Python re.sub() and re.split()

Substitution — the other half of regex

re.sub(pattern, replacement, text) replaces every non-overlapping match of pattern with replacement.

The replacement can be:

A literal string: re.sub(r'\d+', 'NUM', 'a 1 b 22 c 333') → 'a NUM b NUM c NUM'
A pattern with backreferences: r'\1' for group 1, r'\g<name>' for named groups, r'\g<0>' for whole match.
A callable: receives a Match object, returns a string. Most powerful for dynamic logic.

The count parameter

re.sub(pattern, replacement, text, count=N) replaces only the first N matches. Useful for "only fix the first occurrence" or "don't process more than X."

re.subn — get the count back

re.subn returns a tuple (new_string, num_replacements). Useful when you need to know whether anything changed.

re.split — split with a regex separator

re.split(pattern, text) is like str.split but the separator can be any regex. Splits on every match.

If the pattern contains capture groups, the captured text is INCLUDED in the result list. re.split(r'(\W+)', 'hello, world') returns ['hello', ', ', 'world'] — the separator is preserved.

Code

Substitution and split·python

import re

# Simple substitution
re.sub(r'\d+', 'NUM', 'a 1 b 22 c 333')
# 'a NUM b NUM c NUM'

# Backreference in replacement
re.sub(r'(\d{4})-(\d{2})-(\d{2})', r'\2/\3/\1', '2026-05-04')
# '05/04/2026'

# Named group reference
re.sub(r'(?P<year>\d{4})', r'[\g<year>]', 'born 1989, today 2026')
# 'born [1989], today [2026]'

# Callable replacement — full programmatic control
def bracket_if_large(m):
    n = int(m.group())
    return f'[{n}]' if n > 100 else str(n)

re.sub(r'\d+', bracket_if_large, 'small 5 medium 50 large 500')
# 'small 5 medium 50 large [500]'

# subn returns count
re.subn(r'\d+', 'NUM', 'a 1 b 2')
# ('a NUM b NUM', 2)

# split with regex separator
re.split(r'\W+', 'hello, world! how are you?')
# ['hello', 'world', 'how', 'are', 'you', '']

# split preserving the separator
re.split(r'(\W+)', 'hello, world')
# ['hello', ', ', 'world']

Python re.sub() and re.split()

Substitution — the other half of regex

The count parameter

re.subn — get the count back

re.split — split with a regex separator

Code

External links

Exercise

Progress

Comments 0