FreeTDS Development Group

["Donald Sures" <donald_sures AT symantec.com>]

http://www.freetds.org/userguide/a2698.htm


i've read that .. i've read that applying the Character Set = UTF-8 
parameter would allow me to query UTF-8 data from a MSSQL NVarChar column 
.. The data is 9 characters of japanese text .. these characters have the 
first byte as E3 which makes them 3 byte characters .. somewhere in the 
conversion from UCS2 to UTF-8 i'm losing bytes .. 

is iconv responsible for converting ucs2 to utf-8? 

-  Perl: v5.8.0 built for i686-linux
- TDS_VERSION_NO "freetds v0.60"
- Microsoft SQL Server  2000 - 8.00.578 (Intel X86)
- MSSQL Column Type NVCHAR(200)
- MSSQL Column Text : 9 characters
- Perl Length() reports text is 9 characters
- Perl prints first 3 characters of string

- For every 3 Japanese characters in the database, 1 Japanese character is 
displayed in perl
  Here are the hex codes for each byte of the data string:
        テ-E3 83 86
        - E3 82 B9 
        - E3 83 88 
        - E3 83 86 
        - E3 82 B9 
        - E3 83 88 
        - E3 83 86 
        - E3 82 B9 
        - E3 83 88

LOCALES.CONF
==============
[default]
        date format = %b %d %Y %I:%M%p
        char set = UCS-2

[en_US]
        date format = %b %d %Y %I:%M%p
        language = us_english
        char set = UCS-2


FREETDS.CONF
===============
[xxxx]
        host = xxxx
        port = 1433
        tds version = 7.0
        dump file = /tmp/freetds.txt
        debug level = 99
       client charset = UTF-8






---
DJ Sures
Alert Services
Symantec Corporation
V 1.403.213.3939 Ext. 250
F 1.403.233.9179