Skip to content
CDACDA Developer Portal

Quick Start

A working example end to end with the current library, from installation to a typed, validated result. No prior Archi or ArchiMate® knowledge needed.

1. Install

Section titled “1. Install”
Terminal window
npm install @cda/archi-semantic-core

Requirements: Node.js ^20 || ^22 || >=24, ESM-only, TypeScript types bundled in the package.

2. Load

Section titled “2. Load”

parseArchiModel accepts native Archi .archimate XML text. For a real file, read it from disk (or fetch it) and pass the XML string — the archive guide covers zip-archive .archimate files:

import { readFile } from 'node:fs/promises';
import { parseArchiModel } from '@cda/archi-semantic-core';


const xmlText = await readFile('MyModel.archimate', 'utf8');
const model = parseArchiModel(xmlText);

3. Inspect

Section titled “3. Inspect”

The parsed model is a typed ArchiModel with flat collections of folders, elements, relationships, views, diagram objects and connections:

model.elements.map((e) => e.type);      // ["BusinessActor", ...]
model.elements.map((e) => e.name);      // ["Customer", ...]
model.relationships.length;             // 2

Types are namespace-prefix-stripped ("BusinessActor", not "archimate:BusinessActor"), and cross-references such as relationship.sourceId are plain string ids — see IDs and references.

4. Result

Section titled “4. Result”

Run the structural validator on the parsed model — missing or duplicate ids and dangling references are reported as typed issues:

import { validateArchiModel } from '@cda/archi-semantic-core';


const { valid, errors } = validateArchiModel(model);


if (!valid) {
  for (const issue of errors) {
    console.error(`[${issue.code}] ${issue.path} — ${issue.message}`);
  }
  process.exit(1);
}

That is the whole loop: parse a native model, read it as typed data, and gate your pipeline on its structural integrity — the same way linting gates source code.

Next steps

Section titled “Next steps”