Skip to main content
Back to path

Welcome to a Grasp lesson

Create a free account to keep studying and create your own study path.

Lesson 1

COBOL Program Structure and Data Types

Recommended resources

Hello and welcome to your specialized mainframe developer program. We're kicking things off with a focused refresher on COBOL, designed to quickly reactivate your foundational knowledge before we dive into advanced scenarios and modernization topics.

This first lesson serves as a rapid review of the fundamental building blocks of a COBOL program. Our goal is to ensure you can confidently recall the overall program structure, the roles of its divisions and sections, and the syntax for defining data using standard PICTURE clauses. This will create a solid baseline for the more complex COBOL challenges we'll tackle in the lessons to come.

The Anatomy of a COBOL Program

As you know, a COBOL program is rigidly structured into a hierarchy of four divisions. While you've written countless programs, a quick, formal recap ensures we're aligned on the terminology, which is often a target for interview questions.

  1. IDENTIFICATION DIVISION: The program's metadata—PROGRAM-ID, AUTHOR, DATE-WRITTEN. It serves as the program's business card.
  2. ENVIRONMENT DIVISION: Defines the program's external environment. Its most critical part is the FILE-CONTROL section, which maps the logical file names used in the program (SELECT statements) to the system's physical datasets (JCL DD names).
  3. DATA DIVISION: The core of this lesson. It's where every single piece of data the program will touch—files, variables, constants, and parameters—is meticulously described.
  4. PROCEDURE DIVISION: Contains the executable code—the logic, the algorithms, the business rules—that manipulates the data defined in the DATA DIVISION.

This strict, hierarchical structure was born from the era of 80-column punch cards. Although modern IDEs have made development far more fluid, the underlying column-based format of fixed-format COBOL remains.

This diagram illustrates the traditional 80-column layout for a COBOL source line. Margins A (cols 8-11) and B (cols 12-72) are the primary areas for code, with Margin A being reserved for division, section, and paragraph names, as well as 01-level data definitions. The Indicator field (col 7) is famously used for comments (`*`).

A Deep Dive into the DATA DIVISION

The DATA DIVISION is where the contract between your program's logic and its data is formally written. It is itself divided into several key sections. To quickly review their purposes, please read the following sections from the "COBOL DATA DIVISION - Quick Reference" on MainframeMaster.

COBOL DATA DIVISION - Quick Reference - MainframeMaster

This reading will refresh your memory on the purpose and structure of the DATA DIVISION and its primary sections.

Start by reading the DATA DIVISION Overview to re-ground yourself in its overall purpose. Then, pay close attention to the table in the DATA DIVISION Sections subsection, which provides a concise summary of the four key sections.

As the reading highlights, the four main sections each serve a distinct and crucial role:

  • FILE SECTION: This is where you define the record layout for every file your program reads from or writes to. Each FD (File Description) entry here must correspond to a SELECT statement in the ENVIRONMENT DIVISION.
  • WORKING-STORAGE SECTION: This is the program's scratchpad. It holds all the internal data items that are not part of a file record—counters, flags, accumulators, report headings, and intermediate calculation results. Data defined here is typically initialized once when the program starts and persists for the life of the program.
  • LINKAGE SECTION: Essential for modular programming. This section defines the data layout for parameters passed to a subprogram from a calling program. It acts as a template, or a "view," over the calling program's memory, rather than allocating its own storage.
  • LOCAL-STORAGE SECTION: A more modern feature, this section also defines data for subprograms. However, unlike WORKING-STORAGE, its data is allocated and re-initialized each time the subprogram is called. This makes it ideal for re-entrant or recursive programming, a concept that becomes critical when discussing threadsafe CICS programs.

Defining Data: Level Numbers and PICTURE Clauses

Within these sections, data is defined using a combination of level numbers and PICTURE clauses. Level numbers (e.g., 01, 05, 10) create a hierarchy, grouping elementary data items into records. The PICTURE clause (abbreviated as PIC) specifies the type and size of an elementary item.

To review the specific syntax and common data types, please return to the MainframeMaster reference.

COBOL DATA DIVISION - Quick Reference - MainframeMaster

This reading focuses on the most critical syntax for data definition: the PICTURE clause and condition names.

First, read the section on PICTURE Clauses and carefully study the PICTURE Symbols table. This is the fundamental vocabulary for defining data. Next, review the section on Condition Names (88-Level Items). As a senior developer, using these to create self-documenting code is a hallmark of quality.

Let's synthesize these concepts. A typical WORKING-STORAGE entry combines these elements to create a structured record:

       01  WS-TRANSACTION-DATA.
           05  WS-TRANS-ID         PIC 9(10).
           05  WS-TRANS-CODE       PIC XX.
               88  SALE           VALUE 'SA'.
               88  RETURN         VALUE 'RE'.
           05  WS-TRANS-AMOUNT     PIC S9(7)V99.
           05  WS-TRANS-TIMESTAMP  PIC X(26).

In this example:

  • 01 WS-TRANSACTION-DATA is the group level, or record.
  • 05 level items are fields within that record.
  • PIC 9(10) defines a 10-digit unsigned numeric field.
  • PIC XX (or X(2)) defines a two-byte alphanumeric field.
  • The 88 levels (SALE, RETURN) are condition names that provide readable aliases for the values in WS-TRANS-CODE.
  • PIC S9(7)V99 defines a signed numeric field with 7 digits to the left and 2 digits to the right of an implied decimal point. The S is crucial for correctly handling negative values in computations.

To see all these concepts integrated into a single program structure, take a moment to scan the complete example provided in the MainframeMaster resource.

COBOL DATA DIVISION - Quick Reference - MainframeMaster

This final reading provides a full program context, showing how the different divisions and DATA DIVISION sections work together.

In the "Practical Examples" section, find the complete program. Don't analyze every line of logic; instead, observe how the FILE SECTION, WORKING-STORAGE SECTION, and LINKAGE SECTION are populated with 01 level records and various PIC clauses, and how the PROCEDURE DIVISION references them.

Conclusion

In this lesson, we've completed a high-speed refresher of the most fundamental concepts of COBOL program structure and data definition. You should now be able to confidently recall:

  • The purpose of the four main Divisions of a COBOL program.
  • The legacy 80-column structure that dictates code placement.
  • The distinct roles of the FILE, WORKING-STORAGE, LINKAGE, and LOCAL-STORAGE Sections within the DATA DIVISION.
  • The use of level numbers to define data hierarchies and PICTURE clauses to specify data type, size, and format.

This foundational vocabulary is essential. Now that we've re-established it, our next lesson will build directly upon it. We will move from defining data to manipulating it, covering advanced file handling techniques and the powerful string manipulation verbs: STRING, UNSTRING, and INSPECT.