Snippet

Isolated Panic Catching and Hook Testing

Unit testing panic behavior in standard Rust requires intercepting output hooks and wrapping execution inside std::panic::catch_unwind safely.

snippet.rs

rust

use std::panic::{catch_unwind, set_hook, take_hook, AssertUnwindSafe};
use std::sync::atomic::{AtomicBool, Ordering};
use std::sync::Arc;
 
pub fn test_panic_resilience<F>(f: F) -> Result<(), String>
where
    F: FnOnce() + Send + 'static,
{
    let hook_called = Arc::new(AtomicBool::new(false));
    let hook_flag = Arc::clone(&hook_called);
 
    let prev_hook = take_hook();
    set_hook(Box::new(move |_info| {
        hook_flag.store(true, Ordering::SeqCst);
    }));
 
    let result = catch_unwind(AssertUnwindSafe(f));
    set_hook(prev_hook);
 
    if result.is_err() && hook_called.load(Ordering::SeqCst) {
        Ok(())
    } else {
        Err("Expected thread panic was not triggered or caught".to_string())
    }
}

Breakdown

pub fn test_panic_resilience<F>(f: F) -> Result<(), String>

Higher-order test runner accepting a closure to validate panic invocation.

let prev_hook = take_hook();

Saves existing global panic handler prior to registering custom hook.

let result = catch_unwind(AssertUnwindSafe(f));

Executes closure while catching unwinding thread panics without process termination.

set_hook(prev_hook);

Restores original global panic hook to maintain environment clean state.

Previous snippet Next snippet

More Expert Rust exercises →

Isolated Panic Catching and Hook Testing

Related