Python's split() has more gotchas than you think

Akshit — Tue, 06 Oct 2026 15:16:49 +0000

Splitting strings looks trivial until an empty string or an extra space breaks your parser.

1. `split()` vs `split(' ')` are not the same

"a  b".split()     # ['a', 'b']
"a  b".split(' ')  # ['a', '', 'b']

With no argument, split() splits on any whitespace run and drops empty strings. With an explicit separator, every occurrence counts, so consecutive separators produce empty strings.

2. Empty input behaves differently too

"".split()     # []
"".split(',')  # ['']   <- a list with one item, not an empty list

If you read CSV-like fields, this is where an empty field becomes a phantom row.

3. Limit splits with `maxsplit`

"key=value=with=equals".split('=', 1)
# ['key', 'value=with=equals']

4. Split from the right with `rsplit`

"scene.final.blend".rsplit('.', 1)
# ['scene.final', 'blend']

Useful for file extensions. For real paths, prefer os.path.splitext or pathlib.Path.suffix.

5. Use `partition` when you expect exactly one separator

host, _, port = "localhost:8080".partition(':')
# host='localhost', port='8080'

It always returns three items, even if the separator is missing (('localhost', '', '')), so unpacking never raises ValueError.

6. Use `splitlines()` instead of `split('\n')`

"a\r\nb".split('\n')   # ['a\r', 'b']
"a\r\nb".splitlines()  # ['a', 'b']

It handles \r\n, \r, and other line endings, so it's safer for files created on Windows.

7. Multiple delimiters? Use `re.split`

import re

re.split(r'[,;\s]+', "a, b;c  d")
# ['a', 'b', 'c', 'd']

Rule of thumb

split() for whitespace
split(sep, n) or partition() for structured strings
splitlines() for text files
re.split() when there's more than one kind of separator

DEV Community: Akshit

Python's split() has more gotchas than you think

1. split() vs split(' ') are not the same

2. Empty input behaves differently too

3. Limit splits with maxsplit

4. Split from the right with rsplit

5. Use partition when you expect exactly one separator

6. Use splitlines() instead of split('\n')