test: cover SplitTagValue with multi-frame regression cases

Add tests pinning the behavior fixed by SplitTagValue iterating over each
input value. The previous len(values) != 1 short-circuit silently skipped
splitting whenever TagLib merged multiple ID3v2 frames into the same
property (e.g. TMOO + TXXX:MOOD for mood, or duplicate TIPL entries for
composer), as reported in #5065.

Three layers of coverage:
- model/tag_mappings_test.go: direct unit tests on TagConf.SplitTagValue
  covering single/multi-value input, case-insensitive separators, missing
  SplitRx, empty input, and the empty-strings-passed-through contract that
  the downstream metadata pipeline relies on.
- model/metadata/metadata_test.go: end-to-end check that a "mood" tag
  surfaced as two raw values (the exact shape from the bug report) is
  split, trimmed, and deduplicated to the expected three moods.
- model/metadata/map_participants_test.go: parallel multi-value case for
  the COMPOSER tag, ensuring the same fix also corrects multi-frame role
  parsing.

All three new specs fail on the pre-fix code and pass on the patched
SplitTagValue.

model/metadata/map_participants_test.go

@@ -684,6 +684,26 @@ var _ = Describe("Participants", func() {
 				Expect(composers[2].Name).To(Equal("The Album Artist"))
 			})
 		})
+
+		// Sibling fix to https://github.com/navidrome/navidrome/issues/5065: when
+		// multiple frames map to the same role tag (e.g. TIPL producer entries),
+		// the configured split separator must still apply to each value.
+		When("the tag has multiple values", func() {
+			It("should split each value individually", func() {
+				mf = toMediaFile(model.RawTags{
+					"COMPOSER": {"John Doe/Jane Doe", "Someone Else"},
+				})
+
+				participants := mf.Participants
+				Expect(participants).To(HaveKeyWithValue(model.RoleComposer, HaveLen(3)))
+				composers := participants[model.RoleComposer]
+				Expect(composers).To(ConsistOf(
+					HaveField("Name", "John Doe"),
+					HaveField("Name", "Jane Doe"),
+					HaveField("Name", "Someone Else"),
+				))
+			})
+		})
 	})
 
 	Describe("MBID tags", func() {

model/metadata/metadata_test.go

@@ -129,6 +129,21 @@ var _ = Describe("Metadata", func() {
 
 				Expect(md.Strings(model.TagGenre)).To(Equal([]string{"Rock", "Pop", "Punk"}))
 			})
+
+			// Regression test for https://github.com/navidrome/navidrome/issues/5065
+			//
+			// MP3s with both an ID3v2 TMOO frame and a TXXX:MOOD frame are surfaced by
+			// TagLib's PropertyMap as a single "mood" key with multiple values. The split
+			// configuration must still apply to each value individually.
+			It("should split values from multiple frames mapping to the same tag", func() {
+				props.Tags = model.RawTags{
+					// Same shape as the bug report: two frames, comma-separated content.
+					"mood": {"Love, Emotional, Ballad", "Love; Emotional; Ballad"},
+				}
+				md = metadata.New(filePath, props)
+
+				Expect(md.Strings(model.TagMood)).To(ConsistOf("Love", "Emotional", "Ballad"))
+			})
 		})
 
 		DescribeTable("Date",

model/tag_mappings_test.go

@@ -0,0 +1,64 @@
+package model
+
+import (
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("TagConf", func() {
+	Describe("SplitTagValue", func() {
+		var conf TagConf
+
+		BeforeEach(func() {
+			conf = TagConf{Split: []string{";", "/", ","}}
+			conf.SplitRx = compileSplitRegex("test", conf.Split)
+		})
+
+		It("splits a single value on configured separators", func() {
+			Expect(conf.SplitTagValue([]string{"Rock/Pop;Punk"})).To(Equal([]string{"Rock", "Pop", "Punk"}))
+		})
+
+		It("trims whitespace around split values", func() {
+			Expect(conf.SplitTagValue([]string{"Love, Emotional, Ballad"})).To(Equal([]string{"Love", "Emotional", "Ballad"}))
+		})
+
+		// Regression test for https://github.com/navidrome/navidrome/issues/5065
+		//
+		// When multiple ID3v2 frames map to the same logical tag (e.g. TMOO + TXXX:MOOD),
+		// TagLib's PropertyMap merges them into a slice with several entries. Previously
+		// SplitTagValue had a `len(values) != 1` guard that skipped splitting in this case.
+		It("splits each value individually when given multiple inputs", func() {
+			input := []string{"Love, Emotional, Ballad", "Love; Emotional; Ballad"}
+			Expect(conf.SplitTagValue(input)).To(Equal([]string{
+				"Love", "Emotional", "Ballad",
+				"Love", "Emotional", "Ballad",
+			}))
+		})
+
+		It("matches separators case-insensitively when the split pattern allows", func() {
+			c := TagConf{Split: []string{" AND "}}
+			c.SplitRx = compileSplitRegex("test", c.Split)
+			Expect(c.SplitTagValue([]string{"foo and bar AND baz"})).To(Equal([]string{"foo", "bar", "baz"}))
+		})
+
+		It("returns values unchanged when no separators are configured", func() {
+			c := TagConf{}
+			Expect(c.SplitTagValue([]string{"Foo, Bar"})).To(Equal([]string{"Foo, Bar"}))
+			Expect(c.SplitTagValue([]string{"a", "b"})).To(Equal([]string{"a", "b"}))
+		})
+
+		It("returns an empty slice for empty input", func() {
+			Expect(conf.SplitTagValue([]string{})).To(BeEmpty())
+		})
+
+		It("handles a value with no separator as a single-element result", func() {
+			Expect(conf.SplitTagValue([]string{"JustOneMood"})).To(Equal([]string{"JustOneMood"}))
+		})
+
+		It("produces empty strings when separators are adjacent (dedup happens downstream)", func() {
+			// SplitTagValue itself does not filter empties; that is the job of
+			// filterDuplicatedOrEmptyValues in the metadata pipeline.
+			Expect(conf.SplitTagValue([]string{"Rock//Pop"})).To(Equal([]string{"Rock", "", "Pop"}))
+		})
+	})
+})