How to write multiple replace in python?

Question

I have a python dictionary of key, value pairs and I want to replace some words in a string which are the keys in the dictionary with their corresponding values.

I have tried some code which are found online.Here is the example:

    test_dict = {'a/a': 'result1', "a/a b/b c/c": "result2"}

    sentence = "<<a/a>> something <<a/a b/b c/c>> something"

    result = multiple_replace(test_dict, sentence)

    def multiple_replace(dict, text):
        regex = re.compile("(%s)" % "|".join(map(re.escape, dict.keys())))
        return regex.sub(lambda mo: dict[mo.string[mo.start():mo.end()]], text)

I expected the result to be <<result1>> something <<result2>> something

The actual output is <<result1>> something <<result1 b/b c/c>> something

Scott Hunter · Accepted Answer · 2019-08-14 12:24:39Z

1

Your code replaced all of the a/a's it found in the string, meaning that there was no longer a a/a b/b c/c to be replaced.

If you surrounded each key w/ << and >>, searched for that, and replaced it putting back the << & >>, you would avoid this problem.

answered Aug 14, 2019 at 12:24

Scott Hunter

50k12 gold badges64 silver badges107 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

fodma1 · Accepted Answer · 2019-08-14 12:25:26Z

The problem is that <<a/a b/b c/c>> will be replaced by 'a/a': 'result1' resulting in "<<result1 b/b c/c>>" before the a/a b/b c/c replacement would happen.

You should do the replace starting with the more specific to the less specific. One way to accomplish this is to use OrderedDict and define your rules in the other direction:

import re
from collections import OrderedDict

test_dict = OrderedDict([("a/a b/b c/c", "result2"), ('a/a', 'result1'),])

sentence = "<<a/a>> something <<a/a b/b c/c>> something"

def multiple_replace(dict, text):
    regex = re.compile("(%s)" % "|".join(map(re.escape, dict.keys())))
    return regex.sub(lambda mo: dict[mo.string[mo.start():mo.end()]], text)

result = multiple_replace(test_dict, sentence)

The output is: <<result1>> something <<result2>> something

Patrick Artner · Accepted Answer · 2019-08-14 12:28:16Z

Your problem is that your first relpaced key 'a/a' is part of another key 'a/a b/b c/c'. The longer key does not replace anymore because the rule for 'a/a' changes the text so it wont find 'a/a b/b c/c' anymore.

You can avoid this if you sort the keys by decreasing length, so longer ones are replaced first:

import re

def multiple_replace(d, text):
    # sort keys by -len so longer ones come first (you could use reverse=True as well)
    regex = re.compile("(%s)" % "|".join(map(re.escape, 
                                             sorted(d.keys(),key=lambda x:-len(x)))))
    return regex.sub(lambda mo: d[mo.string[mo.start():mo.end()]], text)

test_dict = {'a/a': 'result1', "a/a b/b c/c": "result2"}

sentence = "<<a/a>> something <<a/a b/b c/c>> something"

result = multiple_replace(test_dict, sentence)
print(result)

Output:

<<result1>> something <<result2>> something

You will still have problems if the value that was replaced contains part of a shorter key, it will be partially replaced again.

Collectives™ on Stack Overflow

How to write multiple replace in python?

3 Answers 3

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related