cobol-java-v3/benchmark-programs/31-validation-withdup/README.md

31-validation-withdup: Field Validation with Duplicate Check

电信业务场景

重复CDR检测。对已按CDR-ID排序的明细文件进行WS-PREV-KEY重复检测，首次出现通过，同一ID再次出现判定为重复并拒否。

Description

Reads a sorted input file and checks for duplicate KEY values using the WS-PREV-KEY pattern. The first occurrence of a key is written to FILE-OUT-GOOD. Any subsequent occurrence of the same key is written to FILE-OUT-BAD with an error code (duplicate detection).

Record Layout

Input / Good Output (35 bytes)

Field	Type	Length	Description
KEY	PIC X	5	Sort key
DATA	PIC X	30	Payload data

Bad Output (37 bytes)

Field	Type	Length	Description
KEY	PIC X	5	Duplicated key
DATA	PIC X	30	Original data
ERR	PIC X	2	Error code ('01')

Files

File	Purpose
main-31-validation-withdup.cbl	Main COBOL program
data-gen.sh	Generate sorted test data
run.sh	Compile, run, verify
README.md	This file

Test Data

KEY01  FIRST-RECORD         -> PASS (1st unique)
KEY02  SECOND-RECORD        -> PASS (2nd unique)
KEY03  FIRST-DUP-A          -> PASS (1st)
KEY03  SECOND-DUP-A         -> FAIL (dup)
KEY03  THIRD-DUP-A          -> FAIL (dup)
KEY04  UNIQUE-AFTER-DUP     -> PASS (unique after dups)
KEY05  DUP-GROUP-B-1ST      -> PASS (1st)
KEY05  DUP-GROUP-B-2ND      -> FAIL (dup)
KEY06  TRIPLE-GROUP-1ST     -> PASS (1st)
KEY06  TRIPLE-GROUP-2ND     -> FAIL (dup)
KEY06  TRIPLE-GROUP-3RD     -> FAIL (dup)
KEY99  LAST-UNIQUE          -> PASS (final unique)

Expected: 7 good records, 5 bad records.

Usage

cd 31-validation-withdup
bash data-gen.sh
bash run.sh

Expected Behavior

• The first occurrence of each KEY is always accepted.
• Subsequent occurrences of the same KEY are rejected as duplicates.
• The WS-PREV-KEY is updated only when a record passes validation.
• Duplicate records contain error code '01' in the BAD output.
• Input must be sorted by KEY for correct sequential detection.