255

What's the easiest way to do a case-insensitive string replacement in Python?

Improve this question

11 Answers 11

Reset to default
294

The string type doesn't support this. You're probably best off using the regular expression sub method with the re.IGNORECASE option.

>>> import re >>> insensitive_hippo = re.compile(re.escape('hippo'), re.IGNORECASE) >>> insensitive_hippo.sub('giraffe', 'I want a hIPpo for my birthday') 'I want a giraffe for my birthday'
Improve this answer
Sign up to request clarification or add additional context in comments.

6 Comments

If you're only doing a single replace, or want to save lines of code, it's more efficient to use a single substitution with re.sub and the (?i) flag: re.sub('(?i)' + re.escape('hippo'), 'giraffe', 'I want a hIPpo for my birthday')
Why re.escape for a string of letters only? Thanks.
@Elena, it's not needed for 'hippo', but would be useful if the to-replace value was passed into a function, so it's really more of a good example than anything else.
Well, I think it's a great example, because I use 'hippo' as a metasyntactic variable sometimes, too. 🦛 But I thought the Dandelion Wine song was I want a hippopotamus for Christmas, not 'my birthday'? 🙂
Besides having to re.escape your needle, there's another trap here which this answer fails to avoid, noted in stackoverflow.com/a/15831118/1709587: since re.sub processes escape sequences, as noted in docs.python.org/library/re.html#re.sub, you need to either escape all backslashes in your replacement string or use a lambda.
This doesn't work for replacing r'A\BC' with r'D\EF' in r'xxxA\BCxxxA\BCxxx') - The correct answer is blow, the one from johv
117 
import re pattern = re.compile("hello", re.IGNORECASE) pattern.sub("bye", "hello HeLLo HELLO") # 'bye bye bye'
Improve this answer

3 Comments

Or one-liner: re.sub('hello', 'bye', 'hello HeLLo HELLO', flags=re.IGNORECASE)
Note that re.sub only supports this flag since Python 2.7.
some times it return random strings
63

In a single line:

import re re.sub("(?i)hello","bye", "hello HeLLo HELLO") #'bye bye bye' re.sub("(?i)he\.llo","bye", "he.llo He.LLo HE.LLO") #'bye bye bye'

Or, use the optional "flags" argument:

import re re.sub("hello", "bye", "hello HeLLo HELLO", flags=re.I) #'bye bye bye' re.sub("he\.llo", "bye", "he.llo He.LLo HE.LLO", flags=re.I) #'bye bye bye'
Improve this answer

Comments

25

Continuing on bFloch's answer, this function will change not one, but all occurrences of old with new - in a case insensitive fashion. 

def ireplace(old, new, text): idx = 0 while idx < len(text): index_l = text.lower().find(old.lower(), idx) if index_l == -1: return text text = text[:index_l] + new + text[index_l + len(old):] idx = index_l + len(new) return text
Improve this answer

4 Comments

Very well done. Much better than regex; it handles all kinds of characters, whereas the regex is very fussy about anything non-alphanumeric. Preferred answer IMHO.
All you have to do is escape the regex: the accepted answer is much shorter and easier to read than this.
Escape only works for matching, backslashes in the destination can mess things up still.
Possibly the fastest method for a case-insensitive replace, tested against both using an arrayed string and using regex.
11

Like Blair Conrad says string.replace doesn't support this.

Use the regex re.sub, but remember to escape the replacement string first. Note that there's no flags-option in 2.6 for re.sub, so you'll have to use the embedded modifier '(?i)' (or a RE-object, see Blair Conrad's answer). Also, another pitfall is that sub will process backslash escapes in the replacement text, if a string is given. To avoid this one can instead pass in a lambda.

Here's a function:

import re def ireplace(old, repl, text): return re.sub('(?i)'+re.escape(old), lambda m: repl, text) >>> ireplace('hippo?', 'giraffe!?', 'You want a hiPPO?') 'You want a giraffe!?' >>> ireplace(r'[binfolder]', r'C:\Temp\bin', r'[BinFolder]\test.exe') 'C:\\Temp\\bin\\test.exe'
Improve this answer

Comments

7

This doesn't require RegularExp

def ireplace(old, new, text): """ Replace case insensitive Raises ValueError if string not found """ index_l = text.lower().index(old.lower()) return text[:index_l] + new + text[index_l + len(old):]
Improve this answer

3 Comments

Good one, however this does not change all occurrences of old with new, but only the first occurrence.
It's less readable than the regex version. No need to reinvent the wheel here.
It would be interesting to do a performance comparison between this and the upvoted versions, it might be faster, which matters for some applications. Or it might be slower because it does more work in interpreted Python.
7

An interesting observation about syntax details and options:

# Python 3.7.2 (tags/v3.7.2:9a3ffc0492, Dec 23 2018, 23:09:28) [MSC v.1916 64 bit (AMD64)] on win32 >>> import re >>> old = "TREEROOT treeroot TREerOot" >>> re.sub(r'(?i)treeroot', 'grassroot', old) 'grassroot grassroot grassroot' >>> re.sub(r'treeroot', 'grassroot', old) 'TREEROOT grassroot TREerOot' >>> re.sub(r'treeroot', 'grassroot', old, flags=re.I) 'grassroot grassroot grassroot' >>> re.sub(r'treeroot', 'grassroot', old, re.I) 'TREEROOT grassroot TREerOot'

Using the (?i) prefix in the match expression or adding flags=re.I as a fourth argument will result in a case-insensitive match - however using just re.I as the fourth argument does not result in case-insensitive match.

For comparison:

>>> re.findall(r'treeroot', old, re.I) ['TREEROOT', 'treeroot', 'TREerOot'] >>> re.findall(r'treeroot', old) ['treeroot']
Improve this answer

2 Comments

This does not provide an answer to the question. please edit your answer to ensure that it improves upon other answers already present in this question.
From the re.sub docs it 5 parameters: re.sub(pattern, repl, string, count=0, flags=0) which is why flags=re.I works but trying to pass it as a positional parameter fails, it's in the wrong position.
7

This function uses both the str.replace() and re.findall() functions. It will replace all occurences of pattern in string with repl in a case-insensitive way.

def replace_all(pattern, repl, string) -> str: occurences = re.findall(pattern, string, re.IGNORECASE) for occurence in occurences: string = string.replace(occurence, repl) return string
Improve this answer

Comments

1

I was having \t being converted to the escape sequences (scroll a bit down), so I noted that re.sub converts backslashed escaped characters to escape sequences.

To prevent that I wrote the following:

Replace case insensitive.

import re def ireplace(findtxt, replacetxt, data): return replacetxt.join( re.compile(findtxt, flags=re.I).split(data) )

Also, if you want it to replace with the escape characters, like the other answers here that are getting the special meaning bashslash characters converted to escape sequences, just decode your find and, or replace string. In Python 3, might have to do something like .decode("unicode_escape") # python3

findtxt = findtxt.decode('string_escape') # python2 replacetxt = replacetxt.decode('string_escape') # python2 data = ireplace(findtxt, replacetxt, data)

Tested in Python 2.7.8

Improve this answer

Comments

1 
i='I want a hIPpo for my birthday' key='hippo' swp='giraffe' o=(i.lower().split(key)) c=0 p=0 for w in o: o[c]=i[p:p+len(w)] p=p+len(key+w) c+=1 print(swp.join(o))
Improve this answer

3 Comments

For learning: generally when you do a search and replace on a string, it's better to not have to turn it into an array first. That's why the first answer is probably the best. While it's using an external module, it's treating the string as one whole string. It's also a bit clearer what's happening in the process.
For learning: its very difficult for a developer with no context to read this code and decipher what its doing :)
Any code that has counter++ is bad in general.
1

1 line simple solution without imports :-)

words = 'GREETINGS from EGYPT. GreeTings from Cairo' replace_what, replace_with, = 'Greetings', 'Hello' result = ' '.join([replace_with if word.lower() == replace_what.lower() else word for word in words.split(' ')]) print (result)

The result is:

Hello from EGYPT. Hello from Cairo
Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.