|
View previous topic :: View next topic
|
| Author |
Message |
bijoybabu83
New User
Joined: 15 Jan 2007
Posts: 36
Location: Kerala
|
|
|
|
I have a program reading a blocked vb file. The execuction is failing due to read error. The file status code is 04 which denotes a length mismatch between the vb file and the fd entry for the file record.
The following is the fd entry for the file :-
| Code: |
FD PCFORMT
RECORDING MODE IS V
BLOCK CONTAINS 21104 CHARACTERS
RECORD CONTAINS 1636 TO 20996 CHARACTERS
LABEL RECORDS ARE STANDARD
DATA RECORD IS IN-SUB-XTRCT-REC.
01 IN-SUB-XTRCT-REC.
05 IN-FIRST-BYTE PIC X(01).
05 FILLER PIC X(04).
05 IN-TOTAL-COUNT PIC X(07).
05 FILLER PIC X(12).
05 IN-SUB-NUMBER PIC X(0013).
05 FILLER PIC X(1594).
05 IN-DEPENDENT-CNT PIC 9(5).
88 IN-DEP-CNT-VALID VALUE 0 THRU 20.
05 IN-DEPENDENT-DATA OCCURS 0 TO 20 TIMES DEPENDING ON
IN-DEPENDENT-CNT INDEXED BY PINIDX.
10 FILLER PIC X(0968). |
Code'd
This is the file read code :-
| Code: |
READ PCFORMT.
EVALUATE PCFRMT-FL-STAT
WHEN 00
CONTINUE
WHEN 10
MOVE 'Y' TO END-OF-FILE
GO TO 8999-EXIT
WHEN OTHER
DISPLAY '*****************************'
DISPLAY '** PCFRMT FILE INIT **'
DISPLAY '** READ ERROR **'
DISPLAY '** FILE STATUS IS ' PCFRMT-FL-STAT
DISPLAY '*****************************'
CALL 'CANCEL' USING CC-0020
END-EVALUATE. |
The error sysout it gives is :-
The system or user abend U 020 R=NULL was issued.
From compile unit I290013T at entry point I290013T at compile unit offset +0000FF7E at address 0017AFE.
I looked for the offset in the IDIREPT(may be dump) :-
| Code: |
Offset Hex Instruction
------ -------------- ------------------------------
-2C 9620 9000 OI 0(R9),32
-28 5830 8024 L R3,36(,R8)
-24 9160 3000 TM 0(R3),96
-20 4780 B18A BC 8,394(,R11)
-1C 9610 9000 OI 0(R9),16
-18 9180 C2AF TM 687(R12),128
-14 4780 B194 BC 8,404(,R11)
-10 0B02 BSM 0,R2
-E 18EA LR R14,R10
-C 9807 8014 LM R0,R7,20(R8)
-8 989B 8038 LM R9,R11,56(R8)
-4 5880 8034 L R8,52(,R8)
***** 07FE BCR 15,R14
+2 18BE LR R11,R14
+4 9120 1000 TM 0(R1),32
+8 4770 B012 BC 7,18(,R11)
+C 58E0 1008 L R14,8(,R1)
+10 9680 E000 OI 0(R14),128
+14 5840 1008 L R4,8(,R1)
+18 185F LR R5,R15 |
This is what I can see.
Can anybody help me to identify what is the problem
Also I looked for the file attributes :- These are the file attributes.
| Code: |
Data Set Name . . . . : NPTEST.NASCO.EPZ00074.CORP.FEED.D1006
General Data Current Allocation
Management class . . : MCTSTSTD Allocated tracks . : 787
Storage class . . . : SCNCTEST Allocated extents . : 1
Volume serial . . . : NCT921
Device type . . . . : 3390
Data class . . . . . : DATAV255 Current Utilization
Organization . . . : PS Used tracks . . . . : 787
Record format . . . : VB Used extents . . . : 1
Record length . . . : 21000
Block size . . . . : 27998
1st extent tracks . : 787
Secondary tracks . : 385
Data set name type : SMS Compressible : NO
Creation date . . . : 2009/10/06 Referenced date . . : 2009/11/20 |
Also the file is failing in the reading of the very first record which i can see that has only a size of 52 bytes.
Please help. |
|
| Back to top |
|
 |
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8696
Location: Dubuque, Iowa, USA
|
|
|
|
Do you see any conflicts between
| Quote: |
| BLOCK CONTAINS 21104 CHARACTERS |
and
| Quote: |
| Block size . . . . : 27998 |
You don't have to specify BLOCK CONTAINS but if you do, it had better be accurate since the program is the first source of DCB data. |
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
COBOL2, version/release 4.0 (over 20 years ago), introduced an updated method of reading and writing variable length records in Batch, which has been discussed many times before in the COBOL forum.
Try searching for this, as this updated method (IMHO) is an easier way (and more maintainable and straight forward) than the previous method used in OS/VS COBOL as well as earlier version/releases of COBOL2, prior to 4.0.
Bill |
|
| Back to top |
|
|
bijoybabu83
New User
Joined: 15 Jan 2007
Posts: 36
Location: Kerala
|
|
|
|
To Robert,
Actually I used block contains 0 records and submitted the job. It is giving me the same length error.
And then I changed the block size of the file to 21104 by copying the file to another file and putting the new block size to 21104. But still the program gives the same length error |
|
| Back to top |
|
|
bijoybabu83
New User
Joined: 15 Jan 2007
Posts: 36
Location: Kerala
|
|
|
|
To Bill O'Boyle,
Can you please provide me some links to this updated method IMHO. |
|
| Back to top |
|
|
dbzTHEdinosauer
Global Moderator
Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne
|
|
|
|
This is what Robert indicated that you should change.
BLOCK CONTAINS 21104 CHARACTERS
don't put a block size in the cobol program!!!!!!!!!!!!!!!!!!!
After the OPEN and before each READ, (experience from old old cobol)
move 20 to IN-DEPENDENT-CNT.
and why don't you use better programming methods (join, at least the 20th century) and use the work-area option. I do realize that you need to move data definitions around. for your record area, you can have 2 01's. One would be PIC X(1636) the other would be PIC X(20996). Then put your current record definition into working-storage. |
|
| Back to top |
|
|
Bill O'Boyle
CICS Moderator
Joined: 14 Jan 2008
Posts: 2501
Location: Atlanta, Georgia, USA
|
|
|
|
| bijoybabu83 wrote: |
To Bill O'Boyle,
Can you please provide me some links to this updated method IMHO. |
If you search the Mainframe COBOL forum (as I had suggested), you'll get many hits.
If that doesn't work for you (to quote Robert), "Google is your Friend".
Many members will provide the path, but it's up to you to perform the research.... |
|
| Back to top |
|
|
sachin_star3
Warnings : 1
New User
Joined: 30 Sep 2006
Posts: 78
Location: pune
|
|
|
|
Please add one more filler add into the last filler cluase.
filller pic x(18396).
then it will match the logic proper way. |
|
| Back to top |
|
|
bijoybabu83
New User
Joined: 15 Jan 2007
Posts: 36
Location: Kerala
|
|
|
|
| Please let me know why do u want me to add the filler as u said - filller pic x(18396). |
|
| Back to top |
|
|
bijoybabu83
New User
Joined: 15 Jan 2007
Posts: 36
Location: Kerala
|
|
|
|
Hi I have one find for you...Something like a doubt.
While defining the fd entry for the variable byte format input file, I defined the input layout as shown :-
RECORDING MODE IS V
BLOCK CONTAINS 0 RECORDS
RECORD CONTAINS 1636 TO 20996 CHARACTERS
LABEL RECORDS ARE STANDARD
DATA RECORD IS IN-SUB-XTRCT-REC.
01 IN-SUB-XTRCT-REC.
05 IN-FIRST-BYTE PIC X(01).
05 FILLER PIC X(04).
05 IN-TOTAL-COUNT PIC X(07).
05 FILLER PIC X(12).
05 IN-SUB-NUMBER PIC X(0013).
05 FILLER PIC X(1594).
05 IN-DEPENDENT-CNT PIC 9(5).
88 IN-DEP-CNT-VALID VALUE 0 THRU 20.
05 IN-DEPENDENT-DATA OCCURS 0 TO 20 TIMES DEPENDING ON
IN-DEPENDENT-CNT INDEXED BY PINIDX.
10 FILLER PIC X(0968).
As you can see the variable part of the input file is determined by the value of IN-DEPENDENT-CNT in the input file. It is defined as a decimal 9(05) field. Whereas the in the input file, there are header records which does have spaces in place of IN-DEPENDENT-CNT. Is this the reason why it is failing with the length error. |
|
| Back to top |
|
|
dbzTHEdinosauer
Global Moderator
Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne
|
|
|
|
I assume that IN-FIRST-BYTE is used to determine the type of record: header, detail ???
second, if you have two distinct types of records:- one with the variable length table - detail record?
- one without - header record
if you create two 01's in your FD -- one with the variable length table - detail record?
- one without - header record
you may have better luck.
this is a very easy task and for someone who has no idea of what he is doing,
you are sure dismissive of suggestions. |
|
| Back to top |
|
|
|
|
