feat: Unify Unit Test Workflows with Dynamic Mocking Control

[mpaya5]

Fixes https://linear.app/genlayer-labs/issue/DXP-691/configure-github-actions-and-workflows-part1
Depends on #1329 / https://linear.app/genlayer-labs/issue/DXP-683/separate-re-group-unit-tests

Summary

This PR refactors the unit test workflows to use a single, unified workflow with conditional logic that determines whether to use mocks based on PR approval status. The system now uses a single environment variable TEST_WITH_MOCKS to control all mocking behavior.

Key Changes

🔄 Workflow Consolidation

Before:

• Two separate workflows: unit-tests-pr.yml and unit-tests-pr-approved.yml
• Manual configuration differences between workflows
• Duplicate job definitions

After:

• Single workflow: unit-tests-pr.yml with intelligent trigger detection
• Conditional logic determines mocking behavior
• DRY principle applied - no duplicate code

🎯 Unified Mocking Control

• Single variable: TEST_WITH_MOCKS controls all mocking (LLMs, web requests, external services)
• Replaces: TEST_WITH_MOCK_LLMS for better clarity and consistency
• Applied to: Both unit tests and integration tests for uniformity

🔧 Enhanced Test Fixtures

The tests/unit/conftest.py now provides comprehensive mocking:

• LLM Providers: OpenAI, Anthropic responses
• Web Requests: requests, urllib, aiohttp
• WebDriver: Selenium browser automation
• All controlled by the single TEST_WITH_MOCKS variable

Behavior

Regular PRs (opened, synchronized)

TEST_WITH_MOCKS=true

• ✅ All external services mocked
• ✅ Fast execution
• ✅ No API costs
• ✅ Deterministic results

Approved PRs

TEST_WITH_MOCKS=false

• ❌ Real API calls to LLM providers
• ❌ Real HTTP/HTTPS requests
• ✅ Full integration validation
• ✅ Production-like testing

Push to main

TEST_WITH_MOCKS=false  # or true, depending on requirements

• Runs with appropriate configuration
• Uploads to Codecov and SonarCloud

Trigger Logic

The workflow now intelligently determines when to run:

1. Pull Request Events: opened, synchronized, labeled
2. Pull Request Review: approved
3. Push to main: for coverage reporting
4. Manual dispatch: for debugging

Benefits

1. Maintainability: Single workflow to maintain instead of two
2. Consistency: Same variable name across all test types
3. Flexibility: Easy to adjust mocking behavior without code changes
4. Cost Optimization: Mocks for regular development, real APIs for validation
5. Performance: Faster feedback loop during development

Files Changed

• Deleted: .github/workflows/unit-tests-pr-approved.yml
• Modified: .github/workflows/unit-tests-pr.yml - Complete rewrite with trigger logic
• Modified: tests/unit/conftest.py - Updated to use TEST_WITH_MOCKS
• Modified: tests/unit/consensus/test_helpers.py - Updated variable name
• Modified: .env.example - Changed to TEST_WITH_MOCKS
• Modified: Integration test files for consistency

Testing

Run unit tests with mocking:

TEST_WITH_MOCKS=true pytest tests/unit/

Run unit tests without mocking:

TEST_WITH_MOCKS=false pytest tests/unit/

Migration Notes

• The variable TEST_WITH_MOCK_LLMS has been renamed to TEST_WITH_MOCKS
• All existing configurations should be updated to use the new variable name
• The behavior remains the same, just with better naming consistency

Summary by CodeRabbit

Release Notes

• New Features

  • Added real-time validator synchronization—validators automatically reload when configuration changes are detected.

• Bug Fixes

  • Improved robustness of concurrent validator operations to prevent race conditions.

• Tests

  • Consolidated unit test workflows with unified mocking controls for improved test consistency.

• Chores

  • Updated environment configuration and Docker image references for streamlined deployment.

[sonarqubecloud]

Quality Gate failed

Failed conditions
2 Security Hotspots
23.4% Coverage on New Code (required ≥ 80%)
C Reliability Rating on New Code (required ≥ A)

See analysis details on SonarQube Cloud

Catch issues before they fail your Quality Gate with our IDE extension SonarQube for IDE