COBOL file access

File access in COBOL is how you read and write records: in order (sequential), by key (random), or both (dynamic). The access mode you choose depends on the file organization and how you need to process the data. This page explains sequential, random, and dynamic access and when to use each.

The three access modes

Sequential access

With ACCESS MODE IS SEQUENTIAL you process records one after another. You OPEN the file, then READ repeatedly until end of file (or use START to position first). For sequential and line-sequential files this is the only option. For indexed or relative files, sequential access means reading in key order or relative record number order. Use it when you need to scan the whole file or from a given position to the end.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
       SELECT CUST-FILE ASSIGN TO 'CUSTOMER.DAT'
           ORGANIZATION IS SEQUENTIAL
           ACCESS MODE IS SEQUENTIAL.
 
       *> Read every record in order
       OPEN INPUT CUST-FILE
       PERFORM UNTIL no-more
           READ CUST-FILE
               AT END SET no-more TO TRUE
               NOT AT END
                   PERFORM PROCESS-RECORD
           END-READ
       END-PERFORM
       CLOSE CUST-FILE

Random access

With ACCESS MODE IS RANDOM you go directly to a record by key. For an indexed file you move the key value to the record key field, then READ (or WRITE/REWRITE/DELETE). For a relative file you set the relative key (record number) and then READ or WRITE. You do not read through previous records. Use random access when you need to look up or update a specific record by ID or key.

1
2
3
4
5
6
7
8
       *> Indexed file: read record with primary key = WS-KEY
       MOVE WS-KEY TO CUST-ID
       READ CUST-FILE
           INVALID KEY
               DISPLAY 'Not found'
           NOT INVALID KEY
               PERFORM PROCESS-RECORD
       END-READ

Dynamic access

With ACCESS MODE IS DYNAMIC you can do both. For example: READ by key to find a record, then READ NEXT to get the following records in key order. Or use START to position at a key, then READ NEXT in a loop. Only indexed and relative files support dynamic access. Use it when you need to jump to a key and then scan forward (or backward with READ PREVIOUS where supported).

Which file type supports which mode?

Sequential / line-sequential: SEQUENTIAL only. Indexed: SEQUENTIAL, RANDOM, or DYNAMIC. Relative: SEQUENTIAL, RANDOM, or DYNAMIC. Choose the organization (and key design) first; then choose the access mode that matches how you will read and update the file.

Step-by-step: choosing access mode

• If the file is sequential or line-sequential, use ACCESS MODE IS SEQUENTIAL.
• If the file is indexed or relative and you only need to read in order (or write in order), SEQUENTIAL is enough.
• If you need to read/update by key (or relative record number) without scanning, use RANDOM.
• If you need both “find by key” and “read next in order” in the same program, use DYNAMIC.

Explain like I'm five

Sequential is like reading a book from page 1 to the end. Random is like opening the book to one page number you know. Dynamic is like opening to a page, then turning pages one by one from there. The kind of “book” (file organization) decides which of these you can do.

Related concepts