John:
In the dark resesses of my mind, I remember that MPE allocates extents
as it need them. So on an empty file, the file limit would be small
while at full capacity it would be large.
Just to test, fill the file with full size records and see what
happens. Sometime we learn by seeing the results of our actions.
Regards, Olav.
John Pollard wrote:
>To clarify:
>
>I am already using the FFILEINFO and FLABELINFO intrinsics to get the characteristics of the input file. I chose those intrinsics because they appear to provide the information I need in a format the same (or nearly the same) as that required for the HPFOPEN intrinsic. Doing a programmatic LISTFILE (for example) may get me the same information ... but that information will only be usable in HPFOPEN after a relatively significant amount of internal manipulation.
>
>So far, I haven't seen anything in the documentation, or read anything in any of the replies here, that tells me why I "can't" get the info I need using those (or perhaps other) intrinsics on the opened input file.
>
>My thought is that I'm not sure I am getting/using the correct characteristics in the correct way when I supply them to the HPFOPEN intrinsic for opening the output file.
>
>
>1.) What input file characteristics (available from intrinsics, or elsewhere) do I use to create the correct "capacity" (or FLIMIT) for the output file when the file is a variable record length file?
>
>A variable record length Cobol file created by EDITOR looks like this after a LISTFILE,2:
>
>FILENAME CODE ---------------LOGICAL RECORD---------------- ----SPACE----
> SIZE TYP EOF LIMIT R/B SECTORS #X MX
>
>UTP113 EDTCT 1276B VA 1447 40 1 208 2 2
>
>What characteristics of file UTP113 would plug into this formula,
>
>(blockitem#=recordsize * blockfactor).
>
>and produce a LIMIT of 40?
>
>[I'm already producing an output file with a record SIZE of 1276, a TYP of VA, and an EOF of 1447 ... but my output file has a LIMIT of 1447 (same as EOF), not 40 as the input file does.]
>
>
>2.) Why wouldn't I be able to use the FLABELINFO item 51 KSAM key "record" to pass to HPFOPEN item 54 when creating a new output NM KSAM file?
>
>The documentation for HPFOPEN for item 54 says:
>"54/RECORD KSAM parm:
>Passes a record that defines the keys for a new KSAM file.
>(KSAM XL and KSAM64) For KSAM XL and KSAM64 files, refer to the
>parm parameter discussion in the Using KSAM XL".
>
>As I read the Using KSAM XL documentation, this seems like it should be possible.
>
>"The HPFOPEN and FOPEN intrinsics can be used within a program to create and open a
>KSAM file in a single step. As with the BUILD command, file and key characteristics are
>provided as parameter data".
>
>And
>
>"The unique KSAM file and key characteristics are contained in an array that varies in
>length from 40 to 162 words. The format of the array is shown in Figure 2-9".
>
>It seems to me, Figure 2-9 in Using KSAM XL is the same "ksamparam" parameter returned by FLABELINFO item 51.
>
>"Item 51: KSAM XL/64 Parameters:
>This item returns the ksamparam for KSAM XL/64 files, which is an array that
>specifies the characteristics for up to 16 keys for the file".
>
>
>I realize that if I can't get the info I need from the opened input file, I can't take the approach I originally planned.
>
>But it seems to me that this is more of a problem of "what" input file info I need, and "how to use" that info in HPFOPEN once I have it. And I think some of the first part of that problem and all of the second part of that problem will remain the same no matter what source I use to get the input file characteristics.
>
>* To join/leave the list, search archives, change list settings, *
>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
>
>
>
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|
