Putting in thread because 2 sets of codes
#Day 1 solution on the mainframe
frosty blaze
JCL:
//NRSD225C JOB NRSD225C,
// '&OADOWNER',
// CLASS=S,
// LINES=(999,WARNING),
// MSGCLASS=R,
// TIME=1440
//JOBLIB DD DSN=ISDD0.BATCHUT.LOADLIB,
// DISP=SHR
// DD DSN=SYS1.SCEERUN,
// DISP=SHR
// DD DSN=SYS1.DSNDND1.SDSNLOAD,
// DISP=SHR
// DD DSN=SYS1.DSNDND1.SDSNLOD2,
// DISP=SHR
// DD DSN=NDRD0.NRSD225.PHASELIB,
// DISP=SHR
// DD DSN=NDRD0.NRSD225.LOADLIB,
// DISP=SHR
//*-------------------------------------------------------------------*
//* STEP 0: DELETE OLD FILES
//*-------------------------------------------------------------------*
//D01 EXEC PGM=IEFBR14
//DEL01 DD DSN=NDRD0.NRSD225.AOC.DAY1L,
// DISP=(OLD,DELETE,DELETE)
//DEL02 DD DSN=NDRD0.NRSD225.AOC.DAY1R,
// DISP=(OLD,DELETE,DELETE)
//*-------------------------------------------------------------------*
//* STEP 1: EXTRACT AND SORT LEFT LIST
//*-------------------------------------------------------------------*
//P10 EXEC PGM=SORT
//SYSUDUMP DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SORTIN DD DSN=NDRD0.NRSD225.AOC.DAY1,
// DISP=SHR
//SYSIN DD *
SORT FIELDS=(1,5,CH,A)
OUTREC FIELDS=(1,5)
//SORTOUT DD DSN=NDRD0.NRSD225.AOC.DAY1L,
// DISP=(NEW,CATLG,DELETE),
// UNIT=PERM,
// SPACE=(TRK,(2,1),RLSE),
// DCB=(RECFM=FB,LRECL=5,BLKSIZE=0)
//*-------------------------------------------------------------------*
//* STEP 2: EXTRACT AND SORT RIGHT LIST
//*-------------------------------------------------------------------*
//P20 EXEC PGM=SORT
//SYSUDUMP DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SORTIN DD DSN=NDRD0.NRSD225.AOC.DAY1,
// DISP=SHR
//SYSIN DD *
SORT FIELDS=(9,5,CH,A)
OUTREC FIELDS=(9,5)
//SORTOUT DD DSN=NDRD0.NRSD225.AOC.DAY1R,
// DISP=(NEW,CATLG,DELETE),
// UNIT=PERM,
// SPACE=(TRK,(2,1),RLSE),
// DCB=(RECFM=FB,LRECL=5,BLKSIZE=0)
//*-------------------------------------------------------------------*
//* STEP 3: RUN COBOL SOLUTION
//*-------------------------------------------------------------------*
//P30 EXEC PGM=DAY1
//SYSUDUMP DD SYSOUT=*
//ABENDAID DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//DAY1LIN DD DSN=NDRD0.NRSD225.AOC.DAY1L,
// DISP=SHR
//DAY1RIN DD DSN=NDRD0.NRSD225.AOC.DAY1R,
// DISP=SHR
COBOL
IDENTIFICATION DIVISION.
PROGRAM-ID. DAY1.
AUTHOR. Bobby.
DATE-WRITTEN. 02/12/24.
ENVIRONMENT DIVISION.
INPUT-OUTPUT SECTION.
FILE-CONTROL.
SELECT FILELIN ASSIGN TO DAY1LIN
FILE STATUS IS WS-INPUT-STATUS.
SELECT FILERIN ASSIGN TO DAY1RIN
FILE STATUS IS WS-INPUT-STATUS.
DATA DIVISION.
FILE SECTION.
FD FILELIN
BLOCK CONTAINS 1000 RECORDS
RECORDING MODE IS F
RECORD CONTAINS 5 CHARACTERS.
01 FILELIN-RECORD.
05 LEFT-NUM PIC 9(5).
FD FILERIN
BLOCK CONTAINS 1000 RECORDS
RECORDING MODE IS F
RECORD CONTAINS 5 CHARACTERS.
01 FILELIN-RECORD.
05 RIGHT-NUM PIC 9(5).
WORKING-STORAGE SECTION.
01 WS-INPUT-STATUS PIC X(2).
88 INPUT-OK VALUE "00".
88 INPUT-EOF VALUE "10".
88 INPUT-VALID VALUE "00", "10".
01 WS-EOF-POINT PIC X VALUE 'N'.
01 WS-L-COUNT-RECORDS PIC 9(5) VALUE 0.
01 WS-R-COUNT-RECORDS PIC 9(5) VALUE 0.
01 WS-TOTAL-DISTANCE PIC 9(9) VALUE 0.
01 WS-LEFT-LIST.
05 WS-LEFT-NUM PIC 9(5) OCCURS 1001 TIMES.
01 WS-RIGHT-LIST.
05 WS-RIGHT-NUM PIC 9(5) OCCURS 1001 TIMES.
01 WS-COUNTER PIC 9(5) VALUE 0.
01 WS-L PIC 9(5) VALUE 0.
01 WS-R PIC 9(5) VALUE 0.
01 WS-SIM-SCORE PIC 9(9) VALUE 0.
01 WS-TEMP PIC 9(9) VALUE 0.
01 WS-TEMP2 PIC 9(9) VALUE 0.
01 WS-DIFFERENCE PIC S9(5) VALUE 0.
01 WS-DIFFERENCE-ABS PIC 9(5) VALUE 0.
PROCEDURE DIVISION.
PROGRAM-CONTROL.
OPEN INPUT FILELIN.
PERFORM READ-LEFT-DATA UNTIL WS-EOF-POINT = 'Y'.
CLOSE FILELIN.
MOVE 'N' TO WS-EOF-POINT.
OPEN INPUT FILERIN.
PERFORM READ-RIGHT-DATA UNTIL WS-EOF-POINT = 'Y'.
CLOSE FILERIN.
DISPLAY 'LEFT RECORDS COUNT: ' WS-L-COUNT-RECORDS.
DISPLAY 'RIGHT RECORDS COUNT: ' WS-R-COUNT-RECORDS.
PERFORM PART-A.
PERFORM PART-B.
STOP RUN.
READ-LEFT-DATA.
READ FILELIN AT END MOVE 'Y' TO WS-EOF-POINT.
ADD 1 TO WS-L-COUNT-RECORDS.
MOVE LEFT-NUM TO WS-LEFT-NUM(WS-L-COUNT-RECORDS).
READ-RIGHT-DATA.
READ FILERIN AT END MOVE 'Y' TO WS-EOF-POINT.
ADD 1 TO WS-R-COUNT-RECORDS.
MOVE RIGHT-NUM TO WS-RIGHT-NUM(WS-R-COUNT-RECORDS).
PART-A.
PERFORM TRAVERSE-LISTS VARYING WS-COUNTER FROM 1 BY 1
UNTIL WS-COUNTER > 1000.
DISPLAY 'TOTAL DISTANCE: ' WS-TOTAL-DISTANCE.
TRAVERSE-LISTS.
SUBTRACT WS-LEFT-NUM(WS-COUNTER)
FROM WS-RIGHT-NUM(WS-COUNTER)
GIVING WS-DIFFERENCE.
IF (WS-DIFFERENCE > 0) THEN
ADD WS-DIFFERENCE TO WS-TOTAL-DISTANCE
ELSE
COMPUTE WS-DIFFERENCE-ABS = WS-DIFFERENCE * (-1)
ADD WS-DIFFERENCE-ABS TO WS-TOTAL-DISTANCE
END-IF.
PART-B.
PERFORM VARYING WS-L FROM 1 BY 1 UNTIL WS-L > 1000
MOVE 0 TO WS-TEMP
PERFORM VARYING WS-R FROM 1 BY 1 UNTIL WS-R > 1000
IF WS-LEFT-NUM(WS-L) = WS-RIGHT-NUM(WS-R)
ADD 1 TO WS-TEMP
END-IF
END-PERFORM
MULTIPLY WS-LEFT-NUM(WS-L) BY WS-TEMP GIVING WS-TEMP2
ADD WS-TEMP2 TO WS-SIM-SCORE
END-PERFORM.
DISPLAY 'SIMILARITY SCORE: ' WS-SIM-SCORE.
slate pollen
this looks very cool but I have no idea what it is
or, I have some idea but no details
frosty blaze
first one, "JCL" is "Job Control Language", it is a language used on mainframes to tell the system how to run programs/utilities/subsystems. In this case, my JCL is:
- Deleting the previous sorted files (have to do this or else the system will try to create new files that already exist, which it will fail on. alternatively, could've run the sort steps once then comment out that code, since they do the same thing every time. I just wanted to have something similar to real-world production-ready JCL)
- Running the sort utility to extract the left and right lists and sort them, and then put them into their own files
- Running my cobol program with those 2 files from the previous step as input files
And most people know what COBOL is, though it is a common misconception nowadays that new COBOL isn't written anymore - it most certianly is, and it's part of my job
frosty blaze
because mainframes are very much still around
but yes, the language itself is pretty awful and definately requires a unique way of thinking
I'm hoping there will be a challenge on a later day that requires large data processing because that's what mainframes excel at and I also have an extremely large pool of CPU cycles at my disposal
fading hamlet
I'm hoping there will be a challenge on a later day that requires large data pro...
bad news, https://adventofcode.com/about
Nor do you need a fancy computer; every problem has a solution that completes in at most 15 seconds on ten-year-old hardware
frosty blaze
Oh true
lilac cedar
"has a solution" does not imply an obvious solution
ok did discord change the font for italics again, it looks so trippy
fading hamlet
The problems with obvious suboptimal solutions that a regular PC can't brute force are unfortunately also the kind of problems where every computer on the planet put together probably couldn't brute force in a reasonable timeframe 😛
or they're very serial in nature
lilac cedar
back in my competition days the problems were designed to have several solutions of different complexity