sort Persian strings for python

Question

I want to use sort function of python but it does not work well.

sorted( ['ا', 'ب', 'پ', 'ح', 'س', 'ص', 'ف', 'ک', 'ک', 'ک', 'م', 'م']) = 
 ['ا', 'ب', 'ح', 'س', 'ص', 'ف', 'م', 'م', 'پ', 'ک', 'ک', 'ک']

Could you be more specific about "does not work well"? What is the error? What is the expected output? — Dani Mesejo, Jan 31 '19 at 10:45
in persian character order پ is after ب but here is false. the problem is it use arabic order that is written in unicode. — majid lesani, Jan 31 '19 at 10:48
You will probably need to set a LOCALE to sort stuff. Try the suggestions here- https://stackoverflow.com/questions/11121636/sorting-list-of-string-with-specific-locale-in-python — bhathiya-perera, Jan 31 '19 at 10:53

score 15 · Accepted Answer · edited Jun 22 '22 at 09:26

15

try using PyICU:

import icu
collator = icu.Collator.Collator.createInstance(icu.Locale('fa_IR.UTF-8'))
print ([i for i in sorted(['ا', 'ب', 'پ', 'ح', 'س', 'ص', 'ف', 'ک', 'ک', 'ک', 'م', 'م'], key=collator.getSortKey)])

edited Jun 22 '22 at 09:26

Karl Knechtel

62,466
11
102
153

answered Jan 31 '19 at 10:54

Ria

10,237
3
33
60

Anwarvic · Answer 2 · 2019-01-31T11:00:55.293

No, it works well... I believe sorted sorts characters based on their associated unicode value. The following is the unicode character for each character:

ا : \u0627
ب : \u0628
ح : \u062d
س : \u0633
ص : \u0635
ف : \u0641
م : \u0645
پ : \u067e
ک : \u06a9

As you can see, the unicode of پ is \u067e while the unicode of ب is \u0628. And the reason for that is the ب is also an Arabic character the same as ا , ح, س, ص, ف, and م. While پ and ک are not.

sort Persian strings for python

2 Answers2