On Sun, 9 Feb 2020 21:50:57 -0500, Phil Smith III wrote:
>Mike Schwab wrote:
>
>>Would UTF-16 to UTF-8 be a better conversion? You still have to be
>>certain of the source character set. And is supported by some z/OS
>>software.
>
I believe HLASM on Linux would probably mistake UTF-8 (IBM 1208)
SYSIN for ISO8859-1 and translate it SBCS-wise to IBM-037. It might
even work if ISO code points 128-255 appear only in quoted strings.
E.g. DC C'матрёшка'
COBOL?
>As Cameron indicated, your comment doesn't quite make sense. UTF-16 is just a
>variable-length encoding, in which basic ASCII*
>(0-127) are single-byte, some characters are two-byte, some three-, and some
>four-. More efficient, especially in the "mostly basic
>ASCII" case.
>
You're describing UTF-8. UTF-16 uses 16-bit code units.
https://en.wikipedia.org/wiki/UTF-16
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
