Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/orlovm/resultgroup

A lightweight library for handling results and errors from concurrent tasks, with optional error thresholds and context cancellation.
https://github.com/orlovm/resultgroup

concurrency errgroup golang

Last synced: about 2 months ago
JSON representation

A lightweight library for handling results and errors from concurrent tasks, with optional error thresholds and context cancellation.

Awesome Lists containing this project

README

        

# resultGroup

[![CI](https://github.com/orlovm/resultgroup/actions/workflows/go.yml/badge.svg)](https://github.com/orlovm/resultgroup/actions/workflows/go.yml)
[![codebeat badge](https://codebeat.co/badges/712c7ef7-4ac9-4df9-96b0-bd5c2d547e30)](https://codebeat.co/projects/github-com-orlovm-resultgroup-main)
[![GitHub release](https://img.shields.io/github/release/orlovm/resultgroup.svg?label=version)](https://github.com/orlovm/resultgroup/releases/latest)

resultGroup is a simple and flexible Go library for managing the results and errors of concurrent tasks. It is inspired by the beloved HashiCorp's [go-multierror/Group](https://github.com/hashicorp/go-multierror/blob/master/group.go).

## Motivation

The need to aggregate results from multiple sources is a common requirement in many applications. While libraries such as [errgroup](https://pkg.go.dev/golang.org/x/sync/errgroup) and [go-multierror](https://github.com/hashicorp/go-multierror/blob/master/group.go) can be used for this purpose, they often require additional boilerplate code and the use of mutexes or channels for synchronization. Result Group aims to streamline this process by offering a generic solution that minimizes boilerplate and simplifies the management of concurrent tasks.

## Usage

> **Note**
> ResultGroup works with Go 1.20 and is compatible with Go 1.20 wrapped errors.

To use Result Group, follow these steps:

1. Import the package:

```go
import "github.com/orlovm/resultgroup"
```

2. Create a new Result Group:

```go
group := resultgroup.Group[ResultType]{}
```

Replace `ResultType` with the type of the results you expect to collect.

3. Alternatively, create a new Result Group with an error threshold:

```go
ctx := context.Background()
threshold := 1
group, ctx := resultgroup.WithErrorsThreshold[ResultType](ctx, threshold)
```

4. Run concurrent tasks using the `Go` method:

```go
group.Go(func() ([]ResultType, error) {
// Your concurrent task logic here
})
```

5. Wait for all tasks to complete and collect the results:

```go
results, err := group.Wait()
```

`err` could be used as usual go 1.20 wrapped error, or be easily unwrapped with `Unwrap() []error`

Here's a complete example that demonstrates how to use Result Group to fetch data from multiple sources concurrently:

```go
package main

import (
"context"
"fmt"

"github.com/orlovm/resultgroup"
)

type Data struct {
Source string
Value int
}

func fetchData(source string) ([]Data, error) {
// Simulate fetching data from the source
data := []Data{
{Source: source, Value: 1},
{Source: source, Value: 2},
}

return data, nil
}

func main() {
sources := []string{"source1", "source2", "source3"}

ctx := context.Background()
group, _ := resultgroup.WithErrorsThreshold[Data](ctx, 1)

for _, source := range sources {
source := source
group.Go(func() ([]Data, error) {
return fetchData(source)
})
}

results, err := group.Wait()
if err != nil {
fmt.Println("Error:", err)
fmt.Println("Wrapped errors", err.Unwrap())
}

for _, result := range results {
fmt.Printf("Source: %s, Value: %d\n", result.Source, result.Value)
}
}
```

See [tests](https://github.com/orlovm/resultgroup/blob/main/group_test.go) for more examples.
## License

[MIT](LICENSE)