# Read work file with a variable length record **URL:** \ **Category:** Forum\ **Tags:** Adabas-Natural, Natural, Natural-on-Mainframes, Adabas-Natural-Cloud\ **Created:** [August 24, 2011, 7:32am UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132 "2011-08-24T07:32:38Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![Kevin\_Turner](https://avatars.discourse-cdn.com/v4/letter/k/22d042/32.png) [@Kevin\_Turner](https://techcommunity.softwareag.com/u/Kevin_Turner)\ **Post date:** [August 24, 2011, 7:32am UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/1 "2011-08-24T07:32:38Z")
I have a work file record I am trying to read. It is a variable length record archive of a database file over several years, so it cannot be loaded back in to Adabas easily. It contains 4 fixed fields: field1, field2, field3, field4 and 2 mu fields (field 5 p9(1:50) and field6 a4(1:50)). I’m having difficulty reading the record properly. I mostly get invalid data for numeric input field. There are no delimiters in the record. I can see a 1 byte mu-count fields in the dump of the file for field5 and field6 How do I read this as a variable length work file please. thank you Kevin ---
**Author:** ![system](https://global.discourse-cdn.com/techcommunity/original/3X/c/d/cd19967bb605d28a72d77eb83a1cae22f3df2eb8.svg) [@system](https://techcommunity.softwareag.com/u/system)\ **Post date:** [August 24, 2011, 12:27pm UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/2 "2011-08-24T12:27:04Z")
Assuming field1 to field4 are fixed length and you don’t run into a shifted data issue here right away you will do so as soon as field5 does not come with all 50 elements. You will have to define the record as an array of B1 fields, “parse” the record, find out how many field5 elements there are, move them to an array of P9 fields in the correct length and use that instead of the data out of the work file record directly. Then go on parsing as, assuming the record comes compressed, the field6 MU will immediately follow field5, etc. ---
**Author:** ![Douglas\_Kelly](https://sea1.discourse-cdn.com/techcommunity/user_avatar/techcommunity.softwareag.com/douglas_kelly/32/102_2.png) [@Douglas\_Kelly](https://techcommunity.softwareag.com/u/Douglas_Kelly)\ **Post date:** [August 24, 2011, 9:43pm UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/3 "2011-08-24T21:43:06Z")
The format sounds like the output of the Adabas decompression routine. You should be able to read the file with the Adabas compression utility and load the resulting file with the Adabas load utility into a database file. ---
**Author:** ![Henrialbert1](https://avatars.discourse-cdn.com/v4/letter/h/779978/32.png) [@Henrialbert1](https://techcommunity.softwareag.com/u/Henrialbert1)\ **Post date:** [August 31, 2011, 3:02am UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/4 "2011-08-31T03:02:33Z")
Try this sample (It works for me): ```auto 0010 DEFINE DATA LOCAL 0020 1 #W-RECORD1 (A32760) 0030 * 0040 1 #W-RECORD 0050 2 #FIELD1 (A) DYNAMIC 0060 2 #FIELD2 (A) DYNAMIC 0070 2 #FIELD3 (A) DYNAMIC 0080 2 #FIELD4 (A) DYNAMIC 0090 2 #FIELD5 (B9/1:*) 0100 2 #FIELD6 (A4/1:*) 0110 * 0120 1 #W-F5 (B1) 0130 1 #W-F6 (B1) 0140 * 0150 1 #I (P3) 0160 * 0170 1 #W-LIMF5 (P5) 0180 1 #W-LIMF6 (P5) 0190 * 0200 1 #LEN-F1 (N3) CONST <2> /* LENGTH OF FIELD1 0210 1 #LEN-F2 (N3) CONST <3> /* LENGTH OF FIELD2 0220 1 #LEN-F3 (N3) CONST <8> /* LENGTH OF FIELD3 0230 1 #LEN-F4 (N3) CONST <4> /* LENGTH OF FIELD4 0240 1 #LEN-F5 (N3) CONST <9> /* LENGTH OF FIELD5 0250 1 #LEN-F6 (N3) CONST <4> /* LENGTH OF FIELD6 0260 * 0270 1 #STRT-F1 (P5) /* ADDRESS OF FIELD1 0280 1 #STRT-F2 (P5) /* ADDRESS OF FIELD2 0290 1 #STRT-F3 (P5) /* ADDRESS OF FIELD3 0300 1 #STRT-F4 (P5) /* ADDRESS OF FIELD4 0310 1 #STRT-F5 (P5) /* ADDRESS OF FIELD5 0320 1 #STRT-F6 (P5) /* ADDRESS OF FIELD6 0330 * 0340 END-DEFINE 0350 * 0360 * INITIALISE FIELDS 0370 * 0380 #STRT-F1 := 1 /* ADDRESS OF FIELD1 0390 #STRT-F2 := #LEN-F1 + #STRT-F1 /* ADDRESS OF FIELD2 0400 #STRT-F3 := #LEN-F2 + #STRT-F2 /* ADDRESS OF FIELD3 0410 #STRT-F4 := #LEN-F3 + #STRT-F3 /* ADDRESS OF FIELD4 0420 * 0430 READ WORK 1 #W-RECORD1 0440 * 0450 * ====> GET FIXED FIELDS 1 TO 4 0460 * 0470 #FIELD1 := SUBSTR(#W-RECORD1, #STRT-F1, #LEN-F1) /* FIELD1 0480 #FIELD2 := SUBSTR(#W-RECORD1, #STRT-F2, #LEN-F2) /* FIELD2 0490 #FIELD3 := SUBSTR(#W-RECORD1, #STRT-F3, #LEN-F3) /* FIELD3 0500 #FIELD4 := SUBSTR(#W-RECORD1, #STRT-F4, #LEN-F4) /* FIELD4 0510 * 0520 WRITE '=' #FIELD1 (AL=2) 0530 WRITE '=' #FIELD2 (AL=3) 0540 WRITE '=' #FIELD3 (AL=8) 0550 WRITE '=' #FIELD4 (AL=4) 0560 * 0570 * ====> GET VAR FIELD 5 0580 * 0590 #STRT-F5 := #LEN-F4 + #STRT-F4 /* ADDRESS OF FIELD5 0600 #W-F5 := SUBSTR(#W-RECORD1, #STRT-F5, 1) /* FIELD5 MU COUNTER 0610 #W-LIMF5 := #W-F5 /* FIELD5 TOTAL OF OCURRENCES 0620 EXPAND ARRAY #FIELD5 TO (1: #W-LIMF5) 0630 * 0640 #STRT-F5 := #STRT-F5 + 1 /* REAL POSITION FIELD5 0650 FOR #I EQ 1 TO #W-LIMF5 0660 MOVE SUBSTRING(#W-RECORD1, #STRT-F5, #LEN-F5) TO #FIELD5(#I) 0670 WRITE '=' #FIELD5(#I) 0680 #STRT-F5 := #STRT-F5 + #LEN-F5 0690 END-FOR 0700 * 0710 * ====> GET VAR FIELD 6 0720 * 0730 #STRT-F6 := #STRT-F5 /* 0740 #W-F6 := SUBSTR(#W-RECORD1, #STRT-F6, 1) /* FIELD6 MU COUNTER 0750 #W-LIMF6 := #W-F6 /* FIELD6 TOTAL OF OCURRENCES 0760 EXPAND ARRAY #FIELD6 TO (1: #W-LIMF6) 0770 * 0780 #STRT-F6 := #STRT-F6 + 1 /* REAL POSITION FIELD6 0790 FOR #I EQ 1 TO #W-LIMF6 0800 MOVE SUBSTRING(#W-RECORD1, #STRT-F6, #LEN-F6) TO #FIELD6(#I) 0810 WRITE '=' #FIELD6(#I) 0820 #STRT-F6 := #STRT-F6 + #LEN-F6 0830 END-FOR 0840 END-WORK 0850 END ``` Note: lines 600-690 do the same that lines 740-830 ---
**Author:** ![system](https://global.discourse-cdn.com/techcommunity/original/3X/c/d/cd19967bb605d28a72d77eb83a1cae22f3df2eb8.svg) [@system](https://techcommunity.softwareag.com/u/system)\ **Post date:** [January 17, 2012, 10:38am UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/5 "2012-01-17T10:38:27Z")
One additional note here: If this is an unload created with ADAULD then using ADACMP DECOMPRESS with that unload file as the input to create a fixed length record (using the FORMAT option) may be the easier route. ---
**Author:** ![system](https://global.discourse-cdn.com/techcommunity/original/3X/c/d/cd19967bb605d28a72d77eb83a1cae22f3df2eb8.svg) [@system](https://techcommunity.softwareag.com/u/system)\ **Post date:** [April 2, 2021, 3:00pm UTC](https://techcommunity.softwareag.com/t/read-work-file-with-a-variable-length-record/98132/6 "2021-04-02T15:00:48Z")