:', '3/11/2000', '1:45', 'PM', '\n', 'MR']),
       list(['\n', '[PROV', 'REPORT]', '\n', 'Person', 'Name:', '\n', 'Amy', 'Army', '\n', 'Program', 'Date:', '10/1/2000', '11:45', 'AM', '\n', 'MR'])], dtype=object)

问题

我如何实现我想要的输出

Answer 1

奇怪的结构，但可以实现一些mappings，^{}和^{}

import itertools as it
import re

df['split'] = list(map(lambda text: list(filter(lambda b: b, list(it.chain.from_iterable(map(lambda x: re.split(r'(\n)', x), text))))), df.Text))