desensitize for chinese policy (5 features:name, id, phone, address, card)

1.Using it in full text process

clean() function would do mask job It include two parameters:

• text: you processed text
• replace_with: the replace type:include identifier and compliance

by default, it would process card,cnaddress,cnid,cnname,credential,email,name,phone,url mode, you can use add_detector() or remove_detector() to add or remove mode in it.

import desensitize
de=desensitize.Desensitize()
#de.remove_detector('email')
text=u"13725557496 contact Joe Duffy at [email protected] 370304197709200630 4401250222189922 王猛住在上海市陆家嘴汤臣一品"
de.clean(text)
#['', 'a16a00b1f7d5e4e1b20a5b7517e17463', ' contact ', 'Joe', ' ', 'Duffy', ' at ', '[email protected]', ' ', '370***********0630', ' ', '************9922', ' ', '王*', '住在', '上海市***', '汤臣一品']

2.Using it in csv process

import pandas as pd
df=pd.read_csv("test.csv")
df = df.applymap(str)
pd.save_csv("test1.csv")