25

What is the best way to remove all characters from a string that are not in the alphabet? I mean, remove all spaces, interpunction, brackets, numbers, mathematical operators..

For example:

input: 'as32{ vd"s k!+'
output: 'asvdsk'

5 Answers5

53

You could use re, but you don't really need to.

>>> s = 'as32{ vd"s k!+'
>>> ''.join(x for x in s if x.isalpha())
'asvdsk'    
>>> filter(str.isalpha, s) # works in python-2.7
'asvdsk'
>>> ''.join(filter(str.isalpha, s)) # works in python3
'asvdsk'
timgeb
  • 73,231
  • 20
  • 109
  • 138
29

If you want to use regular expression, This should be quicker

import re
s = 'as32{ vd"s k!+'
print re.sub('[^a-zA-Z]+', '', s)

prints 'asvdsk'

nehem
  • 10,894
  • 6
  • 51
  • 76
3

Here is a method that uses ASCII ranges to check whether an character is in the upper/lower case alphabet (and appends it to a string if it is):

s = 'as32{ vd"s k!+'
sfiltered = ''

for char in s:
    if((ord(char) >= 97 and ord(char) <= 122) or (ord(char) >= 65 and ord(char) <= 90)):
        sfiltered += char

The variable sfiltered will show the result, which is 'asvdsk' as expected.

Patrick Yu
  • 912
  • 1
  • 7
  • 19
0

This simple expression get all letters, including non ASCII letters ok t áàãéèêçĉ... and many more used in several languages.

r"[^\W\d]+"

It means "get a sequence of one or more characters that are not either "non word characters" or a digit.

plpsanchez
  • 175
  • 2
  • 9
0

If you'd like to preserve characters like áàãéèêçĉ that are used in many languages around thw world, try this:

import re
print re.sub('[\W\d_]+', yourString)