Remove unicode from string

Question

I'm trying to remove special character from Arabic String using it's Unicode which I got from this link: https://www.fileformat.info/info/unicode/char/0640/index.htm

This is my code:

TATWEEL = u"\u0640"
text = 'الســلام عليكــم'

text.replace(TATWEEL, '')
print(text)

But I tried it and doesn't work (it prints the same string without removing the character)

This is the special character 'ــ'

I'm using Python3

when you say it *doesn't work* what do you mean? Did you receive an error or the pc you were using exploded? — Adelin, Feb 13 '18 at 05:47
Possible duplicate of [Replacing or substituting in a python string does not work](https://stackoverflow.com/questions/15780139/replacing-or-substituting-in-a-python-string-does-not-work) — DYZ, Feb 13 '18 at 06:19

cco · Answer 1 · 2018-02-13T06:02:32.250

8

The replace method of strings does not change the string it is called on; it returns a new string with the specified character replaced.

This code does what you want:

TATWEEL = u"\u0640"
text = 'الســلام عليكــم'

text2 = text.replace(TATWEEL, '')
print(text2)

To get the exact result you expected, use this:

text = text.replace(TATWEEL, '')
print(text)

edited Feb 13 '18 at 06:02

answered Feb 13 '18 at 05:51

cco

2

@user8393084 If it works for you, please accept it as answer. – skjoshi Feb 13 '18 at 05:54

score 1 · Answer 2 · answered Feb 13 '18 at 06:00

1

If text may contain multiple unicode elements then you should go for regex as below:

import re
TATWEEL = u"\u0640"
text = 'الســلام عليكــم'

unicode_removed_text = re.sub(TATWEEL, '', text)

answered Feb 13 '18 at 06:00

Gahan

How is this better than `replace`? – Adelin Feb 13 '18 at 06:02
refer this [ink](https://stackoverflow.com/questions/5668947/use-pythons-string-replace-vs-re-sub); it's very good explanation on where to use replace() and where re.sub() – Gahan Feb 13 '18 at 06:06
1

First answer says *if you can use `replace`, use it* – Adelin Feb 13 '18 at 06:07
I still don't understand the rationale from your post "*If text may contain multiple unicode elements then you should go for regex*" – Adelin Feb 13 '18 at 06:08

2 Answers2