Unicode Support

[benthicsoftware]

Hi Marco,

I want to run the following query:
select nchr(1513) from dual;
under Oracle 9i and get back a unicode character (it's a hebrew character that looks like a styleized 'W' that I'm using for testing.) I can't seem to get this back using DOA (I'm using some Unicode aware controls and have been able to just display the character without using DOA.) I have been able to get the above query to work using a web page and ASPDotNet (with the Oracle .Net data provider.) Any thoughts on this? My nchar characterset is set for UTF-16 and I'm doing this on Windows XP. I've tried setting BytesPerChar to autodetect as well as other values with no changes. The normal (non nchar) dataset I'm using is a single byte charset.

Thanks,

Mark Ford
Benthic Software

[This message has been edited by benthicsoftware (edited 21 April 2003).]

 

[Marco Kalter]

Direct Oracle Access doesn't do anything with the character data, it is returned as is. What bytes do you get back when you select nchr(1513)?

------------------
Marco Kalter
Allround Automations

 

[benthicsoftware]

I get a single byte chr(191). This is the actual data in the buf field of the TFieldData object in the FieldByIndex call. Does it work for you? Does DOA support nchar/nvarchar types?

Thanks,

-Mark

 

[Marco Kalter]

There is no explicit support for nchar, and I think the character data is converted to the character set specified in NLS_LANG. Setting NLS_LANG to a character set that supports your hebrew characters might fix this.

------------------
Marco Kalter
Allround Automations

 

[benthicsoftware]

Hi Marco,

Thanks. I have done some research and managed to rewrite InternalDescribe and InternalDefine to handle nchar types as unicode (a few other changes were needed as well.) It seems to work so far!

-Mark