gpui-testing

gpui-testing

Testing GPUI applications and components. Use when writing tests, testing async operations, simulating user input, or debugging test failures.

4estrellas
0forks
Actualizado 1/16/2026
SKILL.md
readonlyread-only
name
gpui-testing
description

Testing GPUI applications and components. Use when writing tests, testing async operations, simulating user input, or debugging test failures.

GPUI Testing

This skill covers testing patterns for GPUI applications.

Overview

GPUI provides testing utilities for:

  • TestAppContext: Test harness for GPUI apps
  • Async test execution: with run_until_parked()
  • User input simulation: clicks, key presses
  • Entity testing: Creating and interacting with test entities

Test Setup

Basic Test Structure

#[cfg(test)]
mod tests {
    use super::*;
    use gpui::*;

    #[gpui::test]
    async fn test_my_view(cx: &mut TestAppContext) {
        // Test code here
    }
}

Creating Test Entities

#[gpui::test]
async fn test_counter(cx: &mut TestAppContext) {
    let counter = cx.new(|_| Counter { count: 0 });
    
    assert_eq!(counter.read(cx).count, 0);
}

Async Testing

Using run_until_parked

#[gpui::test]
async fn test_async_operation(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    view.update(cx, |view, cx| {
        view.start_async_work(cx);
    });
    
    // Wait for all async work to complete
    cx.background_executor.run_until_parked();
    
    // Check results
    assert_eq!(view.read(cx).status, "Complete");
}

GPUI Timers in Tests

IMPORTANT: Use GPUI executor timers, not smol::Timer:

#[gpui::test]
async fn test_with_delay(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    // ✅ CORRECT - Use GPUI timer
    cx.background_executor.timer(Duration::from_secs(1)).await;
    
    // ❌ WRONG - Don't use smol::Timer
    // smol::Timer::after(Duration::from_secs(1)).await;
    
    cx.background_executor.run_until_parked();
}

Why: GPUI's scheduler tracks GPUI timers but not smol::Timer, which can cause "nothing left to run" errors in run_until_parked().

Testing Entity Updates

Update and Verify

#[gpui::test]
async fn test_increment(cx: &mut TestAppContext) {
    let counter = cx.new(|_| Counter { count: 0 });
    
    counter.update(cx, |counter, cx| {
        counter.increment(cx);
    });
    
    assert_eq!(counter.read(cx).count, 1);
}

Testing Notify

#[gpui::test]
async fn test_notify(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    let observed = Arc::new(AtomicBool::new(false));
    let observed_clone = observed.clone();
    
    cx.observe(&view, move |_view, _cx| {
        observed_clone.store(true, Ordering::SeqCst);
    });
    
    view.update(cx, |view, cx| {
        view.data = "changed".into();
        cx.notify();
    });
    
    assert!(observed.load(Ordering::SeqCst));
}

Testing Actions

Dispatching Actions

actions!(test, [TestAction]);

#[gpui::test]
async fn test_action_handling(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    // Simulate action dispatch
    view.update(cx, |view, cx| {
        view.handle_action(&TestAction, cx);
    });
    
    assert_eq!(view.read(cx).action_count, 1);
}

Testing Subscriptions

Event Emission

#[derive(Clone, Debug)]
enum MyEvent {
    ValueChanged(i32),
}

impl EventEmitter<MyEvent> for MyView {}

#[gpui::test]
async fn test_event_emission(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    let received_events = Arc::new(Mutex::new(Vec::new()));
    let events_clone = received_events.clone();
    
    cx.subscribe(&view, move |_this, _emitter, event, _cx| {
        events_clone.lock().unwrap().push(event.clone());
    });
    
    view.update(cx, |view, cx| {
        cx.emit(MyEvent::ValueChanged(42));
    });
    
    let events = received_events.lock().unwrap();
    assert_eq!(events.len(), 1);
}

Testing Async Operations

Background Task Completion

#[gpui::test]
async fn test_background_task(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    view.update(cx, |view, cx| {
        cx.spawn(async move |this, cx| {
            let result = cx.background_spawn(async {
                // Expensive computation
                42
            }).await;
            
            this.update(&mut *cx, |view, cx| {
                view.result = Some(result);
                cx.notify();
            })?;
            
            Ok(())
        }).detach();
    });
    
    // Wait for all async work
    cx.background_executor.run_until_parked();
    
    assert_eq!(view.read(cx).result, Some(42));
}

Testing Task Cancellation

#[gpui::test]
async fn test_task_cancellation(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    view.update(cx, |view, cx| {
        view.start_long_task(cx);
    });
    
    // Cancel task
    view.update(cx, |view, cx| {
        view.cancel_task();
    });
    
    cx.background_executor.run_until_parked();
    
    // Task should not have completed
    assert_eq!(view.read(cx).task_completed, false);
}

Assertions and Expectations

Entity State Assertions

#[gpui::test]
async fn test_state_changes(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    // Initial state
    assert_eq!(view.read(cx).count, 0);
    
    // After update
    view.update(cx, |view, cx| {
        view.count = 10;
        cx.notify();
    });
    
    assert_eq!(view.read(cx).count, 10);
}

Using assert Macros

#[gpui::test]
async fn test_with_assertions(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    view.update(cx, |view, cx| {
        view.process_data(vec![1, 2, 3], cx);
    });
    
    cx.background_executor.run_until_parked();
    
    let view_state = view.read(cx);
    assert!(view_state.is_processed);
    assert_eq!(view_state.items.len(), 3);
    assert!(view_state.error.is_none());
}

Testing Patterns

Setup and Teardown

#[gpui::test]
async fn test_with_setup(cx: &mut TestAppContext) {
    // Setup
    let state = cx.new(|_| AppState::default());
    cx.set_global(state.clone());
    
    // Test
    let view = cx.new(|_| MyView::new());
    view.update(cx, |view, cx| {
        view.use_global_state(cx);
    });
    
    // Assertions
    assert!(view.read(cx).has_state);
    
    // Teardown is automatic when test ends
}

Testing Error Cases

#[gpui::test]
async fn test_error_handling(cx: &mut TestAppContext) {
    let view = cx.new(|_| MyView::new());
    
    view.update(cx, |view, cx| {
        cx.spawn(async move |this, cx| {
            // Simulate error
            let result: Result<(), anyhow::Error> = Err(anyhow::anyhow!("Test error"));
            
            this.update(&mut *cx, |view, cx| {
                view.error = result.err().map(|e| e.to_string());
                cx.notify();
            })?;
            
            Ok(())
        }).detach();
    });
    
    cx.background_executor.run_until_parked();
    
    assert!(view.read(cx).error.is_some());
}

Best Practices

  1. Use #[gpui::test] attribute: Required for GPUI tests
  2. Use GPUI timers: cx.background_executor.timer() instead of smol::Timer
  3. Call run_until_parked(): For async operations
  4. Test one thing at a time: Keep tests focused
  5. Use descriptive names: Test names should describe what they test
  6. Clean up resources: Though GPUI handles most cleanup automatically

Common Mistakes

Mistake Problem Solution
Using smol::Timer "Nothing left to run" Use cx.background_executor.timer()
Not calling run_until_parked() Async work doesn't complete Call before assertions
Forgetting #[gpui::test] Test doesn't run properly Use #[gpui::test] attribute
Not handling errors in async Test failures unclear Propagate errors with ?
Testing too much at once Hard to debug failures Split into smaller tests

Example Test Suite

#[cfg(test)]
mod tests {
    use super::*;
    use gpui::*;
    use std::sync::{Arc, atomic::{AtomicBool, Ordering}};

    #[gpui::test]
    async fn test_counter_initialization(cx: &mut TestAppContext) {
        let counter = cx.new(|_| Counter { count: 0 });
        assert_eq!(counter.read(cx).count, 0);
    }

    #[gpui::test]
    async fn test_counter_increment(cx: &mut TestAppContext) {
        let counter = cx.new(|_| Counter { count: 0 });
        
        counter.update(cx, |counter, cx| {
            counter.increment(cx);
        });
        
        assert_eq!(counter.read(cx).count, 1);
    }

    #[gpui::test]
    async fn test_counter_async_increment(cx: &mut TestAppContext) {
        let counter = cx.new(|_| Counter { count: 0 });
        
        counter.update(cx, |counter, cx| {
            cx.spawn(async move |this, cx| {
                // Simulate async delay
                cx.background_executor.timer(Duration::from_millis(100)).await;
                
                this.update(&mut *cx, |counter, cx| {
                    counter.count += 1;
                    cx.notify();
                })?;
                
                Ok(())
            }).detach();
        });
        
        cx.background_executor.run_until_parked();
        assert_eq!(counter.read(cx).count, 1);
    }

    #[gpui::test]
    async fn test_counter_notify(cx: &mut TestAppContext) {
        let counter = cx.new(|_| Counter { count: 0 });
        let notified = Arc::new(AtomicBool::new(false));
        let notified_clone = notified.clone();
        
        cx.observe(&counter, move |_counter, _cx| {
            notified_clone.store(true, Ordering::SeqCst);
        });
        
        counter.update(cx, |counter, cx| {
            counter.count = 5;
            cx.notify();
        });
        
        assert!(notified.load(Ordering::SeqCst));
    }
}

Summary

  • Use #[gpui::test] for GPUI tests
  • Use cx.background_executor.timer() for delays
  • Call run_until_parked() to complete async work
  • Test entity updates with .update() and .read()
  • Use Arc and Mutex for tracking callbacks
  • Avoid smol::Timer in tests

References

You Might Also Like

Related Skills

fix

fix

243Kdev-testing

Use when you have lint errors, formatting issues, or before committing code to ensure it passes CI.

facebook avatarfacebook
Obtener
peekaboo

peekaboo

179Kdev-testing

Capture and automate macOS UI with the Peekaboo CLI.

openclaw avataropenclaw
Obtener
frontend-testing

frontend-testing

128Kdev-testing

Generate Vitest + React Testing Library tests for Dify frontend components, hooks, and utilities. Triggers on testing, spec files, coverage, Vitest, RTL, unit tests, integration tests, or write/review test requests.

langgenius avatarlanggenius
Obtener
frontend-code-review

frontend-code-review

127Kdev-testing

Trigger when the user requests a review of frontend files (e.g., `.tsx`, `.ts`, `.js`). Support both pending-change reviews and focused file reviews while applying the checklist rules.

langgenius avatarlanggenius
Obtener
code-reviewer

code-reviewer

92Kdev-testing

Use this skill to review code. It supports both local changes (staged or working tree) and remote Pull Requests (by ID or URL). It focuses on correctness, maintainability, and adherence to project standards.

google-gemini avatargoogle-gemini
Obtener
session-logs

session-logs

90Kdev-testing

Search and analyze your own session logs (older/parent conversations) using jq.

moltbot avatarmoltbot
Obtener