On Fri, 2 Jan 2026 18:09:28 -0500, Phil Smith III wrote:
>Showing my ignorance of COBOL: Does a MOVE from EBCDIC->UTF-8 mean "Here's a
>variable defined as n EBCDIC characters, please move it to this other variable
>defined as (at least) n UTF-8 characters"?
>
>If so, as Gil intimated, I wouldn't say there "is no technical reason". I can
>think of at least one problem: A UTF-8 string that contains characters from
>multiple UTF-8 blocks. For example, x'52' in EBCDIC 410 (Cyrillic) is a њ; in
>420 (Arabic) it's ؤ. There's no reason a UTF-8 variable cannot contain њؤ but
>if it does, you can't convert back to EBCDIC.
> ...
How can a programmer code a (PIC U) constant or initial value?
I imagine something like:
MOVE 'Latin, русский, español, français, Ελληνικά, עִברִית. '
TO UtF-8-STRING;
Details? Would it be legible in compiler SYSPRINT?
The following works in Regina-Rexx, but hardly practical:
say ,
'4C6174696E2C20D180D183D181D181D0BAD0B8D0B92C2065737061C3B16F6C2C206672616EC3A76169732C20CE95CEBBCEBBCEB7CEBDCEB9CEBACEAC2C20D7A2D6B4D791D7A8D6B4D799D7AA2E20'x
Are there CCSID 1210-savvy terminals and or authoring tools??
Until there are, the less said about UTF-EBCDIC the better.
--
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
