EXECIO in REXX will give you the record count in stem zero. For FB you obtain LRECL using LENGTH function. Multiply the two. No good for VB, so that would require reading whole file.
On Fri, May 15, 2020, 07:05 Gerhard adam <gada...@charter.net> wrote: > > > > > Really? There’s lots that it could do? So why not just read the > DSCB for how much is free and do the calculation that has been on every > 3390 Reference card for 30 years > This is done all the time and in a variety of ways. This problem was > readily assessed for decades using a subroutine. This has always been > something that could readily be done until people start putting language > restrictions on it. It isn’t that the info isn’t available, it is simply > that the information isn’t available given the restrictions placed on the > problem > > > > Get Outlook for iOS > > > > > > > On Thu, May 14, 2020 at 1:56 PM -0700, "Charles Mills" <charl...@mcn.org> > wrote: > > > > > > > > > > > There is lots it could do! Make a decision on "strategy" based on the > volume of input, for example. Do I process it all or cut off after 'n' > records and do more on the next run? Do I read the file into an in-memory > table and access records there, or do I load it into a VSAM file for direct > access? Do I invoke an external sort or use an internal sort? Or a thousand > other things. > > It is wrong as @Shmuel points out to preclude questioning the necessity of > doing something exactly as the OP suggests. It is equally wrong to assume > there could be no good reason for doing it the OP's way. > > Charles > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On > Behalf Of Gerhard adam > Sent: Thursday, May 14, 2020 11:52 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: Is there any z/OS API to get byte file size for non-VSAM, > non-zFS, non-database files? > > > > > > It is easy to say that a COBOL program needs to “know” this but it > is nonsense since there is nothing a COBOL program can do with this info. > If it turns out to really be necessary then a subroutine can be written > (as it has been done for decades) to provide this information. > If the question is simply to bitch about why z/OS does do this > automatically via a call or something then the complaint is directed to the > wrong group > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > > > > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
