when exporting address book to vcf files, the follows problem is observed:

if there are non latin-1 symbols (my tongue is Russian, so all symbols are defenitely not in latin-1) and length of line is more then 51, kaddressbook tries to split line to the new line, but it ruines utf-8 rules when doing so.

for example, line 

NOTE:ранее: Омск\, Кафедра вычислительных с
истем

tranforms into (last word систем) D1 81 D0 0D 0A B8 D1 81 D1 82 D0 B5 D0 BC

when D1 81 0D 0A D0 B8 D1 81 D1 82 D0 B5 D0 BC is probably expected.

this happens in any VCARD's fields.

Exported file is not possible to use in any mobile device nor in any other address book programs.

Also infomation in google (via plugin) is ruined by the same reason.


Reproducible: Always

Steps to Reproduce:
1. open kaddressbook
2. fill information using Cyrillic or any other non-latin based alphabet
3. be unhappy
Actual Results:  
exported vcf file is not correct utf-8 document (also Kate estimate it as UTF-32 Chinese and can not work with it!),

any 3d party programs can to correctly import such files
google contact information is ruined

Expected Results:  
correct utf-8 encoding, without extra 0D 0A (D0) byte

error is present in 4.4.11 as minimum