在 Python 3.6 中替换介于 (') 之间的字符串中的子字符串

Question

在数据框中有一列是一系列消息。每个字符串都有需要替换的不同长度的“字符”块。它看起来像这样：

Index  Message    
0      The value '0000' is not valid.
1      The value 'GA1234' for <category> must be changed.
2      The value '.30' is not a valid entry.
3      The value 'Invitation for Party' is not valid.
4      On line '0010', the value 'CARD' is invalid.

不确定如何替换 (') 之间的系列中的所有字符串。

试过：df.Message.str.replace("The value '%'", 'The value XXXXX') 但没有运气。

试过：df.Message.str.find(" ' ") 它给了我第一个撇号的位置，但不确定如何找到第二个撇号。

有没有办法可以将两者之间的子字符串 (') 替换为 'XXXXX'？

Answer 1

Series.str.replace() 接受正则表达式：

In [184]: df
Out[184]:
   Index                                            Message
0      0                     The value '0000' is not valid.
1      1  The value 'GA1234' for <category> must be chan...
2      2              The value '.30' is not a valid entry.
3      3     The value 'Invitation for Party' is not valid.
4      4       On line '0010', the value 'CARD' is invalid.

In [185]: df['Message'] = df['Message'].str.replace(r"'.*?'", 'XXXXX')

In [186]: df
Out[186]:
   Index                                          Message
0      0                    The value XXXXX is not valid.
1      1  The value XXXXX for <category> must be changed.
2      2            The value XXXXX is not a valid entry.
3      3                    The value XXXXX is not valid.
4      4       On line XXXXX, the value XXXXX is invalid.

Answer 2

使用正则表达式：

import re

pattern = re.compile(r"'[^']+'")  # 1 or more characters between single quotes
result = pattern.sub('XXXXX', inputtext)

上面只接受两个单引号，中间有一个或多个字符（它们本身不是单引号）。

请参阅 Python 文档中包含的Regular Expression HOWTO，了解如何使用此类模式的良好介绍。

演示：

>>> import re
>>> pattern = re.compile(r"'[^']+'")
>>> pattern.sub('XXXXX', "The value '0000' is not valid.")
'The value XXXXX is not valid.'
>>> pattern.sub('XXXXX', "The value 'GA1234' for <category> must be changed.")
'The value XXXXX for <category> must be changed.'
>>> pattern.sub('XXXXX', "The value '.30' is not a valid entry.")
'The value XXXXX is not a valid entry.'
>>> pattern.sub('XXXXX', "The value 'Invitation for Party' is not valid.")
'The value XXXXX is not valid.'
>>> pattern.sub('XXXXX', "On line '0010', the value 'CARD' is invalid.")
'On line XXXXX, the value XXXXX is invalid.'

如果您有熊猫系列，请使用Series.replace()；如果您设置regex=True，它也需要正则表达式，本质上充当re.sub() 调用系列中的所有值：

df.Message.replace(r"'[^']+'", 'XXXXX', regex=True)

演示：

>>> import pandas as df
>>> df = pd.DataFrame({
...     'Message': pd.Series([
...         "The value '0000' is not valid.",
...         "The value 'GA1234' for <category> must be changed.",
...         "The value '.30' is not a valid entry.",
...         "The value 'Invitation for Party' is not valid.",
...         "On line '0010', the value 'CARD' is invalid.",
...     ])
... })
>>> df.Message.replace(r"'[^']+'", 'XXXXX', regex=True)
0                      The value XXXXX is not valid.
1    The value XXXXX for <category> must be changed.
2              The value XXXXX is not a valid entry.
3                      The value XXXXX is not valid.
4         On line XXXXX, the value XXXXX is invalid.
Name: Message, dtype: object