@likec4/language-services

This package provides initialization of language services from given workspace directory, or in-memory from sources.

It is designed to be browser-compatible and bundled within the main likec4 package.

[!WARNING] This package is intended for internal use within other LikeC4 packages.

Please use the main likec4 package instead.

Features

• Browser and Node.js Support - Dual runtime compatibility with separate entry points
• Source Parsing - Parse LikeC4 DSL from strings or workspace directories
• Model Building - Build and validate architecture models from parsed sources
• View Computation - Compute views with predicates and element filtering
• Layout Engine - Graphviz-based automatic layout for diagrams
• Multi-project Support - Handle multiple LikeC4 projects in a workspace

Installation

npm install @likec4/language-services
# or
pnpm add @likec4/language-services

Usage

Browser Environment

import { fromSource, fromSources } from '@likec4/language-services/browser'

// From a single source string
const likec4 = await fromSource(`
  model {
    customer = person 'Customer'
    system = softwareSystem 'System'

    customer -> system 'uses'
  }
`)

// From multiple source files
const likec4 = await fromSources({
  'model.c4': 'model { ... }',
  'views.c4': 'views { ... }',
  /**
   * Optional, but allows to inject project config
   * with theme customization
   */
  'likec4.config.json': '{ "name": "project-name" }',
})

// Get computed model
const model = await likec4.computedModel()

// Get layouted diagrams
const diagrams = await likec4.diagrams()

Node.js Environment

import { fromWorkdir, fromWorkspace } from '@likec4/language-services/node'

// From a workspace directory
const likec4 = await fromWorkspace('./architecture')

// From current working directory
const likec4 = await fromWorkdir()

// Get layouted model
const model = await likec4.layoutedModel()

// Get all projects
const projects = likec4.projects()

[!NOTE] There is a separate entry point for node.js environment without MCP support: @likec4/language-services/node/without-mcp

Useful to reduce bundle size

API

Factory Functions

Browser

• fromSource(source: string): Promise<LikeC4> - Create from a single source string
• fromSources(sources: Record<string, string>): Promise<LikeC4> - Create from multiple source files

[!NOTE] fromSource and fromSources are available in both browser and node.js environments.

Node.js

• fromWorkspace(path: string, options?: FromWorkspaceOptions): Promise<LikeC4> - Create from a workspace directory
• fromWorkdir(options?: FromWorkspaceOptions): Promise<LikeC4> - Create from current working directory
• fromSources(sources: Record<string, string>, options?: InitOptions): Promise<LikeC4> - Create from multiple source files
• fromSource(source: string, options?: InitOptions): Promise<LikeC4> - Create from a single source string

LikeC4 Class

• computedModel(project?: ProjectId): Promise<LikeC4Model.Computed> - Build model with computed views (no layout)
• layoutedModel(project?: ProjectId): Promise<LikeC4Model.Layouted> - Build model with layout applied
• diagrams(project?: ProjectId): Promise<LayoutedView[]> - Get all layouted diagrams
• projects(): NonEmptyArray<ProjectId> - Get all project IDs
• validate(): Promise<ValidationResult> - Validate all documents
• export(format: ExportFormat): Promise<string> - Export to various formats

Runtime Compatibility

The package provides conditional exports for browser and Node.js environments:

• Browser: Uses in-memory workspace, no filesystem access
• Node.js: Full filesystem support, workspace scanning, and file watching

The appropriate version is automatically selected based on your environment.

Getting help

We are always happy to help you get started:

• Join Discord community – it is the easiest way to get help
• GitHub Discussions – ask anything about the project or give feedback

Contributors

Become a contributor

Support development

LikeC4 is a MIT-licensed open source project with its ongoing development made possible entirely by your support.
If you like the project, please consider contributing financially to help grow and improve it.
You can support us via OpenCollective or GitHub Sponsors.

License

This project is released under the MIT License