On 21 Jan 2007 16:45:33 -0800, in bit.listserv.ibm-main
(Message-ID:<[EMAIL PROTECTED]>)
[EMAIL PROTECTED] (Paul Gilmartin) wrote:
BTW, and FWIW, I have solved the problem by adding to the
code a branch to
normal EOF if after opening the input DCB, and before
issuing the first GET,
the DCBBLKSI is zero. Given the full circumstances of the
situation, I see
it as a normal condition -- an empty input dataset -- and
not an error.
But this leaves a pitfall. If a user allocates the data
set supplying
RECFM, LRECL, and BLKSIZE, but never opens it, you are at
the mercy of
the previous content of the allocated space, and can ABEND
on "READ
WRONG LENGTH RECORD" if the first block read exceeds
BLKSIZE, etc.
Not if the dataset is allocated on SMS DASD. As
noted earlier in this thread, such get an automatic EOF
written at allocation time. It's beginning to be rare that
production data is not on SMS disks.
I stand by my belief that the access method should enter
EODAD on any
attempt to read beyond DS1LSTAR. Spare the programmer the
grief of
performing that check himself.
If DS1LSTAR were reliable enough for this, dump
programs wouldn't make it optional to stop at
DS1LSTAR. Last time I remember looking, the FDR *default*
option was to read to end of extent and ignore DS1LSTAR.
--
I cannot receive mail at the address this was sent from.
To reply directly, send to ar23hur "at" intergate "dot" com
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
