spacedrive/core-new/docs/testing.md

Testing Guide

This document covers testing approaches and frameworks available in Spacedrive Core.

Overview

Spacedrive Core includes two test frameworks to handle different testing scenarios:

1. Standard Test Framework (test_framework) - For single-process unit and integration tests
2. Cargo Test Subprocess Framework (test_framework_new) - For multi-device networking tests

Cargo Test Subprocess Framework

When to Use

Use the cargo test subprocess framework when you need:

• Multi-device networking tests (pairing, file transfer, sync)
• Subprocess isolation for Core instances
• Parallel execution of different device roles
• Tests that simulate real network scenarios

How It Works

The framework uses cargo test itself as the subprocess executor:

1. Main Test: Orchestrates the overall test scenario
2. Device Scenarios: Individual test functions for each device role
3. Environment Coordination: Uses env vars to control which role runs
4. Process Management: Spawns and monitors cargo test subprocesses

Basic Structure

use sd_core_new::test_framework_new::CargoTestRunner;
use std::env;

// Device scenario - runs when TEST_ROLE matches
#[tokio::test]
#[ignore] // Only run when explicitly called
async fn alice_scenario() {
    // Exit early if not running as Alice
    if env::var("TEST_ROLE").unwrap_or_default() != "alice" {
        return;
    }
    
    let data_dir = PathBuf::from(env::var("TEST_DATA_DIR").expect("TEST_DATA_DIR required"));
    
    // ALL test logic for Alice goes here
    let mut core = Core::new_with_config(data_dir).await?;
    // ... complete test implementation
}

// Main orchestrator
#[tokio::test]
async fn test_multi_device_scenario() {
    let mut runner = CargoTestRunner::new()
        .with_timeout(Duration::from_secs(90))
        .add_subprocess("alice", "alice_scenario")
        .add_subprocess("bob", "bob_scenario");
    
    runner.run_until_success(|outputs| {
        // Check for success patterns in output
        outputs.get("alice").map(|out| out.contains("SUCCESS")).unwrap_or(false) &&
        outputs.get("bob").map(|out| out.contains("SUCCESS")).unwrap_or(false)
    }).await.expect("Test failed");
}

Environment Variables

The framework uses these environment variables for coordination:

• TEST_ROLE - Specifies which device role to run (alice, bob, etc.)
• TEST_DATA_DIR - Provides isolated temporary directory for each process

Process Communication

Processes coordinate through:

• File-based: Temporary files for sharing data (pairing codes, state)
• Output parsing: Success/failure patterns in stdout/stderr
• Timeouts: Automatic cleanup if tests hang

Example: Device Pairing Test

// Alice's role - all logic in test file
#[tokio::test]
#[ignore]
async fn alice_pairing_scenario() {
    if env::var("TEST_ROLE").unwrap_or_default() != "alice" { return; }
    
    let data_dir = PathBuf::from(env::var("TEST_DATA_DIR").expect("TEST_DATA_DIR required"));
    let mut core = Core::new_with_config(data_dir).await.unwrap();
    
    core.init_networking("test-password").await.unwrap();
    let (pairing_code, _) = core.start_pairing_as_initiator().await.unwrap();
    
    // Share pairing code with Bob
    std::fs::write("/tmp/pairing_code.txt", &pairing_code).unwrap();
    
    // Wait for Bob to connect
    loop {
        let devices = core.get_connected_devices().await.unwrap();
        if !devices.is_empty() {
            println!("PAIRING_SUCCESS: Alice connected");
            break;
        }
        tokio::time::sleep(Duration::from_secs(1)).await;
    }
}

// Bob's role - all logic in test file  
#[tokio::test]
#[ignore]
async fn bob_pairing_scenario() {
    if env::var("TEST_ROLE").unwrap_or_default() != "bob" { return; }
    
    let data_dir = PathBuf::from(env::var("TEST_DATA_DIR").expect("TEST_DATA_DIR required"));
    let mut core = Core::new_with_config(data_dir).await.unwrap();
    
    core.init_networking("test-password").await.unwrap();
    
    // Wait for Alice's pairing code
    let pairing_code = loop {
        if let Ok(code) = std::fs::read_to_string("/tmp/pairing_code.txt") {
            break code.trim().to_string();
        }
        tokio::time::sleep(Duration::from_millis(500)).await;
    };
    
    core.start_pairing_as_joiner(&pairing_code).await.unwrap();
    
    // Wait for connection
    loop {
        let devices = core.get_connected_devices().await.unwrap();
        if !devices.is_empty() {
            println!("PAIRING_SUCCESS: Bob connected");
            break;
        }
        tokio::time::sleep(Duration::from_secs(1)).await;
    }
}

Running Tests

Run the complete test:

cargo test test_device_pairing --nocapture

Debug individual scenarios:

# Run just Alice's scenario
TEST_ROLE=alice TEST_DATA_DIR=/tmp/test cargo test alice_pairing_scenario -- --ignored --nocapture

# Run just Bob's scenario
TEST_ROLE=bob TEST_DATA_DIR=/tmp/test cargo test bob_pairing_scenario -- --ignored --nocapture

Best Practices

1. Use #[ignore] on device scenario functions so they only run when explicitly called
2. Exit early if TEST_ROLE doesn't match to avoid unintended execution
3. Use clear success patterns in output for the condition function to detect
4. Clean up shared files in temporary locations to avoid test interference
5. Set appropriate timeouts based on the complexity of your test scenario
6. Use descriptive test names that clearly indicate the scenario being tested

Debugging

Check test output:

# See detailed output from both processes
cargo test test_device_pairing --nocapture

Run scenarios individually:

# Test Alice's logic in isolation
TEST_ROLE=alice TEST_DATA_DIR=/tmp/debug cargo test alice_scenario -- --ignored --nocapture

Common issues:

• Process hangs: Check timeout settings and success condition logic
• File conflicts: Ensure unique temporary file paths for concurrent tests
• Environment leakage: Make sure TEST_ROLE guards are working correctly

Standard Test Framework

For single-process tests, use the standard Rust testing approach:

#[tokio::test]
async fn test_core_initialization() {
    let core = Core::new().await.unwrap();
    assert!(core.device.device_id().is_ok());
}

Legacy Test Framework

The original test_framework with scenarios is still available but deprecated in favor of the cargo test subprocess approach for multi-device tests.

Writing New Tests

For Single Device Tests

Use standard #[tokio::test] functions.

For Multi-Device Tests

1. Create device scenario functions with #[ignore] and TEST_ROLE guards
2. Create a main orchestrator test using CargoTestRunner
3. Define clear success patterns for the condition function
4. Use appropriate timeouts and cleanup

File Organization

• Put tests in tests/ directory
• Use descriptive filenames: test_device_pairing.rs, test_file_sync.rs, etc.
• Keep all test logic in the test files themselves

Examples

See tests/core_pairing_test_new.rs for a complete example of the cargo test subprocess framework in action.