Simple yet flexible natural sorting in Python.

• Source Code: https://github.com/SethMMorton/natsort
• Downloads: https://pypi.org/project/natsort/
• Documentation: https://natsort.readthedocs.io/
  • Examples and Recipes
  • How Does Natsort Work?
  • API
• Quick Description
• Quick Examples
• FAQ
• Requirements
• Optional Dependencies
• Installation
• How to Run Tests
• How to Build Documentation
• Dropped Deprecated APIs
• History

NOTE: Please see the Dropped Deprecated APIs section for changes.

When you try to sort a list of strings that contain numbers, the normal python sort algorithm sorts lexicographically, so you might not get the results that you expect:

>>> a = ['2 ft 7 in', '1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '7 ft 6 in']
>>> sorted(a)
['1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '2 ft 7 in', '7 ft 6 in']

Notice that it has the order ('1', '10', '2') - this is because the list is being sorted in lexicographical order, which sorts numbers like you would letters (i.e. 'b', 'ba', 'c').

natsort provides a function natsorted() that helps sort lists "naturally" ("naturally" is rather ill-defined, but in general it means sorting based on meaning and not computer code point). Using natsorted() is simple:

>>> from natsort import natsorted
>>> a = ['2 ft 7 in', '1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '7 ft 6 in']
>>> natsorted(a)
['1 ft 5 in', '2 ft 7 in', '2 ft 11 in', '7 ft 6 in', '10 ft 2 in']

natsorted() identifies numbers anywhere in a string and sorts them naturally. Below are some other things you can do with natsort (also see the Examples and Recipes for a quick start guide, or the API for complete details).

Note: natsorted() is designed to be a drop-in replacement for the built-in sorted() function. Like sorted(), natsorted() does not sort in-place. To sort a list and assign the output to the same variable, you must explicitly assign the output to a variable:

>>> a = ['2 ft 7 in', '1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '7 ft 6 in']
>>> natsorted(a)
['1 ft 5 in', '2 ft 7 in', '2 ft 11 in', '7 ft 6 in', '10 ft 2 in']
>>> print(a)  # 'a' was not sorted; "natsorted" simply returned a sorted list
['2 ft 7 in', '1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '7 ft 6 in']
>>> a = natsorted(a)  # Now 'a' will be sorted because the sorted list was assigned to 'a'
>>> print(a)
['1 ft 5 in', '2 ft 7 in', '2 ft 11 in', '7 ft 6 in', '10 ft 2 in']

Please see Generating a Reusable Sorting Key and Sorting In-Place for an alternate way to sort in-place naturally.

• Sorting Versions
• Sort Paths Like My File Browser (e.g. Windows Explorer on Windows)
• Sorting by Real Numbers (i.e. Signed Floats)
• Locale-Aware Sorting (or "Human Sorting")
• Further Customizing Natsort
• Sorting Mixed Types
• Handling Bytes
• Generating a Reusable Sorting Key and Sorting In-Place
• Other Useful Things

natsort does not actually comprehend version numbers. It just so happens that the most common versioning schemes are designed to work with standard natural sorting techniques; these schemes include MAJOR.MINOR, MAJOR.MINOR.PATCH, YEAR.MONTH.DAY. If your data conforms to a scheme like this, then it will work out-of-the-box with natsorted() (as of natsort version >= 4.0.0):

>>> a = ['version-1.9', 'version-2.0', 'version-1.11', 'version-1.10']
>>> natsorted(a)
['version-1.9', 'version-1.10', 'version-1.11', 'version-2.0']

If you need to versions that use a more complicated scheme, please see these version sorting examples.

Prior to natsort version 7.1.0, it was a common request to be able to sort paths like Windows Explorer. As of natsort 7.1.0, the function os_sorted() has been added to provide users the ability to sort in the order that their file browser might sort (e.g Windows Explorer on Windows, Finder on MacOS, Dolphin/Nautilus/Thunar/etc. on Linux).

import os
from natsort import os_sorted
print(os_sorted(os.listdir()))
# The directory sorted like your file browser might show

Output will be different depending on the operating system you are on.

For users not on Windows (e.g. MacOS/Linux) it is strongly recommended to also install PyICU, which will help natsort give results that match most file browsers. If this is not installed, it will fall back on Python's built-in locale module and will give good results for most input, but will give poor results for special characters.

This is useful in scientific data analysis (and was the default behavior of natsorted() for natsort version < 4.0.0). Use the realsorted() function:

>>> from natsort import realsorted, ns
>>> # Note that when interpreting as signed floats, the below numbers are
>>> #            +5.10,                -3.00,            +5.30,              +2.00
>>> a = ['position5.10.data', 'position-3.data', 'position5.3.data', 'position2.data']
>>> natsorted(a)
['position2.data', 'position5.3.data', 'position5.10.data', 'position-3.data']
>>> natsorted(a, alg=ns.REAL)
['position-3.data', 'position2.data', 'position5.10.data', 'position5.3.data']
>>> realsorted(a)  # shortcut for natsorted with alg=ns.REAL
['position-3.data', 'position2.data', 'position5.10.data', 'position5.3.data']

This is where the non-numeric characters are also ordered based on their meaning, not on their ordinal value, and a locale-dependent thousands separator and decimal separator is accounted for in the number. This can be achieved with the humansorted() function:

>>> a = ['Apple', 'apple15', 'Banana', 'apple14,689', 'banana']
>>> natsorted(a)
['Apple', 'Banana', 'apple14,689', 'apple15', 'banana']
>>> import locale
>>> locale.setlocale(locale.LC_ALL, 'en_US.UTF-8')
'en_US.UTF-8'
>>> natsorted(a, alg=ns.LOCALE)
['apple15', 'apple14,689', 'Apple', 'banana', 'Banana']
>>> from natsort import humansorted
>>> humansorted(a)  # shortcut for natsorted with alg=ns.LOCALE
['apple15', 'apple14,689', 'Apple', 'banana', 'Banana']

You may find you need to explicitly set the locale to get this to work (as shown in the example). Please see locale issues and the Optional Dependencies section below before using the humansorted() function.

If you need to combine multiple algorithm modifiers (such as ns.REAL, ns.LOCALE, and ns.IGNORECASE), you can combine the options using the bitwise OR operator (|). For example,

>>> a = ['Apple', 'apple15', 'Banana', 'apple14,689', 'banana']
>>> natsorted(a, alg=ns.REAL | ns.LOCALE | ns.IGNORECASE)
['Apple', 'apple15', 'apple14,689', 'Banana', 'banana']
>>> # The ns enum provides long and short forms for each option.
>>> ns.LOCALE == ns.L
True
>>> # You can also customize the convenience functions, too.
>>> natsorted(a, alg=ns.REAL | ns.LOCALE | ns.IGNORECASE) == realsorted(a, alg=ns.L | ns.IC)
True
>>> natsorted(a, alg=ns.REAL | ns.LOCALE | ns.IGNORECASE) == humansorted(a, alg=ns.R | ns.IC)
True

All of the available customizations can be found in the documentation for the ns enum.

You can also add your own custom transformation functions with the key argument. These can be used with alg if you wish.

>>> a = ['apple2.50', '2.3apple']
>>> natsorted(a, key=lambda x: x.replace('apple', ''), alg=ns.REAL)
['2.3apple', 'apple2.50']

You can mix and match int, float, and str types when you sort:

>>> a = ['4.5', 6, 2.0, '5', 'a']
>>> natsorted(a)
[2.0, '4.5', '5', 6, 'a']
>>> # sorted(a) would raise an "unorderable types" TypeError

natsort does not officially support the bytes type, but convenience functions are provided that help you decode to str first:

>>> from natsort import as_utf8
>>> a = [b'a', 14.0, 'b']
>>> # natsorted(a) would raise a TypeError (bytes() < str())
>>> natsorted(a, key=as_utf8) == [14.0, b'a', 'b']
True
>>> a = [b'a56', b'a5', b'a6', b'a40']
>>> # natsorted(a) would return the same results as sorted(a)
>>> natsorted(a, key=as_utf8) == [b'a5', b'a6', b'a40', b'a56']
True

Under the hood, natsorted() works by generating a custom sorting key using natsort_keygen() and then passes that to the built-in sorted(). You can use the natsort_keygen() function yourself to generate a custom sorting key to sort in-place using the list.sort() method.

>>> from natsort import natsort_keygen
>>> natsort_key = natsort_keygen()
>>> a = ['2 ft 7 in', '1 ft 5 in', '10 ft 2 in', '2 ft 11 in', '7 ft 6 in']
>>> natsorted(a) == sorted(a, key=natsort_key)
True
>>> a.sort(key=natsort_key)
>>> a
['1 ft 5 in', '2 ft 7 in', '2 ft 11 in', '7 ft 6 in', '10 ft 2 in']

All of the algorithm customizations mentioned in the Further Customizing Natsort section can also be applied to natsort_keygen() through the alg keyword option.

• recursively descend into lists of lists
• automatic unicode normalization of input data
• controlling the case-sensitivity
• sorting file paths correctly
• allow custom sorting keys
• accounting for units

How do I debug natsorted()?

The best way to debug natsorted() is to generate a key using natsort_keygen() with the same options being passed to natsorted(). One can take a look at exactly what is being done with their input using this key - it is highly recommended to look at this issue describing how to debug for how to debug, and also to review the How Does Natsort Work? page for why natsort is doing that to your data.

If you are trying to sort custom classes and running into trouble, please take a look at #60. In short, custom classes are not likely to be sorted correctly if one relies on the behavior of __lt__ and the other rich comparison operators in their custom class - it is better to use a key function with natsort, or use the natsort key as part of your rich comparison operator definition.

natsort gave me results I didn't expect, and it's a terrible library!

Did you try to debug using the above advice? If so, and you still cannot figure out the error, then please file an issue.

How does natsort work?

If you don't want to read How Does Natsort Work?, here is a quick primer.

natsort provides a key function that can be passed to list.sort() or sorted() in order to modify the default sorting behavior. This key is generated on-demand with the key generator natsort_keygen(). natsorted() is essentially a wrapper for the following code:

>>> from natsort import natsort_keygen
>>> natsort_key = natsort_keygen()
>>> sorted(['1', '10', '2'], key=natsort_key)
['1', '2', '10']

Users can further customize natsort sorting behavior with the key and/or alg options (see details in the Further Customizing Natsort section).

The key generated by natsort_keygen() always returns a tuple. It does so in the following way (some details omitted for clarity):

1. Assume the input is a string, and attempt to split it into numbers and non-numbers using regular expressions. Numbers are then converted into either int or float.
2. If the above fails because the input is not a string, assume the input is some other sequence (e.g. list or tuple), and recursively apply the key to each element of the sequence.
3. If the above fails because the input is not iterable, assume the input is an int or float, and just return the input in a tuple.

Because a tuple is always returned, a TypeError should not be common unless one tries to do something odd like sort an int against a list.

natsort comes with a shell script called natsort, or can also be called from the command line with python -m natsort. Check out the shell script wiki documentation for more details.

natsort requires Python 3.8 or greater.

The most efficient sorting can occur if you install the fastnumbers package (version >=2.0.0); it helps with the string to number conversions. natsort will still run (efficiently) without the package, but if you need to squeeze out that extra juice it is recommended you include this as a dependency. natsort will not require (or check) that fastnumbers is installed at installation.

It is recommended that you install PyICU if you wish to sort in a locale-dependent manner, see this page on locale issues for an explanation why.

Use pip!

$ pip install natsort

If you want to install the Optional Dependencies, you can use the "extras" notation at installation time to install those dependencies as well - use fast for fastnumbers and icu for PyICU.

# Install both optional dependencies.
$ pip install natsort[fast,icu]
# Install just fastnumbers
$ pip install natsort[fast]

Please note that natsort is NOT set-up to support python setup.py test.

The recommended way to run tests is with tox. After installing tox, running tests is as simple as executing the following in the natsort directory:

$ tox

tox will create virtual a virtual environment for your tests and install all the needed testing requirements for you. You can specify a particular python version with the -e flag, e.g. tox -e py36. Static analysis is done with tox -e flake8. You can see all available testing environments with tox --listenvs.

If you want to build the documentation for natsort, it is recommended to use tox:

$ tox -e docs

This will place the documentation in build/sphinx/html.

In natsort version 6.0.0, the following APIs and functions were removed

• number_type keyword argument (deprecated since 3.4.0)
• signed keyword argument (deprecated since 3.4.0)
• exp keyword argument (deprecated since 3.4.0)
• as_path keyword argument (deprecated since 3.4.0)
• py3_safe keyword argument (deprecated since 3.4.0)
• ns.TYPESAFE (deprecated since version 5.0.0)
• ns.DIGIT (deprecated since version 5.0.0)
• ns.VERSION (deprecated since version 5.0.0)
• versorted() (discouraged since version 4.0.0, officially deprecated since version 5.5.0)
• index_versorted() (discouraged since version 4.0.0, officially deprecated since version 5.5.0)

In general, if you want to determine if you are using deprecated APIs you can run your code with the following flag

$ python -Wdefault::DeprecationWarning my-code.py

By default DeprecationWarnings are not shown, but this will cause them to be shown. Alternatively, you can just set the environment variable PYTHONWARNINGS to "default::DeprecationWarning" and then run your code.

Seth M. Morton

Please visit the changelog on GitHub.