Files
NaviWatcher/internal/scanner/scanner_test.go

132 lines
3.7 KiB
Go

package scanner
import "testing"
func TestSimilarity(t *testing.T) {
tests := []struct {
name string
a string
b string
want float64
epsilon float64
}{
{
name: "exact match scores 1.0",
a: "The Wall",
b: "The Wall",
want: 1.0,
epsilon: 1e-9,
},
{
name: "case-insensitive exact match scores 1.0",
a: "The Wall",
b: "the wall",
want: 1.0,
epsilon: 1e-9,
},
{
name: "remastered variant stays above threshold",
a: "The Wall",
b: "The Wall (Remastered)",
want: 1.0, // parenthesized content is stripped during normalization
epsilon: 1e-9,
},
{
name: "year-suffixed variant stays above threshold",
a: "Abbey Road",
b: "Abbey Road (2019 Remix)",
// After normalization both collapse to "abbey road" → identical.
want: 1.0,
epsilon: 1e-9,
},
{
name: "clearly different titles score below 0.85",
a: "The Wall",
b: "Completely Different Album",
want: 0.0,
epsilon: 1e-9,
},
{
name: "substring-ish title scores moderately",
a: "Dark Side of the Moon",
b: "Dark Side of the Moon Part II",
want: 0.0, // non-empty; value asserted only as below threshold
epsilon: 1e-9,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
got := Similarity(tt.a, tt.b)
switch {
case tt.name == "clearly different titles score below 0.85" ||
tt.name == "substring-ish title scores moderately":
if got >= 0.85 {
t.Errorf("Similarity(%q, %q) = %v, want < 0.85", tt.a, tt.b, got)
}
default:
if diff := got - tt.want; diff > tt.epsilon || diff < -tt.epsilon {
t.Errorf("Similarity(%q, %q) = %v, want %v (+/- %v)", tt.a, tt.b, got, tt.want, tt.epsilon)
}
}
})
}
}
func TestSimilarity_EmptyStrings(t *testing.T) {
// Both empty → no match (0.0).
if got := Similarity("", ""); got != 0.0 {
t.Errorf("Similarity(%q, %q) = %v, want 0.0", "", "", got)
}
// One empty, one non-empty → no similarity.
if got := Similarity("The Wall", ""); got != 0.0 {
t.Errorf("Similarity(%q, %q) = %v, want 0.0", "The Wall", "", got)
}
if got := Similarity("", "The Wall"); got != 0.0 {
t.Errorf("Similarity(%q, %q) = %v, want 0.0", "", "The Wall", got)
}
// Whitespace-only inputs normalize to empty → no match.
if got := Similarity(" ", "The Wall"); got != 0.0 {
t.Errorf("Similarity(%q, %q) = %v, want 0.0", " ", "The Wall", got)
}
}
func TestIsMatch(t *testing.T) {
const threshold = 0.85
tests := []struct {
name string
a string
b string
expected bool
}{
{name: "exact match is a match", a: "The Wall", b: "The Wall", expected: true},
{name: "remastered variant is a match", a: "The Wall", b: "The Wall (Remastered)", expected: true},
{name: "year variant is a match", a: "Abbey Road", b: "Abbey Road (2019)", expected: true},
{name: "clearly different is not a match", a: "The Wall", b: "Random Noise", expected: false},
{name: "empty vs non-empty is not a match", a: "", b: "The Wall", expected: false},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if got := IsMatch(tt.a, tt.b, threshold); got != tt.expected {
t.Errorf("IsMatch(%q, %q, %v) = %v, want %v", tt.a, tt.b, threshold, got, tt.expected)
}
})
}
}
func TestIsMatch_ThresholdBoundary(t *testing.T) {
// A moderately different title should be a match at a low threshold but
// not at a high one, confirming the boundary is inclusive (>=).
a, b := "The Wall", "The Wall Live"
low := IsMatch(a, b, 0.5)
high := IsMatch(a, b, 0.99)
if !low {
t.Errorf("IsMatch(%q, %q, 0.5) = false, want true", a, b)
}
if high {
t.Errorf("IsMatch(%q, %q, 0.99) = true, want false", a, b)
}
}