Better signals api

GUIDE.md

@@ -25,3 +25,41 @@ func run(ctx context.Context, args []string) error {
 	return nil
 }
 ```
+
+## Manually trigger garbage collection
+
+Might be useful when something bad happened and you don't want to restart the binary/instance/container/etc.
+
+```go
+func triggerGC(ctx context.Context) {
+	appx.DoOnSignal(ctx, os.SIGUSR1, func(ctx context.Context) {
+		runtime.GC()
+	})
+}
+```
+
+## Reload config on SIGHUP
+
+Very popular pattern to add zero-restart configs to your application.
+
+```go
+type Config struct{
+	// some fields
+}
+
+func rereadConfig(ctx context.Context) (*Config, error) {
+	var cfg Config
+	var err error
+
+	appx.DoOnSignal(ctx, os.SIGHUP, func(ctx context.Context) {
+		err = json.Unmarshal(&cfg)
+		if err != nil{
+			println("oh, error :(")
+		}
+	})
+
+	return &cfg, err
+}
+```
+
+Took from archived [cristalhq/sigreload](https://github.com/cristalhq/sigreload)

appx.go

@@ -4,7 +4,6 @@ import (
 	"context"
 	"os"
 	"os/signal"
-	"runtime"
 	"syscall"
 	"time"
 )
@@ -22,21 +21,26 @@ func Uptime() time.Duration {
 	return time.Since(startTime)
 }
 
-// DoGCWhen one of the given signal happens (GC means runtime.GC()).
+// DoOnSignal runs fn on every signal.
 // Function is async, context is used to close underlying goroutine.
-// In most cases appx.DoGCWhen(os.SIGUSR1) should be enough.
-func DoGCWhen(ctx context.Context, signals ...os.Signal) {
-	ch := make(chan os.Signal, 1)
-	signal.Notify(ch, signals...)
+func DoOnSignal(ctx context.Context, signal os.Signal, fn func(ctx context.Context)) {
+	ch := newChan(signal)
 
 	go func() {
 		for {
 			select {
+			case <-ch:
+				fn(ctx)
 			case <-ctx.Done():
 				return
-			case <-ch:
-				runtime.GC()
 			}
 		}
 	}()
 }
+
+// newChan returns a channel triggered on every sig.
+func newChan(sig os.Signal) <-chan os.Signal {
+	ch := make(chan os.Signal, 1)
+	signal.Notify(ch, sig)
+	return ch
+}