Refactor random.WeightedChooser, unsing generics

core/external_metadata.go

@@ -267,8 +267,8 @@ func (e *externalMetadata) SimilarSongs(ctx context.Context, id string, count in
 		return nil, ctx.Err()
 	}
 
-	weightedSongs := random.NewWeightedRandomChooser()
+	weightedSongs := random.NewWeightedChooser[model.MediaFile]()
-	addArtist := func(a model.Artist, weightedSongs *random.WeightedChooser, count, artistWeight int) error {
+	addArtist := func(a model.Artist, weightedSongs *random.WeightedChooser[model.MediaFile], count, artistWeight int) error {
 		if utils.IsCtxDone(ctx) {
 			log.Warn(ctx, "SimilarSongs call canceled", ctx.Err())
 			return ctx.Err()
@@ -302,12 +302,12 @@ func (e *externalMetadata) SimilarSongs(ctx context.Context, id string, count in
 
 	var similarSongs model.MediaFiles
 	for len(similarSongs) < count && weightedSongs.Size() > 0 {
-		s, err := weightedSongs.GetAndRemove()
+		s, err := weightedSongs.Pick()
 		if err != nil {
 			log.Warn(ctx, "Error getting weighted song", err)
 			continue
 		}
-		similarSongs = append(similarSongs, s.(model.MediaFile))
+		similarSongs = append(similarSongs, s)
 	}
 
 	return similarSongs, nil

utils/random/weighted_random_chooser.go

@@ -2,42 +2,46 @@ package random
 
 import (
 	"errors"
+	"slices"
 )
 
-type WeightedChooser struct {
+// WeightedChooser allows to randomly choose an entry based on their weights
-	entries     []interface{}
+// (higher weight = higher chance of being chosen). Based on the subtraction method described in
+// https://eli.thegreenplace.net/2010/01/22/weighted-random-generation-in-python/
+type WeightedChooser[T any] struct {
+	entries     []T
 	weights     []int
 	totalWeight int
 }
 
-func NewWeightedRandomChooser() *WeightedChooser {
+func NewWeightedChooser[T any]() *WeightedChooser[T] {
-	return &WeightedChooser{}
+	return &WeightedChooser[T]{}
 }
 
-func (w *WeightedChooser) Add(value interface{}, weight int) {
+func (w *WeightedChooser[T]) Add(value T, weight int) {
 	w.entries = append(w.entries, value)
 	w.weights = append(w.weights, weight)
 	w.totalWeight += weight
 }
 
-// GetAndRemove choose a random entry based on their weights, and removes it from the list
+// Pick choose a random entry based on their weights, and removes it from the list
-func (w *WeightedChooser) GetAndRemove() (interface{}, error) {
+func (w *WeightedChooser[T]) Pick() (T, error) {
+	var empty T
 	if w.totalWeight == 0 {
-		return nil, errors.New("cannot choose from zero weight")
+		return empty, errors.New("cannot choose from zero weight")
 	}
 	i, err := w.weightedChoice()
 	if err != nil {
-		return nil, err
+		return empty, err
 	}
 	entry := w.entries[i]
-	w.Remove(i)
+	_ = w.Remove(i)
 	return entry, nil
 }
 
-// Based on https://eli.thegreenplace.net/2010/01/22/weighted-random-generation-in-python/
+func (w *WeightedChooser[T]) weightedChoice() (int, error) {
-func (w *WeightedChooser) weightedChoice() (int, error) {
+	if len(w.entries) == 0 {
-	if w.totalWeight == 0 {
+		return 0, errors.New("cannot choose from empty list")
-		return 0, errors.New("no choices available")
 	}
 	rnd := Int64(w.totalWeight)
 	for i, weight := range w.weights {
@@ -49,17 +53,18 @@ func (w *WeightedChooser) weightedChoice() (int, error) {
 	return 0, errors.New("internal error - code should not reach this point")
 }
 
-func (w *WeightedChooser) Remove(i int) {
+func (w *WeightedChooser[T]) Remove(i int) error {
+	if i < 0 || i >= len(w.entries) {
+		return errors.New("index out of bounds")
+	}
+
 	w.totalWeight -= w.weights[i]
 
-	w.weights[i] = w.weights[len(w.weights)-1]
+	w.weights = slices.Delete(w.weights, i, i+1)
-	w.weights = w.weights[:len(w.weights)-1]
+	w.entries = slices.Delete(w.entries, i, i+1)
-
+	return nil
-	w.entries[i] = w.entries[len(w.entries)-1]
-	w.entries[len(w.entries)-1] = nil
-	w.entries = w.entries[:len(w.entries)-1]
 }
 
-func (w *WeightedChooser) Size() int {
+func (w *WeightedChooser[T]) Size() int {
 	return len(w.entries)
 }

utils/random/weighted_random_chooser_test.go

@@ -5,35 +5,57 @@ import (
 	. "github.com/onsi/gomega"
 )
 
-var _ = Describe("WeightedRandomChooser", func() {
+var _ = Describe("WeightedChooser", func() {
-	var w *WeightedChooser
+	var w *WeightedChooser[int]
 	BeforeEach(func() {
-		w = NewWeightedRandomChooser()
+		w = NewWeightedChooser[int]()
 		for i := 0; i < 10; i++ {
-			w.Add(i, i)
+			w.Add(i, i+1)
 		}
 	})
 
-	It("removes a random item", func() {
+	It("selects and removes a random item", func() {
 		Expect(w.Size()).To(Equal(10))
-		_, err := w.GetAndRemove()
+		_, err := w.Pick()
 		Expect(err).ToNot(HaveOccurred())
 		Expect(w.Size()).To(Equal(9))
 	})
 
+	It("removes items", func() {
+		Expect(w.Size()).To(Equal(10))
+		for i := 0; i < 10; i++ {
+			Expect(w.Remove(0)).To(Succeed())
+		}
+		Expect(w.Size()).To(Equal(0))
+	})
+
+	It("returns error if trying to remove an invalid index", func() {
+		Expect(w.Size()).To(Equal(10))
+		Expect(w.Remove(-1)).ToNot(Succeed())
+		Expect(w.Remove(10000)).ToNot(Succeed())
+		Expect(w.Size()).To(Equal(10))
+	})
+
 	It("returns the sole item", func() {
-		w = NewWeightedRandomChooser()
+		ws := NewWeightedChooser[string]()
-		w.Add("a", 1)
+		ws.Add("a", 1)
-		Expect(w.GetAndRemove()).To(Equal("a"))
+		Expect(ws.Pick()).To(Equal("a"))
+	})
+
+	It("returns all items from the list", func() {
+		for i := 0; i < 10; i++ {
+			Expect(w.Pick()).To(BeElementOf(0, 1, 2, 3, 4, 5, 6, 7, 8, 9))
+		}
+		Expect(w.Size()).To(Equal(0))
 	})
 
 	It("fails when trying to choose from empty set", func() {
-		w = NewWeightedRandomChooser()
+		w = NewWeightedChooser[int]()
-		w.Add("a", 1)
+		w.Add(1, 1)
-		w.Add("b", 1)
+		w.Add(2, 1)
-		Expect(w.GetAndRemove()).To(BeElementOf("a", "b"))
+		Expect(w.Pick()).To(BeElementOf(1, 2))
-		Expect(w.GetAndRemove()).To(BeElementOf("a", "b"))
+		Expect(w.Pick()).To(BeElementOf(1, 2))
-		_, err := w.GetAndRemove()
+		_, err := w.Pick()
 		Expect(err).To(HaveOccurred())
 	})