• • 测试 Dataframe 开始：**

我们在应用regex函数之前构建以下 Dataframe （我假设结束日期总是在开始日期之前）：

import pandas as pd
import re

### We build dataframe test first ###
s = "Pivot (Triage) Form Entered On:  12/30/2022 23:20 EST    Performed On:  12/30/2022 23:16 EST"

df = pd.DataFrame([('1', s)], columns=['ID', 'RESULT'])

### ----------------------------- ###

ID                                             RESULT
0  1  Pivot (Triage) Form Entered On:  12/30/2022 23...

您可以在代码中使用下面的regex，或者使用下面的代码（它最适合您）regex = r'\d{2}\/\d{2}\/\d{4} \d{2}:\d{2}'

• • 使用您的脚本：**

import pandas as pd
import re

### We build dataframe test first ###
s = "Pivot (Triage) Form Entered On:  12/30/2022 23:20 EST    Performed On:  12/30/2022 23:16 EST"

df = pd.DataFrame([('1', s)], columns=['ID', 'RESULT'])

### ----------------------------- ###
# We define regex
regex = r'Form Entered On:  (\d{1,2}\/\d{1,2}\/\d{4} \d{1,2}:\d{1,2}) EST'
df['END_TIME'] = df['RESULT'].str.extract(regex)
regex = r'Performed On:  (\d{1,2}\/\d{1,2}\/\d{4} \d{1,2}:\d{1,2}) EST'
df['START_TIME'] = df['RESULT'].str.extract(regex)

• • 另一种方式：**

import pandas as pd
import re

### We build dataframe test first ###
s = "Pivot (Triage) Form Entered On:  12/30/2022 23:20 EST    Performed On:  12/30/2022 23:16 EST"

df = pd.DataFrame([('1', s)], columns=['ID', 'RESULT'])

### ----------------------------- ###
# We define regex
regex = r'\d{1,2}\/\d{1,2}\/\d{4} \d{1,2}:\d{1,2}'

df[['END_TIME', 'START_TIME']] = df.apply(lambda x: re.findall(regex, x.iloc[1]), axis=1).iloc[0]

• • df ["结束时间"]：**

0    12/30/2022 23:20
Name: END_TIME, dtype: object

• • df ["开始时间"]：**

0    12/30/2022 23:16
Name: START_TIME, dtype: object

假设总是有2个且只有2个时间戳，并使用更通用的正则表达式模式，我们可以尝试：

test[["END_TIME", "START_TIME"]] = test["RESULT"].str.extract(r'Entered On:\s*(\d{1,2}/\d{1,2}/\d{4} \d{1,2}:\d{1,2}) [A-Z]{3}\s+Performed On:\s*(\d{1,2}/\d{1,2}/\d{4} \d{1,2}:\d{1,2}) [A-Z]{3}')

以下regex demo显示regex模式和捕获组工作正常。