Sample code for ut generation

.github/workflows/README-auto-tests.md

@@ -0,0 +1,240 @@
+# Auto-Generate Unit Tests Workflow
+
+This GitHub Actions workflow automatically generates unit tests for new or modified Go code in pull requests using AI (GitHub Copilot API with OpenAI fallback).
+
+## Features
+
+- 🤖 **AI-Powered Test Generation**: Uses GitHub Copilot or OpenAI to generate comprehensive unit tests
+- 🎯 **Targeted Coverage**: Runs tests only on affected packages, not the entire repo
+- 📊 **Coverage Reporting**: Shows coverage metrics for changed packages
+- ✅ **Manual Approval**: Requires explicit approval before committing generated tests
+- 🔄 **Automatic Updates**: Triggers on every PR update to regenerate tests for new changes
+
+## How It Works
+
+1. **Trigger**: Runs when a PR is opened, updated, or when `/approve-tests` comment is posted
+2. **Detection**: Identifies changed `.go` files (excluding test files and EXCLUDE_PACKAGES)
+3. **Extraction**: Parses Go AST to find exported functions in changed files
+4. **Generation**: Uses AI to generate unit tests following Go best practices
+5. **Testing**: Runs `go test` only on affected packages with coverage analysis
+6. **Review**: Posts PR comment with generated tests and coverage report
+7. **Approval**: Waits for manual approval via `/approve-tests` comment
+8. **Commit**: Commits approved tests to the PR branch
+
+## Setup
+
+### Required Permissions
+
+The workflow requires these GitHub token permissions (already configured):
+- `contents: write` - To commit generated tests
+- `pull-requests: write` - To comment on PRs
+- `issues: write` - To handle comment triggers
+
+### API Configuration
+
+#### Option 1: GitHub Copilot API (Recommended)
+
+Uses the built-in `GITHUB_TOKEN` - no additional setup required! The workflow will automatically use GitHub's Models API endpoint.
+
+**Advantages:**
+- ✅ No external API key needed
+- ✅ Integrated with GitHub
+- ✅ Included with GitHub Copilot subscription
+
+#### Option 2: OpenAI API (Fallback)
+
+If GitHub Copilot API is unavailable, add an OpenAI API key:
+
+1. Get an API key from [OpenAI Platform](https://platform.openai.com/api-keys)
+2. Add it as a repository secret:
+   - Go to: `Settings` → `Secrets and variables` → `Actions`
+   - Click `New repository secret`
+   - Name: `OPENAI_API_KEY`
+   - Value: Your OpenAI API key
+
+The workflow will automatically fall back to OpenAI if GitHub Copilot API fails.
+
+## Usage
+
+### For Contributors
+
+1. **Create a PR** with your Go code changes
+2. **Wait for workflow** to analyze changes and generate tests
+3. **Review the bot comment** showing generated tests and coverage
+4. **Approve tests** by commenting `/approve-tests` if satisfied
+5. **Tests are committed** automatically to your PR branch
+
+### Example Workflow
+
+```yaml
+# PR opened with changes to core/orchestrator_core.go
+
+→ Workflow detects changes
+→ Extracts new/modified functions
+→ Generates tests using AI
+→ Runs coverage on core package only
+→ Posts comment with results
+
+# You review and approve
+/approve-tests
+
+→ Tests committed to PR branch
+→ Workflow acknowledges approval
+```
+
+## Excluded Packages
+
+The following packages are excluded from test generation (matching Makefile EXCLUDE_PACKAGES):
+
+- `client/clientset`
+- `client/informers`
+- `client/listers`
+- `storage/external_test`
+- `astrads/api/v1alpha1`
+- `ontap/api/azgo`
+- `ontap/api/rest`
+- `fake` (any fake packages)
+- `mocks/` (any mock packages)
+- `operator/controllers/provisioner`
+- `storage_drivers/astrads/api/v1beta1`
+
+## Generated Test Quality
+
+The AI generates tests following these guidelines:
+
+✅ **Table-driven test patterns** where appropriate  
+✅ **Edge cases and error scenarios**  
+✅ **Go testing best practices**  
+✅ **Proper test function naming** (`TestFunctionName`)  
+✅ **Mocking for external dependencies**  
+✅ **Standard Go testing package** (compatible with existing tests)
+
+## Coverage Reporting
+
+The workflow provides:
+
+- **Per-package coverage** for affected packages
+- **Overall coverage** for changed code
+- **Non-blocking** - won't fail the PR, just reports metrics
+
+Example output:
+```
+## Coverage Report
+
+| Package | Coverage |
+|---------|----------|
+| core    | 75.3%    |
+| storage | 82.1%    |
+
+**Overall coverage for changed packages:** 78.7%
+```
+
+## Commands
+
+| Command | Action |
+|---------|--------|
+| `/approve-tests` | Approve and commit generated tests to PR branch |
+
+## Customization
+
+### Adjust AI Model
+
+Edit the workflow file and change the model parameter:
+
+```python
+# For GitHub Copilot API
+def call_github_copilot(prompt, model="gpt-4o"):  # Change model here
+
+# For OpenAI API
+def call_openai(prompt, model="gpt-4o-mini"):  # Change model here
+```
+
+### Modify Test Patterns
+
+Update the prompt in the `generate_tests.py` script:
+
+```python
+prompt = f"""Generate a comprehensive unit test...
+Requirements:
+1. Your custom requirement here
+2. Another custom requirement
+...
+```
+
+### Change Coverage Threshold
+
+Add coverage validation in the "Run tests on affected packages" step:
+
+```bash
+COVERAGE_THRESHOLD=70
+if (( $(echo "$OVERALL < $COVERAGE_THRESHOLD" | bc -l) )); then
+  echo "Warning: Coverage below ${COVERAGE_THRESHOLD}%"
+fi
+```
+
+## Troubleshooting
+
+### Tests not generated
+
+**Cause**: No exported functions found or all files excluded  
+**Solution**: Check that your changes include exported (capitalized) functions
+
+### API errors
+
+**Cause**: Rate limits or API unavailable  
+**Solution**: Workflow will try OpenAI fallback if configured
+
+### Coverage not calculated
+
+**Cause**: Test compilation errors or affected packages not found  
+**Solution**: Check workflow logs; tests may need manual fixes
+
+### Commit fails
+
+**Cause**: Insufficient permissions  
+**Solution**: Ensure workflow has `contents: write` permission
+
+## Workflow Triggers
+
+```yaml
+on:
+  pull_request:
+    types: [opened, synchronize, reopened]
+    paths:
+      - '**.go'
+      - '!**_test.go'
+  issue_comment:
+    types: [created]
+```
+
+## Performance
+
+- **Diff analysis**: ~5-10 seconds
+- **Test generation**: ~2-5 seconds per function
+- **Coverage calculation**: ~10-30 seconds (only affected packages)
+- **Total**: Typically < 2 minutes for most PRs
+
+## Best Practices
+
+1. **Review generated tests** - AI is good but not perfect
+2. **Add edge cases** - Supplement AI tests with domain-specific cases
+3. **Update existing tests** - Don't rely solely on generated tests
+4. **Commit incrementally** - Approve tests after each meaningful change
+
+## Limitations
+
+- Only generates tests for **exported functions** (public API)
+- May need manual adjustments for complex mocking scenarios
+- Coverage calculated only for **affected packages** (not entire repo)
+- Requires manual approval - no automatic commits
+
+## Support
+
+For issues or questions:
+1. Check workflow logs in Actions tab
+2. Review generated tests in PR comments
+3. Open an issue with workflow run link
+
+## License
+
+This workflow is part of the NetApp Trident project.