On Wed, 2 Sep 2015 07:38:11 -0500, John McKown wrote:

>I am wondering if anyone else thinks that it might be useful for IBM to
>make an enhancement to BSAM/QSAM support for the CCSID parameter in JCL.
>
>ref:
>http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/iea2b6a0/12.9
><quote>
>
>Purpose
>
>You can request the access method to convert data between the coded
>character set identifier (CCSID) specified on the JOB or EXEC statement and
>the CCSID specified on the DD statement. Data conversion is supported on
>access to ISO/ANSI Version 4 tapes using access methods BSAM or QSAM, but
>not using EXCP.
>
><quote/>
>Why does this only support ASCII tapes? Would it be helpful to generalize
>it to any data source read or written by BSAM/QSAM?
>​ Of course, this is only of use for "pure text" data sets. But it might
>address the recent thread on FTP'ing a VB data set, keeping the LLBB field
>intact but translating the "data" portion to ASCII.
> 
Yes, yes, yes.  This is another instance of IBM's implementing a valuable
facility but with unduly restrictive scope, perhaps at the wrong implementation
layer.  (Does the code reside in the tape driver?  Even if so, why restrict
it to ISO/ANSI tapes?)

Related:  I can use SDSF to extract SYSOUT to UNIX files.  I'd like to tag
those files with CCSID for possible subsequent conversion to Unicode.
How can I determine the effective CCSID for a SYSOUT.  UCS and/or
CHARS appear to carry similar information: how to render code points
as viewable glyphs/graphemes.  Is there a table, somewhere, that
relates UCS or CHARS to CCSID?

And a disappointment.  If I have a file tagged CCSID=1047, FORMAT=NL,
and I use pax to convert it to ASCII, the resultant files appear tagged
CCSID=819, FORMAT=NL.  But the NLs are converted to LFs, so the
FORMAT should be LF.  Submitted SR.  Got WAD.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to