z/OS Extensions

The z/OS extensions for Vim enable direct editing of MVS datasets with integration of z/OS–specific features. These extensions support dataset integrity, platform resource controls, and automatic syntax handling to provide a native editing experience on z/OS.

The extensions support the following high-level capabilities:

Direct editing of MVS datasets and members.
Enforcement of dataset attributes such as LRECL.
z/OS ENQ/DEQ resource serialisation.
Automatic syntax highlighting based on dataset naming conventions.

Opening and Editing Datasets

You can open and edit an MVS dataset member directly from Vim by using the following command:

vim "//'DATASET.NAME(MEMBER)'"

This command opens the specified dataset member in the current Vim window.

Supported Operations

:e, :split, :vsplit: Opens dataset members in the current window, horizontal split, or vertical split. Allows working with multiple datasets simultaneously in different windows.
Vim diff: Compares two dataset members side-by-side using vimdiff or :diffsplit. Highlights differences between versions, useful for code review and change tracking.
Read-only support: Automatically detects and displays [readonly] in the status line for datasets with read-only permissions. Prevents accidental modifications to protected datasets like SYS1.PARMLIB or SYS1.PROCLIB.
Syntax highlighting: Automatically applies syntax highlighting by mapping dataset names to Unix file extensions. For example, a dataset matching *.COBOL is treated as a .cob file, enabling COBOL syntax highlighting.
LRECL enforcement: Enforces the dataset's logical record length (LRECL) for fixed block (FB) datasets. Prevents adding characters beyond the record length boundary, maintaining dataset integrity.
Margin reference: Displays a visual margin indicator at the LRECL boundary. Helps users stay within the fixed record length and avoid accidental overflow, particularly useful for JCL and COBOL editing.
ENQ/DEQ support: Respects z/OS resource serialization. Vim checks for ENQ (enqueue) locks before writing to datasets, preventing concurrent modification conflicts and ensuring data integrity.

Examples

" Edit dataset
vim "//'PROD.JCL(MYJOB)'"

" Compare datasets
vimdiff "//'PROD.COBOL(PROG)'" "//'TEST.COBOL(PROG)'"

" View read-only system dataset
vim "//'SYS1.PROCLIB(JES2)'"

Configuration

The dataset behaviour can be customised using a JSON configuration file. The system searches for configuration files in the following order:

Path specified by the $DIO_CONFIG environment variable
$HOME/.dioconfig.json in the user’s home directory
Default configuration provided with the build

Example Configuration

{
  "*.COBOL": {
    "extension": "cob",
    "type": "text",
    "codepage": "IBM-037"
  },
  "*.JCL": {
    "extension": "jcl",
    "type": "text",
    "codepage": "IBM-1047"
  },
  "*.REXX": {
    "extension": "rexx",
    "type": "text",
    "codepage": "IBM-1047"
  },
  "*.ASM": {
    "extension": "asm",
    "type": "text",
    "codepage": "IBM-037"
  },
  "*.DATA": {
    "extension": "txt",
    "type": "text",
    "codepage": "IBM-1047"
  },
  "*.TEMPLATE": {
    "extension": "tmpl",
    "type": "text",
    "codepage": "IBM-1047"
  }
}

Configuration Fields


Field	Description
Pattern	Wildcard pattern (for example, *.COBOL) used to match dataset names.
extension	Unix file extension mapped to the dataset; determines Vim syntax highlighting.
type	Specifies how Vim should handle the dataset content. Dataset content type: text or binary.
codepage	Character encoding used for EBCDIC–ASCII conversion (for example, IBM-1047, IBM-037)

Default Configuration

If no user-specific configuration is found, Vim uses the default configuration provided with the build.You can override default behaviour by creating your own configuration file.

Limitations

The following limitations apply when using Vim with z/OS datasets:

Cannot allocate new datasets
Cannot browse PDSE like directories
Auto-complete does not work on datasets
Must respect dataset LRECL for fixed block formats