Python – Substituting patterns in text using regex
Last Updated :
29 Dec, 2020
Regular Expression (regex) is meant for pulling out the required information from any text which is based on patterns. They are also widely used for manipulating the pattern-based texts which leads to text preprocessing and are very helpful in implementing digital skills like Natural Language Processing(NLP).
This article demonstrates how to use regex to substitute patterns by providing multiple examples where each example is a unique scenario in its own. It is very necessary to understand the re.sub() method of re (regular expression) module to understand the given solutions.
The re.sub() method performs global search and global replace on the given string. It is used for substituting a specific pattern in the string. There are in total 5 arguments of this function.
Syntax: re.sub(pattern, repl, string, count=0, flags=0)
Parameters:
pattern – the pattern which is to be searched and substituted
repl – the string with which the pattern is to be replaced
string – the name of the variable in which the pattern is stored
count – number of characters up to which substitution will be performed
flags – it is used to modify the meaning of the regex pattern
count and flags are optional arguments.
Example 1: Substitution of a specific text pattern
In this example, a given text pattern will be searched and substituted in a string. The idea is to use the very normal form of the re.sub() method with only the first 3 arguments.
Below is the implementation.
import re
def substitutor():
sentence1 = "It is raining outside."
print(re.sub(r"raining", "sunny", sentence1))
sentence2 = "Thank you very very much."
print(re.sub(r"very", "so", sentence2))
substitutor()
|
Output:
It is sunny outside.
Thank you so so much.
No matter how many time the required pattern is present in the string, the re.sub() function replaces all of them with the given pattern. That’s why both the ‘very’ are replaced by ‘so’ in the above example.
Example 2: Substituting a character set with a specific character
The task is to replace a character set with a given character. A character set means a range of characters. In the re.sub() method a character set is written inside [ ](square brackets).
In this example, the lower case character set i.e., [a-z] will be replaced by the digit 0. Below is the implementation.
import re
def substitutor():
sentence = "22 April is celebrated as Earth Day."
print(re.sub(r"[a-z]", "0", sentence))
substitutor()
|
Output:
22 A0000 00 0000000000 00 E0000 D00.
If there is a need to substitute both lowercase and uppercase character set then we have to introduce the uppercase character set in this way: [a-zA-Z] or the effective way to do is by using flags.
Example 3: Case-insensitive substitution of a character set with a specific character
In this example, both lowercase and uppercase characters will be replaced by the given character. With the use of flags, this task can be carried out very easily.
The re.I flag stands for re.IGNORECASE. By introducing this flag in the re.sub() method and mentioning any one character set i.e., lowercase or uppercase the task can be completed.
Below is the implementation.
import re
def substitutor():
sentence = "22 April is celebrated as Earth Day."
print(re.sub(r"[a-z]", "0", sentence, flags = re.I))
substitutor()
|
Output:
22 00000 00 0000000000 00 00000 000.
Example 4: Perform substitution up to a certain number of character
In this example, substitution will be up to a specific number of characters and not on the whole string. To perform this type of substitution the re.sub() method has an argument count.
By providing a numeric value to this argument, the number of characters on which substitution will occur can be controlled. Below is the implementation.
import re
def substitutor():
sentence = "Follow your Passion."
print(re.sub(r"[a-z]", "0", sentence, 8, flags = re.I))
substitutor()
|
Output:
000000 00ur Passion.
Example 5: Substitution using shorthand character class and preprocessing of text
Regex module provides many shorthand character class for those character sets which are very common during preprocessing of text. Usage of shorthand character class results in writing efficient code and lessen the need to remember the range of every character set.
To get a detail explanation of shorthand character class and how to write regular expression in python for preprocessing of text click here. Following are some of the commonly used shorthand character classes:
\w: matches alpha numeric characters
\W: matches non-alpha numeric characters like @, #, ‘, +, %, –
\d: matches digit characters
\s: matches white space characters
Meaning of some syntax:
adding a plus(+) symbol after a character class or set: repetition of preceding character class or set for at least 1 or more times.
adding an asterisk(*) symbol after a character class or set: repetition of preceding character class or set for at least 0 or more times.
adding a caret(^) symbol before a character class or set: matching position is determined for that character class or set at the beginning of the string.
adding a dollar($) symbol after a character class or set: matching position is determined for that character class or set at the end of the string.
This example demonstrates the use of mentioned shorthand character classes for the substitution and preprocessing of text to get clean and error-free strings. Below is the implementation.
import re
def substitutor():
S = ["2020 Olympic games have @# been cancelled",
"Dr Vikram Sarabhai was +%--the ISRO’s first chairman",
"Dr Abdul Kalam, the father of India's missile programme"]
for i in range(len(S)):
S[i] = re.sub(r"\W", " ", S[i])
S[i] = re.sub(r"\d", " ", S[i])
S[i] = re.sub(r"\s+", " ", S[i])
S[i] = re.sub(r"\s+[a-z]\s+", " ", S[i], flags = re.I)
S[i] = re.sub(r"^\s+", "", S[i])
S[i] = re.sub(r"\s+$", "", S[i])
for i in range(len(S)):
print(S[i])
substitutor()
|
Output:
Olympic games have been cancelled
Dr Vikram Sarabhai was the ISRO first chairman
Dr Abdul Kalam the father of India missile programme
Like Article
Suggest improvement
Share your thoughts in the comments
Please Login to comment...