Infact i did the export/import having nls_lang set to american_america.utf8
Svend Jensen wrote:
> Stan wrote:
>
>> We have a front-end application thats running on java
>> servlets/perl/jdbc connected to an oracle 8.1.6 database and recently
>> we did a conversion of database from US7ASCII to UTF8 to support
>> multi-lingual features. As US7ASCII is a strict superset to UTF8, i
>> was able to change the database characterset as well national
>> character set without any problems. Also, i was told to do an export
>> and import with NLS_LANG properly set to avoid any non-ascii chars
>> dangling in the database without being converted to UTF8. So
>> everything went well without any issues, but now all the non-ascii
>> characters i.e. > 128 that were in US7ASCII database, now got
>> converted to ? in UTF8 database. Is there a way to avoid these ? marks
>> as am getting sun.io.MalformedInput Exception error on our java
>> front-ends. But if i do an update on the tables like below, i get away
>> with these errors.
>>
>> update x set col1=convert(col1, 'UTF8', 'US7ASCII')
>> where col1!=convert(col1, 'US7ASCII');
>>
>> Note : all our frontend applications input forms are also utf8 complaint.
>>
>> any solutions or workarounds.
>>
>> -Stan
>>
>
> Please clarify, did you export with nls_lang = us7ascii and import with
> nls_lang = utf8, or was both exp and imp done with nls_lang = utf8?
>
> /Svend
>
Received on Fri Mar 01 2002 - 01:18:16 CST
