Feature 10: Recipe grounding and photos (#78)

Co-authored-by: dev_agent <dev_agent@cubecraftlabs.com>

internal/ai/harness/dedupe.go

@@ -53,7 +53,14 @@ func Normalize(name string) string {
 // similarity of their normalized forms is at least SimilarThreshold. A name
 // with no words left after normalizing is similar to nothing.
 func Similar(a, b string) bool {
-	return similarity(tokenSet(a), tokenSet(b)) >= SimilarThreshold
+	return Similarity(a, b) >= SimilarThreshold
+}
+
+// Similarity is the token-set (Jaccard) similarity of a's and b's normalized
+// forms, from 0 (no shared word) to 1 (the same words). A name with no words
+// left after normalizing scores 0 against everything.
+func Similarity(a, b string) float64 {
+	return similarity(tokenSet(a), tokenSet(b))
 }
 
 // tokenSet returns the distinct words of name's normalized form.

internal/ai/harness/dedupe_test.go

@@ -114,6 +114,22 @@ func TestSimilar(t *testing.T) {
 	}
 }
 
+func TestSimilarity(t *testing.T) {
+	for _, tc := range []struct {
+		a, b string
+		want float64
+	}{
+		{"Easy Chicken Curry", "chicken curry", 1},
+		{"Thai Green Curry", "Chicken curry", 0.25},
+		{"Beef Stew", "Chicken curry", 0},
+		{"The Recipe", "chicken curry", 0},
+	} {
+		if got := Similarity(tc.a, tc.b); got != tc.want {
+			t.Errorf("Similarity(%q, %q) = %v, want %v", tc.a, tc.b, got, tc.want)
+		}
+	}
+}
+
 func TestDeduper(t *testing.T) {
 	d := NewDeduper(DedupeSources{
 		Regulars: []string{"Beef Tacos", "Spaghetti Bolognese"},