The Dignus assembler works on Linux (z/Linux and/or x86 Linux).
We had to cross this bridge a long time ago, since we have to handle
the translation issue when assembling on ASCII platforms.
The Dignus assembler defaults to the 1047 code-page, but has
options that allow the user to control the translation should the
situation warrant it.. they are very crude, allowing you to select
various code points and define a different translation… nothing
as fancy as “use codepage blah-de-blah”.
However - to be useful and answer the question, I believe these
would be the bytes that move-around when comparing code-page 037
and code-page 0147:
0x5E - circmflex
0xac - PL/I style NOT
0x5b - Left bracket
0x5d - Right bracket
0xdd - ???
0xa8 - ???
(Not sure what the 0xdd and 0xa8 bytes are…)
You can run the Dignus assembler in “ASMA90” mode, where it
operates “just like” IBM’s assembler under z/Linux. We had to
adjust our translation tables to accommodate that, which is where
I gleaned this list (since we are “just like” HLASM, we had to
use their code-page.)
- Dave Rivers -
--
[email protected] Work: (919) 676-0847
Get your mainframe programming tools at http://www.dignus.com
On Jan 13, 2014, at 12:03 PM, Paul Gilmartin <[email protected]> wrote:
> On 2014-01-13, at 08:50, Ed Jaffe wrote:
>
>> On 1/13/2014 6:35 AM, John McKown wrote:
>>> I had never considered this until Gil posted his problem with DC CA'[]'
>>> where the [] characters are the IBM-1047 code points for those glyphs
>>> instead of the CP-037 code points. I have the same problem . I have set my
>>> 3270 emulator to "Latin1", or IBM-1047, so that I can use OMVS and have use
>>> some special characters such as: {, }, [, and ]. which UNIX expects in the
>>> IBM-1047 character set, not the CP-037.
>>
>> IBM-1047 has become the most common code page used by z/OS programmers.
>>
> If I were JWG I'd insist on a citation. Which I'd ignore if it
> contradicted my prejudices.
>
> What fraction of z/OS users is "programmers". I suppose all HLASM
> users are programmers.
>
> It's dismaying:
>
> o That HLASM lacks support for the most common code page
> used by z/OS programmers.
>
> o That HLASM lacks support for the code page which is standard
> for Unix System Services. Yah, Assembler far antedates USS,
> and perhaps IBM-1047. But HLASM is bucking the historical
> dialectic.
>
> o That HLASM fails to report as an error the appearance of a
> code point in a CA'...' constant that isn't properly
> translatable to an ISO8859-1 equivalent.
>
> Even though I recognize that admitting a second code page opens
> the gate to a Babel of additional code pages.
>
> What characters significant to HLASM are not stable between 037
> and 1047?
>
> How do C'...' and CA'...' work in Linux. Am I correct that
> HLASM is available under Linux?
>
> -- gil
>
