Python partition vs split: Which One Should You Use?

Python partition vs split comes down to this: str.partition(sep) cuts once at the first separator and always returns a 3-tuple, even when the separator is missing. str.split(sep) returns a list of every piece. Use partition() for one key=value cut you want to unpack safely, and split() when you need all the parts.

What does each method return?

We ran this on Python 3.13.5:

"key=value".partition("=")   # ('key', '=', 'value')
"key=value".split("=", 1)    # ['key', 'value']
"key=a=b".partition("=")     # ('key', '=', 'a=b')
"key=a=b".split("=")         # ['key', 'a', 'b']

partition() returns a tuple of exactly three items: the part before the separator, the separator, and the rest. split() returns a list whose length depends on the input.

What happens when the separator is missing?

This is the biggest difference:

"novalue".partition("=")   # ('novalue', '', '')
"novalue".split("=")       # ['novalue']
k, _, v = "novalue".partition("=")   # works: k='novalue', v=''
k, v = "novalue".split("=", 1)
# ValueError: not enough values to unpack (expected 2, got 1)

With partition() you can always unpack three names, and an empty middle value tells you the separator was not found. With split() you have to check the length first or catch the error.

How do rpartition and rsplit differ?

They cut from the right:

"key=a=b".rpartition("=")   # ('key=a', '=', 'b')
"key=a=b".rsplit("=", 1)    # ['key=a', 'b']
"novalue".rpartition("=")   # ('', '', 'novalue')

Note the flip: when the separator is missing, rpartition() puts the whole string last, not first.

Which edge cases surprise people?

• "=x".partition("=") gives ('', '=', 'x') and "x=".partition("=") gives ('x', '=', '').
• "".partition("=") gives ('', '', ''), while "".split("=") gives [''].
• An empty separator fails for both: ValueError: empty separator.
• partition() needs an argument. Calling it bare raises TypeError: str.partition() takes exactly one argument (0 given).
• Only split() has a no-argument mode: " a b ".split() gives ['a', 'b'], but .split(" ") gives ['', '', 'a', '', 'b', ''].
• Both work on bytes: b"k=v".partition(b"=") gives (b'k', b'=', b'v').

Is partition faster than split?

Slightly, in our runs. With s = "user@example.com":

The gap gets big when there are many separators and you only need the first cut. On "a=" repeated 1,000 times, partition("=") took 58.7 ns, split("=", 1) took 65.9 ns, and a plain split("=") took 4,640.9 ns because it builds a list of 1,001 pieces.

Which one should you use?

• One cut, like key=value, user@host or a header line: partition(), or split(sep, 1) if you want a list.
• Last cut, like a file extension: rpartition().
• Every piece, like CSV-ish fields or words: split().
• Whitespace-separated words: split() with no argument.

FAQ

Does partition() support a maxsplit?No. It always cuts exactly once. Use split(sep, maxsplit) for more control.

Does partition() accept a regex?No. Both methods take a plain string. Use re.split() for patterns.

Is partition() in Python 2?Yes, it dates back to Python 2.5. Our tests ran on Python 3.13.5.

Related

• removeprefix vs lstrip in Python
• Python isdigit vs isnumeric
• csv.reader vs split lab

Sources

• 01docs.python.org/3/library/stdtypes.html
• 02docs.python.org/3/library/stdtypes.html
• 03docs.python.org/3/library/stdtypes.html