In this article, we will discuss how to replace only the duplicate occurrences of certain words in a string that is, replace a word from its second occurrence onwards, while keeping its first occurrence unchanged.

Example:

Input: Gfg is best. Gfg also has Classes now. Classes help understand better.
Output: Gfg is best. It also has Classes now. They help understand better.

Using list comprehension + set

This is the cleanest and most efficient approach. We use a set to keep track of seen words and a list comprehension to perform replacements in a single line.

Steps:

1. Split the string into individual words.
2. Iterate over each word.
3. If it exists in the replacement dictionary and hasn’t been seen yet, keep it as-is.
4. If it appears again, replace it with the mapped value.
5. Join the modified words back into a single string.

s1='Gfg is best . Gfg also has Classes now. Classes help understand better .'rep={'Gfg':'It','Classes':'They'}seen=set()res=[rep[word]ifwordinrepandwordinseenelse(seen.add(word)orword)forwordint1.split()]s2=' '.join(res)print(s2)

Gfg is best . It also has Classes now. They help understand better .

This approach uses regex patterns to detect repeated occurrences of target words and replaces them usingre.sub().

Steps:

1. Import the re module.
2. Define a regex pattern to match the target words from the replacement dictionary.
3. Use a function in re.sub() to perform conditional replacements based on previous occurrences.

importres1='Gfg is best . Gfg also has Classes now. Classes help understand better .'d={'Gfg':'It','Classes':'They'}pattern=r'\b('+'|'.join(re.escape(k)forkind.keys())+r')\b'seen=set()deffun(m):word=m.group(1)ifwordinseen:returnd[word]seen.add(word)returnwordres=re.sub(pattern,fun,s1)print(res)

Gfg is best . It also has Classes now. They help understand better .

Explanation:

• \b(...)\b:matches complete words only.
• re.escape(): safely handles special regex characters in keys.
• re.sub(pattern, fun, s1): calls fun() for each match and replaces only later duplicates using a seen set.

Using split() + enumerate() + Loop

This is a more explicit and beginner-friendly approach using loops and sets. It manually iterates through the words, tracks first occurrences, and replaces duplicates.

s1='Gfg is best . Gfg also has Classes now. Classes help understand better .'rep={'Gfg':'It','Classes':'They'}words=s1.split()seen=set()fori,wordinenumerate(words):ifwordinrep:ifwordinseen:words[i]=rep[word]else:seen.add(word)res=' '.join(words)print(res)

Gfg is best . It also has Classes now. They help understand better .

Explanation:

• s1.split():splits text into tokens.
• if word in rep and word in seen:replaces repeat occurrences using mapping.
• seen.add(word):records first encounter.

Using keys() + index() + List Comprehension

This method uses the list.index() function inside a list comprehension to find if a word has appeared before. It replaces only the repeated words and keeps the first one as it is.

s1='Gfg is best . Gfg also has Classes now. Classes help understand better .'d={'Gfg':'It','Classes':'They'}words=s1.split()res=' '.join([d.get(word)ifwordindandwords.index(word)!=ielsewordfori,wordinenumerate(words)])print(res)

Gfg is best . It also has Classes now. They help understand better .

Explanation:

• rep.get(word): retrieves replacement from dictionary.
• words.index(word) != i: ensures only later occurrences are replaced.

 Related Articles:

• Python | Replace multiple occurrence of character by single
• Remove all consecutive duplicates from a string