sort accented string

[Laurent]

Hello,

I'm french and I have a small sorting problem with python (and zope's
zcatalog):

In a python shell python, try :

'é' > 'z'

The answer is true. Then if you try

test=['a','b','f','é','z']
print test.sort()

You almost get :

['a','b','f','z','é']

That's not what we want

Any help will be greatly appriciated.

PS : I got the same behaviour with zope's zcatalog during a sort_on request.
Even if I start zope with the correct linux locale (fr_FR).

THX

Laurent.

 

[Peter Otten]

I'm french and I have a small sorting problem with python (and zope's

Try locale.strcoll():

import locale
locale.setlocale(locale.LC_ALL, "fr_FR") 'fr_FR'
test = list("abéfgz")
test.sort()
test ['a', 'b', 'f', 'g', 'z', '\xe9']
test.sort(locale.strcoll)
test ['a', 'b', '\xe9', 'f', 'g', 'z']

Peter

 

[Laurent]

Thank you very much Peter, that's a really great tip. You should try to
publish this one on the Internet 'cause I didn't find any web site
explaining this.

Bye.

Laurent

I'm french and I have a small sorting problem with python (and zope's

Try locale.strcoll():

import locale
locale.setlocale(locale.LC_ALL, "fr_FR") 'fr_FR'
test = list("abéfgz")
test.sort()
test ['a', 'b', 'f', 'g', 'z', '\xe9']
test.sort(locale.strcoll)
test ['a', 'b', '\xe9', 'f', 'g', 'z']

Peter

 

[Laurent]

Peter what about using this kind of tip with cmp() built-in ??

def myCompare(self,a,b):
cmp(a,b)

with a and b being french accented chars ???

Any idea ??

Thanks again.

I'm french and I have a small sorting problem with python (and zope's

Try locale.strcoll():

import locale
locale.setlocale(locale.LC_ALL, "fr_FR") 'fr_FR'
test = list("abéfgz")
test.sort()
test ['a', 'b', 'f', 'g', 'z', '\xe9']
test.sort(locale.strcoll)
test ['a', 'b', '\xe9', 'f', 'g', 'z']

Peter

 

[Peter Otten]

Peter what about using this kind of tip with cmp() built-in ??

def myCompare(self,a,b):
cmp(a,b)

with a and b being french accented chars ???

Any idea ??

From the documentation of the locale module:

"""
strxfrm(string)

Transforms a string to one that can be used for the built-in function cmp(),
and still returns locale-aware results. This function can be used when the
same string is compared repeatedly, e.g. when collating a sequence of
strings.
"""

Peter

 

[David Fraser]

Laurent wrote:




From the documentation of the locale module:

"""
strxfrm(string)

Transforms a string to one that can be used for the built-in function cmp(),
and still returns locale-aware results. This function can be used when the
same string is compared repeatedly, e.g. when collating a sequence of
strings.
"""

Peter

Note however that you need to transform both strings being compared...
e.g., with a french locale:
True

David