InterBase And Hebrew language

By: Borland Staff

Abstract: InterBase does not currently have a Hebrew or Arabic driver.

Problem:
Does InterBase provide support for Hebrew and Arabic Languages?

Solution:
This article applies to all versions of InterBase through 6.0

InterBase does not currently have a Hebrew or Arabic driver. 
                           
In Hebrew and Arabic (called bi-directional languages for a reason 
you'll see in a moment) - the "normal" display order is right-to-left, 
however, the storage order is "left-to-right". Furthermore, embedded 
"foreign" words and numbers are displayed left-to-right. 

Assume uppercase characters are Hebrew, and lower are English. 

The string 'HELLO world REPEATED 123 TIMES' is displayed 
as 'SEMIT 123 DETAEPER world OLLEH' 

Indices are built using storage order, not display order. 

Display is also complicated by the script nature of Arabic - where the 
glyph that represents a character is determined by the glyphs of 
the characters to the left-and-right. OS Display primitives generally 
take care of all this. 

Another big However: 
Binary indices (character set order) of Arabic and Hebrew are easy 
to make. Dictionary indices become quite a bit harder. 

Even though InterBase does not support Hebrew and Arabic - there is an 
ability to add character set and collations to the product. Perhaps 
if you submit requests to Inprise this can be documented in the future. 

If you don't need to use translitation (multiple character sets in the same 
product) - you could simply store your character data in CHARACTER SET NONE 
fields. You won't get Hebrew or Arabic dictionary collation, but you 
will get back the same byte sequence you put into the database. 
UNICODE_FSS is an alternate - but your client end may have to convert 
the UNICODE_FSS to Unicode or Hebrew character set before IO. 

Source:  Dave Brookstone Schnepper 
4/29/99   Listserve                

Server Response from: ETNASC03