python - 如何通过定义分隔符前后来提取子字符串

Question

我有包含 URL 的数据框，我想在两者之间提取一些东西。

df
    URL
    https://storage.com/vision/Glass2020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg
    https://storage.com/vision/Carpet5020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg
    https://storage.com/vision/Metal8020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg

期望的输出是这样的

            URL                                                                           Type
 https://storage.com/vision/Glass2020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg      Glass2020
 https://storage.com/vision/Carpet5020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg     Carpet5020
 https://storage.com/vision/Metal8020/2020-02-04_B8I8FZHl-xJ_2236301468348443721.jpg      Metal8020

我会使用 df['URL'].str.extract 但要了解如何定义定界符前后。

Answer 1

一个想法是使用 Series.str.split通过索引选择倒数第二个值:

df['Type'] = df['URL'].str.split('/').str[-2]
print (df)
                                                 URL        Type
0  https://storage.com/vision/Glass2020/2020-02-0...   Glass2020
1  https://storage.com/vision/Carpet5020/2020-02-...  Carpet5020
2  https://storage.com/vision/Metal8020/2020-02-0...   Metal8020

编辑:要在预期输出之外指定不同的值，请使用 Series.str.extract :

df['Type'] = df['URL'].str.extract('vision/(.+)/2020')
print (df)
                                                 URL        Type
0  https://storage.com/vision/Glass2020/2020-02-0...   Glass2020
1  https://storage.com/vision/Carpet5020/2020-02-...  Carpet5020
2  https://storage.com/vision/Metal8020/2020-02-0...   Metal8020