topic Re: issues with f$file_format lexical function in Operating System - OpenVMS

issues with f$file_format lexical function

SAMI AHMAD — Wed, 24 Jul 2013 19:31:32 GMT

I am unable to get the number of records in a file using the following

why is "file_length_hint" not working ? and why **bleep**/rms/stats is showing File Length Hint (Data Byte Count): -1 (invalid)?

the vms version is Alpha 7.3-2

$ REC = f$file("A.A","file_length_hint")

$ show sym rec

REC = "(-1,-1)"

$ **bleep**/RMS/stat a.a

RMS File Statistics 24-JUL-2013 15:21:24.39 Page 1

PASDISK5:[MDX.OUTGOING.PROCESSED]A.A;1

FILE HEADER

File Spec: PASDISK5:[MDX.OUTGOING.PROCESSED]A.A;1

File ID: (46161,486,0)

Owner UIC: [INTEROP,MDX]

Protection: System: RWED, Owner: RWED, Group: RE, World:

Creation Date: 24-JUL-2013 14:54:40.07

Revision Date: 24-JUL-2013 14:54:40.07, Number: 1

Expiration Date: none specified

Backup Date: none posted

Contiguity Options: none

Performance Options: none

Reliability Options: none

Journaling Enabled: none

RMS FILE ATTRIBUTES

File Organization: sequential

Record Format: variable

Record Attributes: carriage-return

Maximum Record Size: 0

Longest Record: 7

Blocks Allocated: 61, Default Extend Size: 0

End-of-File VBN: 1, Offset: %X'001C'

File Monitoring: disabled

File Length Hint (Record Count): -1 (invalid)

File Length Hint (Data Byte Count): -1 (invalid)

Global Buffer Count: 0

The analysis uncovered NO errors.

**bleep**/RMS/STAT A.A

Re: issues with f$file_format lexical function

Steven Schweda — Wed, 24 Jul 2013 21:08:14 GMT

Re: issues with f$file_format lexical function

Steven Schweda — Wed, 24 Jul 2013 21:33:00 GMT

Re: issues with f$file_format lexical function

John McL — Thu, 25 Jul 2013 01:50:41 GMT

You can't get recod counts that way.

The following command procedure will do it for you, P1 is the target file.

$!
$       pipe search/stat 'p1' XYZZY | search sys$input "records searched" | -
                 (read sys$input module_name ; define/job MY_REC &module_name)
$       my_sym = f$trnlnm ("MY_REC", "LNM$JOB")
$       deass/job my_rec
$       recs = f$integer(f$extract(26, 10, my_sym))
$       write sys$output "No of records is ''recs'"
$!
$       exit

Modify to suit your purpose...

Re: issues with f$file_format lexical function

Hein van den Heuvel — Thu, 25 Jul 2013 03:07:58 GMT

>> You can't get recod counts that way.

Yes you can, Steven proved that in his second reply.

>> pipe search/stat 'p1' XYZZY

1) it is my understanding the an early match if quicker than no-match, so i use "" /WIN=0

2) In any reasonably recent openVMS version SEARCH will define statistics symbols for your convenience.

$ define/user sys$output nl:
$ searc /stat /win=0  ""  tmp.tmp
$ show symb sear*
:
  SEARCH$RECORDS_SEARCHED = "13"
$

Works on all file types, contrary to the file hints.

hth,

Hein

Re: issues with f$file_format lexical function

Steven Schweda — Thu, 25 Jul 2013 03:21:40 GMT

Re: issues with f$file_format lexical function

SAMI AHMAD — Thu, 25 Jul 2013 17:59:21 GMT

i found the problem after reading openVMS notes from web , f$file("<file name>","file_length_hint")

works only if

- file type is sequential

- record type is VAR or VFC

- volume is ODS-5

to fix the problem I :

- generated the FDL for the file

- modified the FDL to correct the file type and record type.

- converted the file using the modified FDL

and then f$file("<file name> ","file_length_hint") worked and gave me the record count.

Re: issues with f$file_format lexical function

Steven Schweda — Thu, 25 Jul 2013 18:21:11 GMT

Re: issues with f$file_format lexical function

John McL — Thu, 25 Jul 2013 22:16:54 GMT

Hein,

Thanks for the tip but your search command seems to have the search string and the filename the wrong way around.

Also, I found that I was seeing error messages, so ...

$ def/user sys$output nl:
$ def/user sys$error nl:
$ search/stat/win=0 [-]login.com XYZZY

(If sys$error not redefined or messages not disabled, here you'll see "%SEARCH-I-NOMATCHES, no strings matched")

$ sho sym search*
SEARCH$CHARACTERS_SEARCHED = "2300"
SEARCH$FILES_SEARCHED = "1"
SEARCH$LINES_PRINTED = "0"
SEARCH$RECORDS_MATCHED = "0"
SEARCH$RECORDS_SEARCHED = "94"

Re: issues with f$file_format lexical function

cachaza — Wed, 31 Jul 2013 12:45:27 GMT

Hi : ...i dont speak english...only spanish (Chile-south America).....

I know a way to calculate de records quantity ..without read the file ,but only for sequential file (fixed format)

1.- dcl : receive de file name by parameter,extract and send file attributes to a cobol program.

2.- Cobol: calculate de records quantity and display it....send this quantity to dcl too.

3.- below you will find dcl + cobol

$ Set NoVer

$ Status:= " "

$ Ws := "Write Sys$output "

$ On Error Then Goto Err_Com

$!----------------------------------------------------------------*

$! El Pgma Cobol Util_ctareg.sco Devuelve La Variable "Numregs", *

$! La Cual Contiene El Numero De Registros Que Tiene El Archivo *

$! Nota: solo acepta archivos Secuenciales de Largo Fijo *

$!----------------------------------------------------------------*

$ Archivo :='P1'

$ Archivo = F$edit(archivo,"upcase,trim,compress")

$ Archivo = F$string(archivo)

$ Ws " Archivo Ingresado : ''Archivo'"

$!-----------------------------------------------------------------------------------------------*

$ parOrg = f$file_attributes(archivo,"org")

$ if parOrg.NeS. "SEQ"

$ then

$ Ws "*** Error : Este archivo NO es Secuencial ***"

$ Exit

$ endif

$ parRfm = f$file_attributes(archivo,"rfm")

$ if parRfm.NeS. "FIX"

$ then

$ Ws "*** Error : Solo admite registros de Largo Fijo ***"

$ Exit

$ endif

$!-----------------------------------------------------------------------------------------------*

$ Pareof = F$file_attributes(archivo,"eof")

$ Pareof9 = F$fao("!(09ZL)",f$integer(pareof))

$!---------------------------------------------------*

$ Parffb = F$file_attributes(archivo,"ffb")

$ Parffb9 = F$fao("!(09ZL)",f$integer(parffb))

$!---------------------------------------------------*

$ Parmrs = F$file_attributes(archivo,"mrs")

$ Parmrs9 = F$fao("!(09ZL)",f$integer(parmrs))

$!-----------------------------------------------------------------------------------------------*

$!=====================================================

$ Run/nodeb Dka1:[optiafc.cta.exe]util_ctareg.exe

$!=====================================================

$! Ws " Numero De Registros Recibido: ''Numregs'"

$ If Status.eqs."-1"

$ Then

$ Ws "***===>>> Util_ctareg Mal-Mal-Mal <<<===***"

$ Else

$ Ws "***===>>> Util_ctareg Ok-Ok-Ok <<<===***"

$ Endif

$ Exit

$!--------

$ Err_Com:

$!--------

$ Set Nover

$ Ws "***===>>> Util_ctareg con ERROR a nivel de .Com <<<===***"

$ Exit

$!-----------------------------------------------------------------------------------------------*

IDENTIFICATION DIVISION.

*************************

PROGRAM-ID. UTIL_CTAREG.

*AUTHOR CACHAZA.

ENVIRONMENT DIVISION.

**********************

CONFIGURATION SECTION.

*=======================

SPECIAL-NAMES.

*--------------

DECIMAL-POINT IS COMMA.

DATA DIVISION.

***************

WORKING-STORAGE SECTION.

*========================

01 VARIABLES-INICIO.

02 PAR-EOF9 PIC 9(09).

02 PAR-FFB9 PIC 9(09).

02 PAR-MRS9 PIC 9(09).

02 WS-PIC9 PIC 9(09).

02 WX-LARGO PIC 9(04) COMP.

02 NUM-REG PIC 9(12) COMP.

02 NUM-REGZ PIC ZZZ.ZZZ.ZZZ.ZZ9.

********************

PROCEDURE DIVISION.

********************

AAXX.

INITIALIZE VARIABLES-INICIO

CALL "LIB$GET_SYMBOL" USING BY DESCRIPTOR "PAREOF9", PAR-EOF9,

BY REFERENCE WX-LARGO.

IF WX-LARGO = 0

DISPLAY "PARAMETRO PAR-EOF9 NO DEFINIDO. CANCELA . . ."

STOP RUN.

move 0 to WX-LARGO

CALL "LIB$GET_SYMBOL" USING BY DESCRIPTOR "PARFFB9",PAR-FFB9

BY REFERENCE WX-LARGO.

IF WX-LARGO = 0

DISPLAY "PARAMETRO PAR-FFB9 NO DEFINIDO. CANCELA . . ."

STOP RUN.

move 0 to WX-LARGO

CALL "LIB$GET_SYMBOL" USING BY DESCRIPTOR "PARMRS9",PAR-MRS9

BY REFERENCE WX-LARGO.

IF WX-LARGO = 0

DISPLAY "PARAMETRO PAR-MRS9 NO DEFINIDO. CANCELA . . ."

STOP RUN.

IF FUNCTION REM (PAR-MRS9 , 2) = 1

ADD 1 TO PAR-MRS9

END-IF.

COMPUTE NUM-REG ROUNDED = (((( PAR-EOF9 - 1 ) * 512 ) + PAR-FFB9 ) / PAR-MRS9 )

MOVE NUM-REG TO NUM-REGZ

DISPLAY "*--------------------------------------------------*".

DISPLAY "*** ==== > ARCHIVO TIENE:"NUM-REGZ CONVERSION" REGS"

DISPLAY "*--------------------------------------------------*".

CALL "LIB$SET_SYMBOL" USING BY DESCRIPTOR "NUMREGS",NUM-REG

STOP RUN.

*****************************************************************************

Re: issues with f$file_format lexical function

Hein_vdHeuvel_d — Wed, 31 Jul 2013 13:33:51 GMT

Thank you cachaza for your contribution. It is clear enough.

You are correct, IF-and-only-IF, the file is fixed length record sequential then you can calculate the number of records from eof-block * 512 + ffb / size.

Note: Size should be picked up from LRL, not MRS, and should be rounded up to even value before dividing.

I the file is larger than 2GB (4194303 blocks) you indeed need a (cobol) helper program to do the math with more than 32 bits, but for most files you can just use DCL divides.

Too bad that RMS does not do file-length-hints for this easy file type huh!

I had an Email exchange with John... Searching for and empty string makes sure that there is an immediate hit for each record, so that's quicker and lookign at each bytes. Also, this avoids the 'no match found' message to sys$error.

Cheers,

Hein

Re: issues with f$file_format lexical function

H.Becker — Wed, 31 Jul 2013 17:08:10 GMT

> COMPUTE NUM-REG ROUNDED = (((( PAR-EOF9 - 1 ) * 512 ) + PAR-FFB9 ) / PAR-MRS9 )

>You are correct, IF-and-only-IF, the file is fixed length record sequential then you can calculate the number of records from eof-block * 512 + ffb / size.

The right formula - subtracting 1 or not - depends on ffb. The one in your Cobol source is correct for ffb>0. The other one - insert the appropriate parentheses - is correct for ffb==0 (which makes the addition somehow obsolete).

$ f:=whatever

$ cre 'f

whatever

Exit

$ write sys$output f$file(f,"eof")

$ write sys$output f$file(f,"ffb")

$ ! here rfm=var and the file size is 10: (eof-1)*512+ffb

$ convert/fdl=tt: 'f 'f.fix/pad=%x20

FILE

ORGANIZATION sequential

RECORD

CARRIAGE_CONTROL none

FORMAT fixed

SIZE 512

Exit

$ write sys$output f$file(f+".fix","eof")

$ write sys$output f$file(f+".fix","ffb")

$ ! the file size is 512: eof*512+ffb