> There might be an error with mapping number 0-9 digits (0x30-0x39) for
> ISO8859-6 into Unicode.
> These numbers are in the ASCII sections below 0x7F and should stay as they
> are for any code that manipulates digits in string to work.
> Currently the ISO8859-6 mapping table maps the 0-9 digits (0x30-0x39) into
> the Hindu numbers in Unicode (0x0660 - 0x0669).
This is a known issue with that mapping table, 8859-6.TXT, which unfortunately,
has not been updated in a long time.
> However if you compare with Windows Arabic character set Cp1256 the mapping
> table correctly mapps 0x30-0x39 into 0x0030-0x0039.
> Can you please advice?
You are correct. Any implementation of CP1256 and of ISO8859-6 should be
mapping 0x30..0x39 to U+0030..U+0039.
> Isn't it more proper for the Hinu numbers to be moved from the 0x0660-0x0669
> section into the Arabic Presentation section 0xFE7-0XFEFC since they are
> just a different shape from the ASCII 0x30-0x39 digits?
No. This is definitely not the case. The Hindu number shapes used in
Arabic are not simply a different shape for the ASCII digits. They
are sometimes *implemented* that way in 8-bit Arabic code pages, but
in general they have distinct existence and behavior in Unicode.
Also, there can be no moves of characters from one place to another
in the standard. The value of the standard lies primarily in its
being standard: everything stays put, and does not change out
from under people who have implemented it.
> Please reply directly to me, I'm not on the list.
> Many Thanks.
> Omar Halaseh Applications Integration
> 500 Oracle Parkway # Work: 650.506.8707
> Mail Stop 3op445, # Fax: 650.506.7292
> Redwood Shores, CA 94065 # E-Mail: OHalaseh@Us.Oracle.Com
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:44 EDT