fix(server): ensure that similar artists retrieved from provider are no more than limit (#4267)

* fix(provider): ensure that similar artists retreived from provider are no more than limit

* add overlimit multiplier

core/external/provider.go

@@ -560,7 +560,7 @@ func (e *provider) callGetSimilar(ctx context.Context, agent agents.ArtistSimila
 		return
 	}
 	start := time.Now()
-	sa, err := e.mapSimilarArtists(ctx, similar, includeNotPresent)
+	sa, err := e.mapSimilarArtists(ctx, similar, limit, includeNotPresent)
 	log.Debug(ctx, "Mapped Similar Artists", "artist", artist.Name, "numSimilar", len(sa), "elapsed", time.Since(start))
 	if err != nil {
 		return
@@ -568,7 +568,7 @@ func (e *provider) callGetSimilar(ctx context.Context, agent agents.ArtistSimila
 	artist.SimilarArtists = sa
 }
 
-func (e *provider) mapSimilarArtists(ctx context.Context, similar []agents.Artist, includeNotPresent bool) (model.Artists, error) {
+func (e *provider) mapSimilarArtists(ctx context.Context, similar []agents.Artist, limit int, includeNotPresent bool) (model.Artists, error) {
 	var result model.Artists
 	var notPresent []string
 
@@ -591,21 +591,33 @@ func (e *provider) mapSimilarArtists(ctx context.Context, similar []agents.Artis
 		artistMap[artist.Name] = artist
 	}
 
+	count := 0
+
 	// Process the similar artists
 	for _, s := range similar {
 		if artist, found := artistMap[s.Name]; found {
 			result = append(result, artist)
+			count++
+
+			if count >= limit {
+				break
+			}
 		} else {
 			notPresent = append(notPresent, s.Name)
 		}
 	}
 
 	// Then fill up with non-present artists
-	if includeNotPresent {
+	if includeNotPresent && count < limit {
 		for _, s := range notPresent {
 			// Let the ID empty to indicate that the artist is not present in the DB
 			sa := model.Artist{Name: s}
 			result = append(result, sa)
+
+			count++
+			if count >= limit {
+				break
+			}
 		}
 	}
 

core/external/provider_topsongs_test.go

@@ -42,10 +42,6 @@ var _ = Describe("Provider - TopSongs", func() {
 		p = NewProvider(ds, ag)
 	})
 
-	BeforeEach(func() {
-		// Setup expectations in individual tests
-	})
-
 	It("returns top songs for a known artist", func() {
 		// Mock finding the artist
 		artist1 := model.Artist{ID: "artist-1", Name: "Artist One", MbzArtistID: "mbid-artist-1"}
@@ -248,4 +244,31 @@ var _ = Describe("Provider - TopSongs", func() {
 		ag.AssertExpectations(GinkgoT())
 		mediaFileRepo.AssertExpectations(GinkgoT())
 	})
+
+	It("only returns requested count when provider returns additional items", func() {
+		// Mock finding the artist
+		artist1 := model.Artist{ID: "artist-1", Name: "Artist One", MbzArtistID: "mbid-artist-1"}
+		artistRepo.On("GetAll", mock.AnythingOfType("model.QueryOptions")).Return(model.Artists{artist1}, nil).Once()
+
+		// Mock agent response
+		agentSongs := []agents.Song{
+			{Name: "Song One", MBID: "mbid-song-1"},
+			{Name: "Song Two", MBID: "mbid-song-2"},
+		}
+		ag.On("GetArtistTopSongs", ctx, "artist-1", "Artist One", "mbid-artist-1", 1).Return(agentSongs, nil).Once()
+
+		// Mock finding matching tracks (both returned in a single query)
+		song1 := model.MediaFile{ID: "song-1", Title: "Song One", ArtistID: "artist-1", MbzRecordingID: "mbid-song-1"}
+		song2 := model.MediaFile{ID: "song-2", Title: "Song Two", ArtistID: "artist-1", MbzRecordingID: "mbid-song-2"}
+		mediaFileRepo.On("GetAll", mock.AnythingOfType("model.QueryOptions")).Return(model.MediaFiles{song1, song2}, nil).Once()
+
+		songs, err := p.TopSongs(ctx, "Artist One", 1)
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(songs).To(HaveLen(1))
+		Expect(songs[0].ID).To(Equal("song-1"))
+		artistRepo.AssertExpectations(GinkgoT())
+		ag.AssertExpectations(GinkgoT())
+		mediaFileRepo.AssertExpectations(GinkgoT())
+	})
 })