From cff6c1baff5e0bcea49a9369883c75ea255ede37 Mon Sep 17 00:00:00 2001
From: Samuel Enocsson <samuel.enocsson@gmail.com>
Date: Fri, 22 May 2026 11:41:33 +0200
Subject: [PATCH] feat(discovery): tag-agnostic image name normalisation

---
 internal/discovery/matching.go      | 33 +++++++++++++++++++
 internal/discovery/matching_test.go | 51 +++++++++++++++++++++++++++++
 2 files changed, 84 insertions(+)
 create mode 100644 internal/discovery/matching.go
 create mode 100644 internal/discovery/matching_test.go

diff --git a/internal/discovery/matching.go b/internal/discovery/matching.go
new file mode 100644
index 0000000..c872009
--- /dev/null
+++ b/internal/discovery/matching.go
@@ -0,0 +1,33 @@
+package discovery
+
+import "strings"
+
+// NormaliseImage strips the tag and digest from an image reference,
+// returning the bare repository name.
+//
+// Tricky case: "localhost:5000/foo:v1" — the first colon is a port,
+// not a tag. We disambiguate by splitting on "/" first and only
+// treating colons in the last segment as tag separators.
+func NormaliseImage(ref string) string {
+	if at := strings.Index(ref, "@"); at >= 0 {
+		ref = ref[:at]
+	}
+	slash := strings.LastIndex(ref, "/")
+	if slash < 0 {
+		if colon := strings.Index(ref, ":"); colon >= 0 {
+			return ref[:colon]
+		}
+		return ref
+	}
+	prefix, last := ref[:slash], ref[slash+1:]
+	if colon := strings.Index(last, ":"); colon >= 0 {
+		last = last[:colon]
+	}
+	return prefix + "/" + last
+}
+
+// ImagesMatch reports whether two image references resolve to the same
+// repository, ignoring tag and digest. Case-sensitive per spec section 5.2.
+func ImagesMatch(a, b string) bool {
+	return NormaliseImage(a) == NormaliseImage(b)
+}
diff --git a/internal/discovery/matching_test.go b/internal/discovery/matching_test.go
new file mode 100644
index 0000000..5f20fb1
--- /dev/null
+++ b/internal/discovery/matching_test.go
@@ -0,0 +1,51 @@
+package discovery_test
+
+import (
+	"testing"
+
+	"github.com/shcizo/package-updater/internal/discovery"
+	"github.com/stretchr/testify/require"
+)
+
+func TestNormaliseImage(t *testing.T) {
+	cases := []struct {
+		in   string
+		want string
+	}{
+		{"registry.example.com/myapp", "registry.example.com/myapp"},
+		{"registry.example.com/myapp:v1.2.3", "registry.example.com/myapp"},
+		{"registry.example.com/myapp:latest", "registry.example.com/myapp"},
+		{"registry.example.com/myapp@sha256:abc123", "registry.example.com/myapp"},
+		{"registry.example.com/myapp:v1.2.3@sha256:abc123", "registry.example.com/myapp"},
+		{"nginx", "nginx"},
+		{"nginx:1.25-alpine", "nginx"},
+		{"library/nginx:latest", "library/nginx"},
+		{"gcr.io/proj/svc:tag", "gcr.io/proj/svc"},
+		{"localhost:5000/myimg:v1", "localhost:5000/myimg"},
+	}
+	for _, c := range cases {
+		t.Run(c.in, func(t *testing.T) {
+			require.Equal(t, c.want, discovery.NormaliseImage(c.in))
+		})
+	}
+}
+
+func TestImagesMatch(t *testing.T) {
+	require.True(t, discovery.ImagesMatch(
+		"registry.example.com/myapp",
+		"registry.example.com/myapp:v1.2.3",
+	))
+	require.True(t, discovery.ImagesMatch(
+		"registry.example.com/myapp:v1.0.0",
+		"registry.example.com/myapp:v9.9.9",
+	))
+	require.False(t, discovery.ImagesMatch(
+		"registry.example.com/myapp",
+		"registry.example.com/otherapp",
+	))
+	// Case-sensitive per spec section 5.2
+	require.False(t, discovery.ImagesMatch(
+		"registry.example.com/MyApp",
+		"registry.example.com/myapp",
+	))
+}