Re: [Boa Constr] Search text and unicode build

[le d. <le....@la...>]

Werner F. Bruhin a écrit :
> Hi,
> 
> le dahut wrote:
>> It are normal .py files.
>>
>> I made some tests and it appears if there are some "caractères 
>> accentués". If I have :
>> """
>> coucou
>> psqfqsfq
>> coucou
>> sdfs fsdf sdf sd fcoucou sdfsd
>> sdkj fmqlkj fkqmsj fkq coucou  sdfsq
>> éé
>> sdlkfjds fsd
>> sd f
>> q
>> f qf coucou
>> sdfds coucousdfsd
>> """
>>
>> and I search for "coucou", it is correctly highlighted before the 
>> characters "éé". After the blue highlight is 2 characters before where 
>> it should be.
>> It seems that the shift corresponds with the number of accentuated 
>> characters.
> Yes, I see this some strange things if the source contains special 
> characters, probably some issue with the highlighting causing a problem 
> with the unicode escaped character which follows.
> 
> Would be interesting to see if other editors which are based on 
> wxStyledTextCtrl show the same problem.
>> I have modified C:\Python25\Lib\site.py line 370 with :
>> encoding = "iso8859-1"
> I would suggest to use a sitecustomize.py in your python root folder for 
> this, with the following content:
> 
> import sys
> sys.setdefaultencoding('utf-8')

Why utf-8 ? windows is iso8859-1 encoded isn't it ?

> 
> Werner
> 
>