query-orchestration/test/README.md

Integration Tests

This directory contains two integration test implementations that test the same core API workflow but serve different purposes.

Test Files

text_extraction_integration_test.go - Automated CI/CD Testing

Purpose: Automated integration testing for CI/CD pipelines.

How it works:

• Self-contained using testcontainers (spins up PostgreSQL, S3 mock, and API server automatically)
• Uses the generated Go API client for type-safe requests
• Runs as part of task test:functional or task fullsuite:ci

When to use:

• Local development verification before committing
• CI/CD pipeline testing
• Automated regression testing

Run with:

go test -v ./test/... -run TestE2EDocWorkflow
# Or via task
task test:functional

---

text_extraction_integration_test.sh - Manual Testing

Purpose: Manual end-to-end testing against a running API server.

How it works:

• Uses curl/jq to make REST API calls
• Requires an external API server already running (auth bypass mode)
• Creates comprehensive test data with realistic healthcare contract values
• Demonstrates API usage for UI developers

When to use:

• Testing against local docker-compose stack
• Testing against AWS dev/UAT deployments
• Populating a development environment with sample data
• Demonstrating API endpoints to UI developers

Run with:

# Against local stack
./test/text_extraction_integration_test.sh http://localhost:8080

# Against AWS deployment
./test/text_extraction_integration_test.sh http://your-alb-url.us-east-2.elb.amazonaws.com

Requirements:

• bash 4.0+
• curl
• jq
• zip
• Target API must be running in auth bypass mode

---

Test Coverage

Both tests exercise the same core workflow:

1. Client Setup: Create, verify, list, update client
2. Document Upload: Batch ZIP upload, single document upload
3. Label Operations: Apply labels, verify, query by label
4. Folder Metrics: List folders, get metrics per folder
5. Field Extractions: Create, version history, retrieve specific versions

The Go test includes additional scenarios:

• Deep folder hierarchy (12 levels)
• Large array field extraction (100 items)
• Label reapplication history
• Cross-feature integration (folders + labels + field extractions)