How to replace all those Special Characters with w

Question

How to replace all those special characters with white spaces in python ?

I have a list of names of a company . . .

Ex:-[myfiles.txt]

MY company.INC

Old Wine pvt

master-minds ltd

"apex-labs ltd"

"India-New corp"

Indo-American pvt/ltd

Here, as per the above example . . . I need all the special characters[-,",/,.] in the file myfiles.txt must be replaced with a single white space and saved into another text file myfiles1.txt.

Can anyone please help me out?

Answer 1

import re
strs = "how much for the maple syrup? $20.99? That's ricidulous!!!"
strs = re.sub(r'[?|$|.|!]',r'',strs) #for remove particular special char
strs = re.sub(r'[^a-zA-Z0-9 ]',r'',strs) #for remove all characters
strs=''.join(c if c not in map(str,range(0,10)) else '' for c in strs) #for remove numbers
strs = re.sub('  ',' ',strs) #for remove extra spaces
print(strs) 

Ans: how much for the maple syrup Thats ricidulous

Answer 2

import string

specials = '-"/.' #etc
trans = string.maketrans(specials, ' '*len(specials))
#for line in file
cleanline = line.translate(trans)

e.g.

>>> line = "Indo-American pvt/ltd"
>>> line.translate(trans)
'Indo American pvt ltd'

Answer 3

Assuming you mean to change everything non-alphanumeric, you can do this on the command line:

cat foo.txt | sed "s/[^A-Za-z0-99]/ /g" > bar.txt

Or in Python with the re module:

import re
original_string = open('foo.txt').read()
new_string = re.sub('[^a-zA-Z0-9\n\.]', ' ', original_string)
open('bar.txt', 'w').write(new_string)

Answer 4

At first i thought to provide a string.maketrans/translate example, but maybe you are using some utf-8 encoded strings and the ord() sorted translate-table will blow in your face, so i thought about another solution:

conversion = '-"/.'
text =  f.read()
newtext = ''
for c in text:
    newtext += ' ' if c in conversion else c

It's not the fastest way, but easy to grasp and modify.

So if your text is non-ascii you could decode conversion and the text-strings to unicode and afterwards reencode in whichever encoding you want to.

Answer 5

While maketrans is the fastes way to do it, I never remerber the syntax. Since speed is rarely an issue and I know regular expression, I would tend to do this:

>>> line = "-[myfiles.txt] MY company.INC"
>>> import re
>>> re.sub(r'[^a-zA-Z0-9]', ' ',line)
'  myfiles txt  MY company INC'

This has the additional benefit of declaring the character you accept instead of the one you reject, which feels easier in this case.

Of couse if you are using non ASCII caracters you'll have to go back to removing the characters you reject. If there are just punctuations sign, you can do:

>>> import string
>>> chars = re.escape(string.punctuation)
>>> re.sub(r'['+chars+']', ' ',line)
'  myfiles txt  MY company INC'

But you'll notice