RE: Unicode App and Tables

From: Smita Desai (
Date: Mon Jul 22 1996 - 13:05:33 EDT

Thanks Jean for the info. The conversion that I need to do is for Win
95 and NT, between Unicode and a char set on the server, for example,
Japanese EUC, so WideChartoMultiByte type calls would not work. There
are algorithms to do these char set conversions between Japanese EUC
and Windows client (Ken Lunde's UJIP), but none I suppose for Unicode.
The longer way would be to convert to ANSI or Windows client and then
use MultiBytetoWideChar. But that will cause a lot of performance hit
and is not ideal.


>>Reply to your message of 7/22/96 11:37 AM

>Smita Desai wrote:
>____________________ Begin Original Message ___________________________
>Date: Fri Jul 19 07:03:22 -0700 1996
>From: internet!Unicode.ORG!unicode
>Subject: Unicode App and Tables
>To: internet!Unicode.ORG!unicode
>Content-Type: text
>Content-Length: 808
>Hi Everyone,
>Can anyone please help with the following?
>1. Any example of the source code for a generic Unicode app? The
>reason for requesting this is the question in 2 below.
>2. A set of Unicode tables. If I have an international app, does
anyone =
>have any thoughts on how to load and map the tables? Should I table
>lookup the whole table or be locale-specific? Is that possible? I
>believe I have to keep the whole table available due to global nature
>of the app. One of my concern is performance impact due to the size
of =
>the table. My app needs to map data to Unicode.
>Any help would be greatly appreciated.
>Smita Desai
>International Engineering
>FTP Software, Inc.
>Tel: 508.684.6921
>Fax: 508.684.6105
>Smita, this is Jean-Baptiste,
>1) Dealing with your first problem, I give you some places at the end
of the message.
retrieve two
>Warning: They are not exhaustive, can't be compiled but can help with
>WideChartomultibyte/multibytetowideChar samples.
>2) If I understand your question you wonder if it is possible to
"load" only the part
>of the table that you need.
>This depend on the system you use, on Windows NT or PLAN9 (new AT&T
Bell system)
>UNICODE is inside the system. Files coded with UNICODE can be
processed with system
>calls without doind anything. On Windows 3.1, 95 you'll havw to use
calls such as
>WideChartomultibyte. On Unix you'll have most of the time to load a
table and perform
>the conversion.
>It's up to you to create a table with the full UNICODE encoding or
just the codes you
>need, but be carefull, if one day you retrieve or a user fill a data
that is not in the
>table, you'll crash !
>Here are some information:
>Where to search ? :
>- tips on I18N.
>- and
>- Dejanews server (, mix searches with: UNICODE,
>- (Good site!)
retrieve two
>Warning: They are not exhaustive, can't be compiled but can help with
>WideChartomultibyte/multibytetowideChar samples.
>- FTP site, Start with the /pub/ directory
>- This package handling UTF-7, UTF-8, GB 2312 and Big-5 conversion.
>- conversion tables extracted from cjkxref.txt,
>found on the site.
>- for conversion tools on all platforms.
>- Microsoft Win32 SDK Help: Win32/Overview/International Features. (45
>- Nadine Kano's Book "Developping International Software for W95 and
NT", Microsoft
>- Newsgroup: will provide discussions and :
>Archive-name: internationalization/font-faq
>Archive-name: internationalization/programming-faq
>- Newsgroup
>- Newsgroup comp.std.internat
>- subscribe to listserv and send subscribe

Smita Desai
International Engineering
FTP Software, Inc.

Tel: 508.684.6921
Fax: 508.684.6105

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:31 EDT