Python 正则表达式删除字母数字字符而不删除字符串末尾的单词

Question

I'm trying to clean some text by removing alphanumeric characters from the end of the string, but I'm also removing normal words as shown on the output . Can someone help me achieve the expected result?

re.sub(r'[a-zA-Z0-9/]{5,}$', '', text)

asus zenfone 3s max zc521tl
asus zenfone max plus (m1) zb570tl
asus zenfone max pro (m1) zb601kl/zb602k
nokia 3.1 c
nokia 3
asus zenfone 3 zoom ze553k
asus zenfone 3 deluxe zs570kl
blackberry keyone
htc explorer
lg tribute
acer liquid z520

Output:

asus zenfone 3s max 
asus zenfone max plus (m1) 
asus zenfone max pro (m1) 
nokia 3.1 c
nokia 3
asus zenfone 3 zoom 
asus zenfone 3 deluxe 
blackberry 
htc 
lg 
acer liquid z520

Expected output:

asus zenfone 3s max
asus zenfone max plus (m1) 
asus zenfone max pro (m1)
nokia 3.1 c
nokia 3
asus zenfone 3 zoom 
asus zenfone 3 deluxe 
**blackberry keyone**
**htc explorer**
**lg tribute**
acer liquid z520

Answer 1

You can add a positive look-ahead to the regex that requires the word at the end to contain at least one digit for it to be removed: (?=\D*\d) . That will prevent it from removing normal words that don't contain numbers.

The complete program:

#!/usr/bin/env python3
import re

texts = [
    'asus zenfone 3s max zc521tl',
    'asus zenfone max plus (m1) zb570tl',
    'asus zenfone max pro (m1) zb601kl/zb602k',
    'nokia 3.1 c',
    'nokia 3',
    'asus zenfone 3 zoom ze553k',
    'asus zenfone 3 deluxe zs570kl',
    'blackberry keyone',
    'htc explorer',
    'lg tribute',
    'acer liquid z520',
]

for text in texts:
    print(re.sub(r'(?=\D*\d)[a-zA-Z0-9/]{5,}$', '', text))

It outputs:

asus zenfone 3s max 
asus zenfone max plus (m1) 
asus zenfone max pro (m1) 
nokia 3.1 c
nokia 3
asus zenfone 3 zoom 
asus zenfone 3 deluxe 
blackberry keyone
htc explorer
lg tribute
acer liquid z520

Answer 2

If it should be the last word in a string and there are always multiple words, you might use:

[ \t]+(?=[a-zA-Z0-9/]{5})[a-zA-Z/]*[0-9][a-zA-Z0-9/]*[A-Za-z]$

• [ \t]+ Match 1+ spaces or tabs
• (?=[a-zA-Z0-9/]{5}) Assert at least 5 chars of any of the listed
• [a-zA-Z/]* Match 0+ times any of the listed
• [0-9] Match a digit
• [a-zA-Z0-9/]* Match 0+ times any of the listed in the character class
• [A-Za-z] Match a char a-zA-Z
• $ End of string

Regex demo

In the replacement use an empty string.