Functional Characteristics › File Matching

File Matching

To perform file matching, CA JARS Wizard processes data from a primary input file and up to three secondary files in a run. A secondary file can be either of the following:

• A table that is read into storage and searched as each primary input file record is processed.
• A data file, properly sorted, that is read record by record, and matched with records from the primary input file, based upon a common data field.

The primary input file should be defined by a glossary. Secondary files do not have to be defined by a glossary since items can be referenced in general or by DEFINE statements in the run.

CA JARS Wizard has an fCCa-b notation to refer to data fields in secondary files. 2CCa-b defines positions in the first secondary file. 3CCa-b refers to the second, and 4CCa-b to the third. For example, 2CC20-25 refers to positions 20 to 25 in the first secondary file, and 3CC20-25 refers to the same positions in the second secondary file.

The keywords MATCHED and MISSING can be used in SELECT and COMPUTE statements to determine whether a match has occurred between the primary input file and the secondary files. Processing can then be directed accordingly, based upon this determination.

Example:

INPUT PRIMARY-INPUT-FILE
SELECT REC/TYPE EQ 'E'
INPUT FILE KEY=2CC1-4 MATCH=APPL CODE TABLE=2CC5-20
INPUT FILE KEY=3CC1-4 MATCH=APPL CODE
SELECT MATCHED1 AND MATCHED2 AND DIV CODE EQ 'T' AND MONTH EQ '05'
COMPUTE APPL NAME(K)(C)(16) EQ 2CC5-20,
        PROGRAMMER TIME(N)(4) EQ 3CC48-51
SORT APPL CODE
BREAK APPL CODE
LIST(SUM) APPL NAME, APPL CODE, PROGRAMMER TIME, ACTIVE TIME

Use this example when introducing secondary input to the run. When INPUT is used it sets up the "key" pointing to the secondary input file and also indicates what it should "MATCH" in the glossary. COMPUTE if matched will match based on what criteria you give it.

This example assumes the following:

• The first secondary file has an application code in positions 1-4 and a system development application name in positions 5-20.
• The second secondary file contains an application code in positions 1-4. Positions 48-51 contain the total number of programmer hours for the project for the month of May.
• The first secondary file doesn't require any sorting since the TABLE parameter initiates binary lookup procedure.
• The second secondary file, however, must be sorted in key sequence, (positions 1-4 of the record), prior to the CA JARS Wizard run.

The following processing actions are made in this example (the exact sequence of operations varies from run to run, according to the specific requirements for the run):

• The primary input file is sorted by application code (SORT statement). This sort is necessary because of the record-for-record matching with the secondary input file.
• The entire first secondary file is read into storage and sorted. The TABLE parameter in this file's INPUT statement tells CA JARS Wizard that it is a table lookup file. A binary search is used to locate matches.
• A record is read from the primary input file and evaluated based upon the selection criteria. The record is selected if the DIV CODE is a T, the month portion of the date field is 05, and CA JARS Wizard is able to find matching records in both of the secondary files.
• CA JARS Wizard can now reference two data fields of records where a match occurred with the two secondary files. They are the application name field in positions 5-20 of the first secondary file and the programmer time field in positions 48-51 of the second secondary file. The COMPUTE statement moves to these fields.
• Normal processing is performed for the primary input record.
• The primary file could contain multiple records for the same application code. Consequently, the application name requires a (K) suffix, because character fields that are not break items are not printed on total lines (this is the default). The programmer time work field is defined as numeric data so that it is accumulated as a volume and displayed in the summary report print line.

The product of this example would be a summary report with the month's totals for each application code. The line would contain the application name, the application code, the total programmer hours from the tape file, and total active time from all records for the application in the file.

Note: For more information about multiple file processing, see the INPUT, SELECT, and COMPUTE statement discussions in the "Control Statements" chapter.