feat: cross-script transliteration matching for Tidal/Qobuz and skip-downloaded option for CSV import

go_backend/matching_test.go

@@ -0,0 +1,59 @@
+package gobackend
+
+import "testing"
+
+func TestQobuzTitlesMatchCrossScript(t *testing.T) {
+	t.Run("rejects unrelated cross-script titles", func(t *testing.T) {
+		if qobuzTitlesMatch("パンツ脱げるもん！", "Warrior of the Darkness") {
+			t.Fatalf("expected unrelated cross-script titles to not match")
+		}
+	})
+
+	t.Run("accepts transliterated japanese title", func(t *testing.T) {
+		if !qobuzTitlesMatch("パンツ脱げるもん！", "Pantsu Nugeru Mon") {
+			t.Fatalf("expected transliterated japanese title to match")
+		}
+	})
+}
+
+func TestQobuzArtistsMatchCrossScript(t *testing.T) {
+	t.Run("rejects unrelated cross-script artists", func(t *testing.T) {
+		if qobuzArtistsMatch("TakeponG", "陳奕迅") {
+			t.Fatalf("expected unrelated cross-script artists to not match")
+		}
+	})
+
+	t.Run("accepts transliterated japanese artist", func(t *testing.T) {
+		if !qobuzArtistsMatch("たけぽんぐ", "takepong") {
+			t.Fatalf("expected transliterated japanese artist to match")
+		}
+	})
+}
+
+func TestTidalTitlesMatchCrossScript(t *testing.T) {
+	t.Run("rejects unrelated cross-script titles", func(t *testing.T) {
+		if titlesMatch("パンツ脱げるもん！", "Warrior of the Darkness") {
+			t.Fatalf("expected unrelated cross-script titles to not match")
+		}
+	})
+
+	t.Run("accepts transliterated japanese title", func(t *testing.T) {
+		if !titlesMatch("パンツ脱げるもん！", "Pantsu Nugeru Mon") {
+			t.Fatalf("expected transliterated japanese title to match")
+		}
+	})
+}
+
+func TestTidalArtistsMatchCrossScript(t *testing.T) {
+	t.Run("rejects unrelated cross-script artists", func(t *testing.T) {
+		if artistsMatch("TakeponG", "陳奕迅") {
+			t.Fatalf("expected unrelated cross-script artists to not match")
+		}
+	})
+
+	t.Run("accepts transliterated japanese artist", func(t *testing.T) {
+		if !artistsMatch("たけぽんぐ", "takepong") {
+			t.Fatalf("expected transliterated japanese artist to match")
+		}
+	})
+}

go_backend/qobuz.go

@@ -81,13 +81,160 @@ func qobuzArtistsMatch(expectedArtist, foundArtist string) bool {
 	expectedLatin := qobuzIsLatinScript(expectedArtist)
 	foundLatin := qobuzIsLatinScript(foundArtist)
 	if expectedLatin != foundLatin {
-		GoLog("[Qobuz] Artist names in different scripts, assuming match: '%s' vs '%s'\n", expectedArtist, foundArtist)
-		return true
+		if qobuzCrossScriptEquivalent(expectedArtist, foundArtist) {
+			GoLog("[Qobuz] Artist names in different scripts but transliteration matched: '%s' vs '%s'\n", expectedArtist, foundArtist)
+			return true
+		}
 	}
 
 	return false
 }
 
+func qobuzNormalizeScriptAware(value string) string {
+	normalized := strings.ToLower(strings.TrimSpace(value))
+	normalized = CleanToASCII(JapaneseToRomaji(normalized))
+	normalized = strings.Join(strings.Fields(normalized), " ")
+	return strings.TrimSpace(normalized)
+}
+
+func qobuzCrossScriptEquivalent(expected, found string) bool {
+	normExpected := qobuzNormalizeScriptAware(expected)
+	normFound := qobuzNormalizeScriptAware(found)
+
+	if normExpected == "" || normFound == "" {
+		return false
+	}
+
+	if normExpected == normFound {
+		return true
+	}
+
+	compactExpected := strings.ReplaceAll(normExpected, " ", "")
+	compactFound := strings.ReplaceAll(normFound, " ", "")
+	if len(compactExpected) >= 6 && len(compactFound) >= 6 {
+		if compactExpected == compactFound ||
+			strings.Contains(compactExpected, compactFound) ||
+			strings.Contains(compactFound, compactExpected) {
+			return true
+		}
+
+		shorterLen := len(compactExpected)
+		if len(compactFound) < shorterLen {
+			shorterLen = len(compactFound)
+		}
+
+		maxDistance := 1
+		if shorterLen >= 10 {
+			maxDistance = 2
+		}
+		if shorterLen >= 16 {
+			maxDistance = 3
+		}
+
+		if qobuzEditDistanceWithin(compactExpected, compactFound, maxDistance) {
+			if qobuzCommonPrefixLen(compactExpected, compactFound) >= 4 ||
+				qobuzCommonSuffixLen(compactExpected, compactFound) >= 4 {
+				return true
+			}
+		}
+	}
+
+	return false
+}
+
+func qobuzCommonPrefixLen(a, b string) int {
+	max := len(a)
+	if len(b) < max {
+		max = len(b)
+	}
+	count := 0
+	for i := 0; i < max; i++ {
+		if a[i] != b[i] {
+			break
+		}
+		count++
+	}
+	return count
+}
+
+func qobuzCommonSuffixLen(a, b string) int {
+	i := len(a) - 1
+	j := len(b) - 1
+	count := 0
+	for i >= 0 && j >= 0 {
+		if a[i] != b[j] {
+			break
+		}
+		count++
+		i--
+		j--
+	}
+	return count
+}
+
+func qobuzEditDistanceWithin(a, b string, maxDistance int) bool {
+	if maxDistance < 0 {
+		return false
+	}
+
+	if a == b {
+		return true
+	}
+
+	lenA := len(a)
+	lenB := len(b)
+	diff := lenA - lenB
+	if diff < 0 {
+		diff = -diff
+	}
+	if diff > maxDistance {
+		return false
+	}
+
+	prev := make([]int, lenB+1)
+	for j := 0; j <= lenB; j++ {
+		prev[j] = j
+	}
+
+	for i := 1; i <= lenA; i++ {
+		curr := make([]int, lenB+1)
+		curr[0] = i
+		minInRow := curr[0]
+
+		for j := 1; j <= lenB; j++ {
+			cost := 0
+			if a[i-1] != b[j-1] {
+				cost = 1
+			}
+
+			insertCost := curr[j-1] + 1
+			deleteCost := prev[j] + 1
+			replaceCost := prev[j-1] + cost
+
+			best := insertCost
+			if deleteCost < best {
+				best = deleteCost
+			}
+			if replaceCost < best {
+				best = replaceCost
+			}
+
+			curr[j] = best
+			if best < minInRow {
+				minInRow = best
+			}
+		}
+
+		if minInRow > maxDistance {
+			return false
+		}
+
+		prev = curr
+	}
+
+	return prev[lenB] <= maxDistance
+}
+
 func qobuzSplitArtists(artists string) []string {
 	normalized := artists
 	normalized = strings.ReplaceAll(normalized, " feat. ", "|")
@@ -177,8 +324,10 @@ func qobuzTitlesMatch(expectedTitle, foundTitle string) bool {
 	expectedLatin := qobuzIsLatinScript(expectedTitle)
 	foundLatin := qobuzIsLatinScript(foundTitle)
 	if expectedLatin != foundLatin {
-		GoLog("[Qobuz] Titles in different scripts, assuming match: '%s' vs '%s'\n", expectedTitle, foundTitle)
-		return true
+		if qobuzCrossScriptEquivalent(expectedTitle, foundTitle) {
+			GoLog("[Qobuz] Titles in different scripts but transliteration matched: '%s' vs '%s'\n", expectedTitle, foundTitle)
+			return true
+		}
 	}
 
 	return false
@@ -702,8 +851,11 @@ func (q *QobuzDownloader) SearchTrackByMetadataWithDuration(trackName, artistNam
 	GoLog("[Qobuz] Title matches: %d out of %d results\n", len(titleMatches), len(allTracks))
 
 	tracksToCheck := titleMatches
-	if len(titleMatches) == 0 {
-		GoLog("[Qobuz] WARNING: No title matches for '%s', checking all %d results\n", trackName, len(allTracks))
+	if strings.TrimSpace(trackName) != "" && len(titleMatches) == 0 {
+		return nil, fmt.Errorf("no tracks found with matching title (expected '%s')", trackName)
+	}
+
+	if strings.TrimSpace(trackName) == "" {
 		for i := range allTracks {
 			tracksToCheck = append(tracksToCheck, &allTracks[i])
 		}
@@ -1152,10 +1304,16 @@ func downloadFromQobuz(req DownloadRequest) (QobuzDownloadResult, error) {
 	if track == nil {
 		GoLog("[Qobuz] Trying metadata search: '%s' by '%s'\n", req.TrackName, req.ArtistName)
 		track, err = downloader.SearchTrackByMetadataWithDuration(req.TrackName, req.ArtistName, expectedDurationSec)
-		if track != nil && !qobuzArtistsMatch(req.ArtistName, track.Performer.Name) {
-			GoLog("[Qobuz] Artist mismatch from metadata search: expected '%s', got '%s'. Rejecting.\n",
-				req.ArtistName, track.Performer.Name)
-			track = nil
+		if track != nil {
+			if !qobuzTitlesMatch(req.TrackName, track.Title) {
+				GoLog("[Qobuz] Title mismatch from metadata search: expected '%s', got '%s'. Rejecting.\n",
+					req.TrackName, track.Title)
+				track = nil
+			} else if !qobuzArtistsMatch(req.ArtistName, track.Performer.Name) {
+				GoLog("[Qobuz] Artist mismatch from metadata search: expected '%s', got '%s'. Rejecting.\n",
+					req.ArtistName, track.Performer.Name)
+				track = nil
+			}
 		}
 	}
 

go_backend/tidal.go

@@ -1196,13 +1196,160 @@ func artistsMatch(spotifyArtist, tidalArtist string) bool {
 	spotifyLatin := isLatinScript(spotifyArtist)
 	tidalLatin := isLatinScript(tidalArtist)
 	if spotifyLatin != tidalLatin {
-		GoLog("[Tidal] Artist names in different scripts, assuming match: '%s' vs '%s'\n", spotifyArtist, tidalArtist)
-		return true
+		if crossScriptEquivalent(spotifyArtist, tidalArtist) {
+			GoLog("[Tidal] Artist names in different scripts but transliteration matched: '%s' vs '%s'\n", spotifyArtist, tidalArtist)
+			return true
+		}
 	}
 
 	return false
 }
 
+func normalizeScriptAware(value string) string {
+	normalized := strings.ToLower(strings.TrimSpace(value))
+	normalized = CleanToASCII(JapaneseToRomaji(normalized))
+	normalized = strings.Join(strings.Fields(normalized), " ")
+	return strings.TrimSpace(normalized)
+}
+
+func crossScriptEquivalent(expected, found string) bool {
+	normExpected := normalizeScriptAware(expected)
+	normFound := normalizeScriptAware(found)
+
+	if normExpected == "" || normFound == "" {
+		return false
+	}
+
+	if normExpected == normFound {
+		return true
+	}
+
+	compactExpected := strings.ReplaceAll(normExpected, " ", "")
+	compactFound := strings.ReplaceAll(normFound, " ", "")
+	if len(compactExpected) >= 6 && len(compactFound) >= 6 {
+		if compactExpected == compactFound ||
+			strings.Contains(compactExpected, compactFound) ||
+			strings.Contains(compactFound, compactExpected) {
+			return true
+		}
+
+		shorterLen := len(compactExpected)
+		if len(compactFound) < shorterLen {
+			shorterLen = len(compactFound)
+		}
+
+		maxDistance := 1
+		if shorterLen >= 10 {
+			maxDistance = 2
+		}
+		if shorterLen >= 16 {
+			maxDistance = 3
+		}
+
+		if editDistanceWithin(compactExpected, compactFound, maxDistance) {
+			if commonPrefixLen(compactExpected, compactFound) >= 4 ||
+				commonSuffixLen(compactExpected, compactFound) >= 4 {
+				return true
+			}
+		}
+	}
+
+	return false
+}
+
+func commonPrefixLen(a, b string) int {
+	max := len(a)
+	if len(b) < max {
+		max = len(b)
+	}
+	count := 0
+	for i := 0; i < max; i++ {
+		if a[i] != b[i] {
+			break
+		}
+		count++
+	}
+	return count
+}
+
+func commonSuffixLen(a, b string) int {
+	i := len(a) - 1
+	j := len(b) - 1
+	count := 0
+	for i >= 0 && j >= 0 {
+		if a[i] != b[j] {
+			break
+		}
+		count++
+		i--
+		j--
+	}
+	return count
+}
+
+func editDistanceWithin(a, b string, maxDistance int) bool {
+	if maxDistance < 0 {
+		return false
+	}
+
+	if a == b {
+		return true
+	}
+
+	lenA := len(a)
+	lenB := len(b)
+	diff := lenA - lenB
+	if diff < 0 {
+		diff = -diff
+	}
+	if diff > maxDistance {
+		return false
+	}
+
+	prev := make([]int, lenB+1)
+	for j := 0; j <= lenB; j++ {
+		prev[j] = j
+	}
+
+	for i := 1; i <= lenA; i++ {
+		curr := make([]int, lenB+1)
+		curr[0] = i
+		minInRow := curr[0]
+
+		for j := 1; j <= lenB; j++ {
+			cost := 0
+			if a[i-1] != b[j-1] {
+				cost = 1
+			}
+
+			insertCost := curr[j-1] + 1
+			deleteCost := prev[j] + 1
+			replaceCost := prev[j-1] + cost
+
+			best := insertCost
+			if deleteCost < best {
+				best = deleteCost
+			}
+			if replaceCost < best {
+				best = replaceCost
+			}
+
+			curr[j] = best
+			if best < minInRow {
+				minInRow = best
+			}
+		}
+
+		if minInRow > maxDistance {
+			return false
+		}
+
+		prev = curr
+	}
+
+	return prev[lenB] <= maxDistance
+}
+
 func splitArtists(artists string) []string {
 	normalized := artists
 	normalized = strings.ReplaceAll(normalized, " feat. ", "|")
@@ -1292,8 +1439,10 @@ func titlesMatch(expectedTitle, foundTitle string) bool {
 	expectedLatin := isLatinScript(expectedTitle)
 	foundLatin := isLatinScript(foundTitle)
 	if expectedLatin != foundLatin {
-		GoLog("[Tidal] Titles in different scripts, assuming match: '%s' vs '%s'\n", expectedTitle, foundTitle)
-		return true
+		if crossScriptEquivalent(expectedTitle, foundTitle) {
+			GoLog("[Tidal] Titles in different scripts but transliteration matched: '%s' vs '%s'\n", expectedTitle, foundTitle)
+			return true
+		}
 	}
 
 	return false