Parser Explanation in docker-to-iac

A parser in docker-to-iac translates Docker configurations (either Docker run commands or docker-compose.yml files) into Infrastructure as Code (IaC) or One-Click Deploy templates. Each parser is designed to target a specific IaC language or cloud provider template format.

Input Types

docker-to-iac can process two types of input:

Docker Run Commands

docker run -d -p 8080:80 -e NODE_ENV=production nginx:latest

Docker Compose Files

version: '3'
services:
  web:
    image: nginx:latest
    ports:
      - "8080:80"
    environment:
      NODE_ENV: production

API

For detailed API documentation, see the parser API reference.

Default Parser Config

Each parser includes default configurations specific to its target cloud provider. These defaults are necessary because providers have different compute specifications and limitations.

Example: AWS Fargate has a minimum CPU allocation of 256, while DigitalOcean's minimum setting is 1 vCPU. The default parser config handles these provider-specific requirements.

To retrieve default parser configurations through the API, see the parser info documentation.

Parser vs. Language

The ParserInfo type separates variables between Provider and Language. This separation exists because some cloud providers support multiple IaC languages.

For example, AWS infrastructure can be defined using:

• CloudFormation
• AWS CDK (for TypeScript, Python, etc.)
• Terraform

When adding new parsers, consider whether multiple IaC languages are possible for your target provider. This affects how you name your parser file in src/parsers/<IAC_LANGUAGE_FILE_NAME>.ts. It's why the translate() method requires the target IaC language name (e.g., CFN) rather than the provider name (e.g., AWS).

Parser Implementation Notes

Creating parsers for multi-cloud IaC tools like Terraform presents additional challenges. Terraform's extensive provider ecosystem means a Terraform parser would need complex logic to handle various provider-specific implementations, making maintenance more difficult.

In contrast, single-provider languages like AWS CloudFormation have a one-to-one relationship with their cloud provider, simplifying parser implementation and maintenance.