Does Python 3 string ordering depend on locale?

Question

Do Python's str.__lt__ or sorted order characters based on their unicode index or by some locale-dependent collation rules?

possible duplicate of [String Comparison Technique Used by Python](http://stackoverflow.com/questions/4806911/string-comparison-technique-used-by-python) — Adriano Repetti, Oct 22 '14 at 10:50
@AdrianoRepetti, the question you linked seems to be more about the principle of lexicographic ordering. — Aivar, Oct 22 '14 at 11:03
no, it's same question you asked (does "<" performs culture aware comparison?) — Adriano Repetti, Oct 22 '14 at 11:08

Martijn Pieters · Accepted Answer · 2014-10-22T11:03:42.010

9

No, string ordering does not take locale into account. It is based entirely on the Unicode codepoint sort order.

The locale module does provide you with a locale.strxform() function that can be used for locale-specific sorting:

import locale

sorted(list_of_strings, key=locale.strxfrm)

This tool is quite limited; for any serious collation task you probably want to use the PyICU library:

import PyICU

collator = PyICU.Collator.createInstance(PyICU.Locale(locale_spec))
sorted(list_of_strings, key=collator.getSortKey)

edited Oct 22 '14 at 11:03

answered Oct 22 '14 at 10:48

Martijn Pieters

1

[it may fail](http://stackoverflow.com/q/3412933/4279), [PyICU could be used instead](http://stackoverflow.com/a/16701346/4279) – jfs Oct 22 '14 at 10:57
@J.F.Sebastian: agreed; added a reference. – Martijn Pieters Oct 22 '14 at 11:03

1 Answers1