Testing Strategies

Relevant source files

The following files were used as context for generating this wiki page:

• tests/phpunit.xml
• jest.config.js
• tests/unit/cadLoader.spec.ts
• composer.json
• package.json
• AGENTS.md

The Nextcloud CAD Viewer project employs a multi-layered testing strategy to ensure the reliability of both its PHP backend and Vue.js frontend. The strategy encompasses unit testing, static analysis, and linting to maintain high code quality across the entire stack.

The primary goal is to validate core functionalities, such as CAD file format detection and application bootstrapping, while ensuring compatibility with Nextcloud 33 and PHP 8.2. Developers can refer to the Developer Guide for environment setup.

Backend Testing Architecture

The backend testing suite is primarily driven by PHPUnit and supplemented by static analysis tools. It focuses on the logic within the lib/ directory.

Unit Testing with PHPUnit

The project uses PHPUnit 10.4 for unit tests. The configuration is defined in tests/phpunit.xml, which specifies that tests are located in the tests/unit directory and should cover the lib source directory.

Sources: tests/phpunit.xml:1-16, composer.json:15

Static Analysis and Linting

To ensure type safety and adherence to coding standards, several tools are integrated:

• Psalm & PHPStan: Used for deep static analysis to find bugs without executing code.
• PHP CS Fixer: Enforces the Nextcloud coding standard.
• Linting: A custom script validates PHP syntax across the project (excluding vendor directories).

Sources: composer.json:13-18, composer.json:33-36, AGENTS.md:46-49

Test Execution Commands

Tool	Command	Purpose
PHPUnit	composer test:unit	Executes backend unit tests.
Psalm	composer psalm	Performs static analysis.
PHP CS Fixer	composer cs:check	Verifies coding standard compliance.
Syntax Lint	composer lint	Checks for PHP syntax errors.

Sources: composer.json:32-37, AGENTS.md:42-45

Frontend Testing Architecture

The frontend uses Jest as its primary test runner, configured specifically for a Vue 3 and TypeScript environment.

Jest Configuration

The jest.config.js file defines how the frontend tests are executed. Key configurations include:

• Environment: Uses jsdom to simulate a browser environment.
• Transforms: Employs @vue/vue3-jest for .vue files and babel-jest for JavaScript/TypeScript files.
• Mocks: Mocks core dependencies like @mlightcad/cad-viewer and element-plus to isolate components during testing.
• Globals: Provides global Nextcloud objects like OC and OCA that are expected at runtime.

Sources: jest.config.js:1-50, package.json:52-56

Frontend Test Flow

The following diagram illustrates how Jest processes a test file:

The diagram shows the transformation pipeline used by Jest to handle different file extensions before executing them in the jsdom environment.
Sources: jest.config.js:5-13, package.json:14

Automated Verification Targets

The project focuses its testing efforts on critical integration points and utility logic.

CAD Format Validation

A key area of testing is the cadLoader utility, which ensures that only supported MIME types (DWG and DXF) are processed by the viewer.

describe('isSupportedCADFormat', () => {
  it('should return true for DWG MIME types', () => {
    expect(isSupportedCADFormat('application/acad')).toBe(true);
    expect(isSupportedCADFormat('image/vnd.dwg')).toBe(true);
  });
});

Sources: tests/unit/cadLoader.spec.ts:5-10

Coverage and Quality Requirements

The project configuration includes strict requirements for test execution:

• PHP Version: Analysis and tests must target PHP 8.2.
• Platform Constraints: Backend tests may require specific Nextcloud core classes, which might lead to failures if not properly mocked in local environments.
• Frontend Coverage: Jest is configured to collect coverage from all .ts and .vue files in the src/ directory, excluding entry points like main.ts.

Sources: AGENTS.md:42-45, jest.config.js:42-47, composer.json:47-49

Continuous Integration Logic

The testing strategy is designed to be executed within GitHub Actions, following Nextcloud organization patterns.

The sequence diagram details the automated workflow triggered upon code submission, involving both backend and frontend validation steps.
Sources: AGENTS.md:51-54, composer.json:37

Conclusion

The testing strategy for Nextcloud CAD Viewer provides a robust framework for catching regressions. By combining PHPUnit for backend logic, Jest for frontend components, and comprehensive static analysis via Psalm and ESLint, the project ensures that the integration of the browser-based CAD viewer remains stable across updates to the Nextcloud ecosystem.