RE: using UTF-8 in openldap 2.0.7

["Smits.Dolf" <Dolf.Smits@siemens.nl>]

According to the LDif rfc, Any attributevalue which contains UTF8
characters, should be BASE64 encode.
As soon as a "plain text" attribute value contains UTF8 characters, you will
get that error messagage you quoted.

Depending on the vendor, nearly always latin1 is supported an can be loaded
without any special precautions.

I think that just loading the file with umlauts in latin1 will do.

Success, Dolf Smits


> -----Original Message-----
> From: Stephan Cremer [mailto:scremer@dohle.com]
> Sent: woensdag 9 mei 2001 16:46
> To: openldap-software@OpenLDAP.org
> Subject: using UTF-8 in openldap 2.0.7
> 
> 
> Hello openldap-software,
> 
> I have a problem importing LDIF files containing German Umlauts.
> Starting with an iso8859-1 (latin1) encoded file, I used the 
> "recode" program to convert the data into the UTF-8 format:
> 
> ---------- snip ----------
> $ recode l1..u8 mydata.ldif
> ---------- snip ----------
> 
> After that, trying to import the converted LDIF-file using ldapadd,
> I got the following message:
> 
> ---------- snip ----------
> $  ldapadd -x -P 3 -D "<binddn>" -W -f /etc/openldap/ldif/akr.ldif 
> ldap_add: Invalid syntax
>         additional info: value contains invalid data
> 
> ldif_record() = 21
> ---------- snip ----------
> 
> -> I browsed the LDAP mailing lists and found a couple of messages
>    containing programs and scripts to convert between latin1 and 
>    utf8. I compared their results to that of recode and found out 
>    they all produce exactly the same output as recode.
> 
> -> browsing those messages I take it that I have to use UTF-8 encoded
>    LDIF data. However, if I use latin1-encoded data, ldapadd will take
>    it without any complaints, while it produces the above 
> error message
>    when using UTF-8 encoded data.
> 
> -> since those discussions were mostly from 1998, I suppose this task
>    should not be a problem...
> 
> I am using openldap 2.0.7 - I tried the percompiled version from the
> SuSE 7.1 distribution as well as compiling the source code myself.
> The results were always the same :-(
> 
> What am I doing wrong?
> Suggestions, comments or hints would be greatly appreciated.
> Best regards,
> 
> Stephan Cremer.
>