Harald Vajkonny napsal(a):
> Dušan Pavlica schrieb:
>
>> Try to convert file to latin1, if it's possible, create database with
>> latin1 charset, create table with required structure (you can set utf8
>> charset to string fields ) and then load data.
>>
> I can not convert the file into latin1, because it is multilingual (i.e.
> European, Japanese, Korean etc.).
>
>
>> What client do you use to check whether data are OK? I can recommend
>> MySQL Query Browser where I don't have to care about setting charset
>> and collation parameters.
>>
>>
> I checked the file in MySQL Query Browser too, but even there I get it
> wrong, when I imported it with a latin session. I get it displayed
> right, when I import it with utf8, but then I have the message with
> duplicate keys and get only half of the data.
>
What's the charset and collation of your primary field in the table?

Dusan