Re: UTF-8 to Japanese encoding scheme for IBM Mainframe OS 390

From: Jungshik Shin (jshin@mailaps.org)
Date: Tue Nov 12 2002 - 15:14:36 EST

  • Next message: Michael Everson: "Roadmaps reorganized"

    On Tue, 12 Nov 2002, sourav mazumder wrote:

    > Need an urgent help regarding UTF-8 data conversion in
    > IBM Mainframe 390.

    > 1. What is the encoding scheme is generally used in
    > IBM Mainframe 390 for Japanese language ? I'm aware of
    > IBM EUC. But this is not working.

    I guess it uses IBM code page 0300(IBM Japanese DBCS-Host encoding),
    EBCDIC-based modal encoding. (ref. Ken Lunde, CJKV Information
    Processing, p.583, p. 607, p.623). According to him, it's not
    possible to map IBM Japanese EUC and IBM Japanese PC(almost
    identical to Shift_JIS) to/from IBM Japanese DBCS-Host encoding(code
    page 0300) algorithmically. (the mapping table is necessary.).
    <ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/IBM/> has the following
    in README.TXT file:

    --------
    IBM has many conversion tables to Unicode available in a binary (machine
    readable) format on a CD-ROM included with the document SC09-2190-00
    'CDRA Reference and Registry'. IBM has also defined new code pages which
    contain the Euro Sign. For further information, please contact the IBM National
    Language Technical Centre (nltc@ca.ibm.com).
    ---------

    > 2. Is there any API(s) available in Windows platform
    > (2000) using which I can use to convert the UTF-8 data
    > to the encoding scheme supported by IBM Mainframe 390
    > for Japanese ?

      I doubted WideCharToMultiByte API (since you have your
    data in UTF-8, it should be easy to convert it to 'WideChar' - UTF-16
    in Windows-) supports IBM code page 0300, until googling 'IBM codepage
    300 WideCharToMultiByte' turned up the following page:

      http://www.sharmahd.com/tm/codepages.html

    This page should save your day, I think.

    > 3. Does Java has any equivalent encoding scheme which
    > can use to convert the UTF-8 data to the encoding
    > scheme supported by IBM Mainframe 390 for Japanese ?

      I looked up the Java I18N API reference. Unfortunately,
    IBM Japanese (EBCDIC-based) code page 0300 is not listed among supported
    encodings. Bruno Haible's libiconv(ported to Windows as well) doesn't,
    either. ICU may support it, but somehow I couldn't find the list of
    codepages/encodings supported by ICU's UConvert API. If ICU supports it,
    you'll hear from Markus Scherer very soon (if not already).

      Jungshik



    This archive was generated by hypermail 2.1.5 : Tue Nov 12 2002 - 15:54:01 EST