Handling of individual task errors during concurrent execution

[1ch0]

Does the library currently provide a mechanism to continue execution of concurrent tasks even if some tasks encounter errors? I'm interested in a feature that allows all tasks to complete and then returns a summary of successes and failures. If this functionality exists, please point me to the relevant documentation.

[destel]

There are 3 ways to achieve this. Suppose you have a pipeline that returns a results channel in the end.

Go's for-range loop

results is a regular Go channel and can be processed with for-range. Just make sure you add a deferred DrainNB call, to prevent leaks in case of early return

defer rill.DrainNB(results)

for result := range results {
	if result.Error != nil {
		fmt.Println("Error:", result.Error)
	} else {
		fmt.Println("Value:", result.Value)
	}
}

using rill.Catch

resultsWoErrors := rill.Catch(results, 1, func(err error) error {
	fmt.Println("Error:", err)
	return nil // Do not pass this error to the next stage
})

_ = rill.ForEach(resultsWoErrors, 1, func(result int) error {
	fmt.Println("Value:", result)
	return nil
})

Go 1.23 iterators and rill.ToSeq2

This is similar to the first approach, but does not require the DrainNB call.There's a slight risk that ToSeq2 function will be deprecated and renamed in the future, since (value, error) iterators are not conventional.

for value, err := range rill.ToSeq2(results) {
	if err != nil {
		fmt.Println("Error:", err)
	} else {
		fmt.Println("Value:", value)
	}
}