在python

Question

我有一个数据框，如下所示：

 varA                   json_data
 'string1'      {"str":{"str":"string","str":12345,"str":"str","str":"str","xyz":"1234","zyx":"string1","str":[{"str":"str","str":"str"}],"str":["str","str"],"str":["str"]},"str":"str"}

 'string2'      {"str":{"str":"string","str":12345,"str":"str","str":"str","xyz":"4567","zyx":"string2","str":[{"str":"str","str":"str"}],"str":["str","str"],"str":["str"]},"str":"str"}

 'string3'      {"str":{"str":"string","str":12345,"str":"str","str":"str","xyz":"8910","zyx":"string3","str":[{"str":"str","str":"str"}],"str":["str","str"],"str":["str"]},"str":"str"}

我需要创建新列，并将从json_data中提取的数字填充到第一个，＆＃34; string1＆＃34;，＆＃34; string2＆＃34;和＆＃34; string3＆＃34;值到第二个新列。结果数据集如下所示：

   varA        json_data    xyz   zyx
 'string'        ...       1234  string1
 'string'        ...       4567  string2
 'string'        ...       8910  string3

我的代码如下：

df['xyz'] = df['json_data'].str.split('xyz":')[1].split('","zyx')[0]

然而，我收到一个错误：

AttributeError: 'list' object has no attribute 'split'

我该如何解决？还有其他选择吗？