Find and replace text with Python

Tips of Python programming.

Basic find and replace

Search and replace text in Python is simple, you can find a specific string with find() or index() method, it will return the index of first match occasion.

>>> s = 'Cat and Dog'
>>> s.find('Dog')
8
>>> s.index('Dog')
8
>>> s.find('Duck')
-1

To replace Cat to Dog, you can simply call replace() method.

>>> s = 'Cat and Dog'
>>> s.replace('Cat', 'Dog')
'Dog and Dog'

Wildcards matching

So how about searching string with wildcards pattern? You should try fnmatch library, it is built-in python.

>>> s = 'Cat and Dog'
>>> import fnmatch
>>> fnmatch.fnmatch(s,'Cat*')
True
>>> fnmatch.fnmatch(s,'C*and*D?')
False
>>> fnmatch.fnmatch(s,'C*and*D*')
True

Regex find and replace

To use advanced text search and replacement, regular expression is your best friend. To find string with pattern, here is an example:

>>> import re
>>> s = 'We will fly to Thailand on 2016/10/31'
>>> pattern = r'\d+'
>>> re.findall(pattern, s)
['2016', '10', '31']
>>> re.search(pattern, s)
<_sre.SRE_Match object at 0x03A8FD40>
>>> re.search(pattern, s).group()
'2016'

To replace string with pattern, hmm, it is an advanced feature, you might want to try re.sub() function(sub => substitution).

>>> s = "I like {color} car."
>>> re.sub(r'\{color\}','blue',s)
'I like blue car.'

>>> s = 'We will fly to Thailand on 10/31/2016'
>>> re.sub('(\d+)/(\d+)/(\d+)', r'\3-\1-\2', s)
'We will fly to Thailand on 2016-10-31'

The re.sub() function is really powerful, in above example, {color} is a pattern that might be updated when string finally published. You can create pattern like this as a template. And r'\3-\1-\2' is the reference to regex matching groups.

Let’s see another example:

s = "Tom is talking to Jerry."
name1 = "Tom"
name2 = "Jerry"

pattern = r'(.*)({0})(.*)({1})(.*)'.format(name1, name2)
print re.sub(pattern, r'\1\4\3\2\5', s)
# Jerry is talking to Tom.

Let’s see how to customize the replace function.

def change_date(m):
    from calendar import month_abbr
    mon_name = month_abbr[int(m.group(1))]
    return '{} {} {}'.format(m.group(2), mon_name, m.group(3))

s = 'We will fly to Thailand on 10/31/2016'
pattern = r'(\d+)/(\d+)/(\d+)'
print re.sub(pattern, change_date, s)
# We will fly to Thailand on 31 Oct 2016

OK, the ultimate example goes here. Hope you enjoy :)

def match_case(word):
    def replace(m):
        text = m.group()
        if text.isupper():
            return word.upper()
        elif text.islower():
            return word.lower()
        elif text[0].isupper():
            return word.capitalize()
        else:
            return word
    return replace

s = "LOVE PYTHON, love python, Love Python"
print re.sub('python', match_case('money'), s, flags=re.IGNORECASE)
# LOVE MONEY, love money, Love Money

Summary

Oh, last but not least, do you want to do use re.sub() for wildcards, yes, you can do it! fnmatch provide a function to let you translate wildcards pattern into regular expression pattern.

>>> fnmatch.translate('C*and*D*')
'C.*and.*D.*'