f5-corkscrew

A TypeScript-based tool for extracting and parsing F5 BIG-IP TMOS configurations from various sources (.conf, .ucs, .qkview) into structured JSON format for analysis, migration, and automation workflows.

Features

• Multi-format Support: Parse .conf files, UCS archives, and qkview files
• String Input: Parse config directly from string (no file needed) - great for MCP servers
• Comprehensive Parsing: Universal recursive parser handles all TMOS object types
• Discovery APIs: List partitions and apps for efficient filtering
• Filtered Extraction: Extract apps by partition or specific names
• Structured Output: JSON format for easy consumption by automation tools
• CLI Tool: Simple command-line interface for quick analysis
• Library Usage: Import as npm package for programmatic use
• High Performance: Processes large configs (6MB, 13K objects) in ~20 seconds
• Excellent Test Coverage: 95.5% code coverage with comprehensive test suite

Quick Start

Installation

Global CLI Installation:

npm install -g f5-corkscrew

As Project Dependency:

npm install f5-corkscrew

Requirements

• Node.js: v22 or higher
• npm: v10 or higher

Check your versions:

node --version && npm --version

Usage

Command Line Interface

Basic Usage:

corkscrew --file /path/to/bigip.conf

Process UCS Archive:

corkscrew --file /path/to/backup.ucs > output.json

Process qkview with XML stats:

corkscrew --file /path/to/diagnostics.qkview --includeXmlStats > output.json

CLI Options:

• --file - Path to config file (.conf, .ucs, .qkview, .tar.gz)
• --no_sources - Exclude source config files from output (default: true)
• --no_file_store - Exclude filestore files (certs/keys) from output (default: true)
• --no_command_logs - Suppress command execution logs (default: true)
• --no_process_logs - Suppress parsing logs (default: true)
• --includeXmlStats - Include qkview XML statistics (default: false)

Programmatic Usage

import BigipConfig from 'f5-corkscrew';

async function parseConfig() {
  const bigip = new BigipConfig();

  // Load and parse config
  await bigip.loadParseAsync('/path/to/config.ucs');

  // Extract all applications
  const explosion = await bigip.explode();

  console.log(`Found ${explosion.config.apps.length} applications`);
  console.log(`Parsed in ${explosion.stats.parseTime}ms`);
}

Parse from String (NEW)

import BigipConfig from 'f5-corkscrew';

// Parse config from string (useful for MCP servers, APIs)
const configText = `#TMSH-VERSION: 15.1.0
ltm virtual /Common/app_vs {
    destination /Common/10.0.0.1:443
    pool /Common/app_pool
}`;

const bigip = new BigipConfig();
await bigip.loadParseString(configText);

// List partitions
const partitions = bigip.listPartitions();  // ['Common']

// List apps (optionally filter by partition)
const apps = bigip.listApps();              // ['/Common/app_vs']
const tenant1Apps = bigip.listApps('Tenant1');

// Extract specific apps
const appDetails = await bigip.apps({ partition: 'Tenant1' });
const specificApps = await bigip.apps({ apps: ['/Common/app_vs'] });

Discovery and Filtering API

import BigipConfig from 'f5-corkscrew';

const bigip = new BigipConfig();
await bigip.loadParseAsync('/path/to/config.ucs');
// Or: await bigip.loadParseString(configText);

// Discovery - find what's in the config
const partitions = bigip.listPartitions();      // ['Common', 'Tenant1', 'Tenant2']
const allApps = bigip.listApps();               // All virtual servers
const tenant1Apps = bigip.listApps('Tenant1');  // Filter by partition

// Lightweight summaries (good for display)
const summaries = bigip.listAppsSummary('Tenant1');
// [{ name: 'app_vs', fullPath: '/Tenant1/app_vs', partition: 'Tenant1', 
//    destination: '10.0.0.1:443', pool: '/Tenant1/app_pool' }]

// Filtered extraction - only dig what you need
const apps1 = await bigip.apps({ partition: 'Tenant1' });
const apps2 = await bigip.apps({ partitions: ['Tenant1', 'Tenant2'] });
const apps3 = await bigip.apps({ apps: ['/Common/vs1', '/Tenant1/vs2'] });

Working with Output

Using jq for JSON Processing:

# List all virtual servers
corkscrew --file config.ucs | jq '.output.config.apps[].name'

# Get specific application details
corkscrew --file config.ucs | jq '.output.config.apps[] | select(.name == "/Common/app1_vs")'

# Extract statistics
corkscrew --file config.ucs | jq '.output.stats'

Example Output Structure:

{
  "output": {
    "config": {
      "apps": [
        {
          "name": "/Common/app1_vs",
          "config": "ltm virtual /Common/app1_vs { ... }",
          "map": {
            "name": "/Common/app1_vs",
            "destination": "192.168.1.10:443",
            "pool": "/Common/app1_pool"
          }
        }
      ]
    },
    "stats": {
      "objectCount": 153,
      "objects": {
        "virtuals": 7,
        "pools": 7,
        "nodes": 10,
        "monitors": 6
      },
      "parseTime": 5478.3
    }
  }
}

Supported TMOS Objects

The universal parser handles all TMOS object types with full depth parsing. Key objects include:

LTM (Local Traffic Manager)

• Virtual Servers (with all nested properties)
• Pools & Pool Members (including FQDN nodes)
• Nodes
• Monitors (HTTP, HTTPS, TCP, and all types)
• Profiles (HTTP, TCP, Client-SSL, Server-SSL, and all types)
• iRules (with proper TCL bracket handling)
• Local Traffic Policies (LTPs)
• Persistence Profiles
• SNAT Pools
• Data Groups
• Virtual Addresses

GTM/DNS (Global Traffic Manager)

• Wide IPs (A, AAAA, CNAME, MX, etc.)
• Pools (all DNS record types)
• Servers
• Data Centers
• Regions
• Topology records

APM (Access Policy Manager)

• Access Profiles
• Access Policies

ASM/WAF (Application Security Manager)

• Security Policies

Security

• Bot Defense Profiles
• DoS Profiles

System

• Partitions and folders
• Global settings
• Provisioning
• And more...

Architecture

Core Components

• BigipConfig Class (src/ltm.ts) - Main parsing orchestrator and public API
• UniversalParse (src/universalParse.ts) - Recursive TMOS parser (full-depth)
• UnPacker (src/unPackerStream.ts) - Streams archives without full memory load
• DigConfigs (src/digConfigs.ts) - Application extraction (VS + dependencies)
• RegExTree (src/regex.ts) - Version-specific regex patterns
• XmlStats (src/xmlStats.ts) - qkview statistics processing

Data Flow

1. Input Processing

   • loadParseAsync(): Archives streamed via UnPacker, .conf files read directly
   • loadParseString(): Config text parsed directly (no file needed)

2. Universal Parsing (universalParse.ts)

   • Recursive bracket matching handles any nesting depth
   • iRule-aware parsing (proper TCL bracket handling)
   • Edge cases: multiline strings, pseudo-arrays, empty objects
   • Outputs hierarchical JSON structure

3. Discovery & Extraction

   • listPartitions(): Quick scan for unique partitions
   • listApps(): List virtual servers with optional partition filter
   • apps(): Full extraction with filter options
   • explode(): Complete extraction of all apps + metadata

See CLAUDE.md for detailed architecture documentation.

Development

Setup

# Clone repository
git clone https://github.com/f5devcentral/f5-corkscrew.git
cd f5-corkscrew

# Install dependencies
npm install

# Compile TypeScript
npm run compile

# Run tests
npm test

# Lint code
npm run lint

Project Scripts

• npm run compile - Compile TypeScript to JavaScript
• npm run watch - Watch mode for development
• npm test - Run test suite with coverage
• npm run lint - Run TypeScript compiler check and ESLint
• npm run build-package - Build and create npm package
• npm run build-code-docs - Generate TypeDoc documentation

Testing

Comprehensive test suite with 95.5% coverage:

# Run all tests
npm test

# Tests are organized by feature
tests/
├── 010_json_objects.test.ts      # Core object parsing
├── 020_unPacker.tests.ts         # Archive extraction
├── 030_dnsDetails.tests.ts       # GTM/DNS parsing
├── 037_ltmDetails.tests.ts       # LTM parsing
├── 040_waf.tests.ts              # ASM/WAF parsing
└── 050_conf_file.tests.ts        # Config file processing

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes with tests
4. Follow JSDoc documentation standards
5. Ensure tests pass (npm test)
6. Submit a pull request

See ENHANCEMENTS.md for improvement opportunities.

Performance

Typical Performance:

• 6MB config file
• ~300 virtual servers
• 223K lines
• 13K TMOS objects
• Processing Time: ~20 seconds

Memory Efficient:

• Streaming architecture for archives
• Minimal memory footprint
• Handles configs up to 50MB

Statistics Tracking:

• File sizes and object counts
• Parse time breakdown
• Performance metrics included in output

Documentation

• CHANGELOG.md - Version history and release notes
• CLAUDE.md - Architecture and development guide for AI assistants
• ENHANCEMENTS.md - Improvement opportunities and technical debt
• testingStats.md - Performance benchmarks

Version History

v1.6.0 (2025-12-31) - Universal Parser & MCP APIs

• Universal Parser: Replaced selective parsing with full-depth recursive parser from tmos-converter
• String Input: New loadParseString() method for parsing config text directly
• Discovery APIs: New listPartitions(), listApps(), listAppsSummary() methods
• Filtered Extraction: Enhanced apps() with partition and app name filters
• MCP-Friendly: APIs designed for AI agent workflows (drift detection, selective extraction)
• See PARSER_ANALYSIS.md for technical details

v1.5.0 (2025-10-30)

• Updated all dependencies to latest versions
• Upgraded to Node 22 and ES2022 target
• Major dependency updates: TypeScript 5.9.3, Mocha 11.7.4, ESLint 9.38.0
• Improved TypeScript type safety
• See CHANGELOG.md for complete details

Previous Versions

• v1.4.2 - Bug fixes for partition special characters
• v1.4.1 - Object counter bug fixes
• v1.4.0 - VS rank nesting and UCS parsing improvements

Support

Community Support:

• 📝 GitHub Issues - Report bugs or request features
• 💬 F5 DevCentral - Community discussions

Note: This is a community-supported project and is not officially supported by F5 Networks.

License

This project is licensed under the Apache License 2.0 - see LICENSE file for details.

Copyright

Copyright 2014-2025 F5 Networks Inc.

F5 Networks Contributor License Agreement

Before contributing to any project sponsored by F5 Networks, Inc. (F5) on GitHub, you will need to sign a Contributor License Agreement (CLA).

If you are signing as an individual, we recommend that you talk to your employer (if applicable) before signing the CLA since some employment agreements may have restrictions on your contributions to other projects.

If you are signing on behalf of a company, you represent that you are legally entitled to grant the license recited therein.

Community Code of Conduct

Please refer to the F5 DevCentral Community Code of Conduct.