fix(oci): cache tag lists and normalize manifest variants

docs/architecture.md

@@ -353,7 +353,7 @@ Eviction can be implemented as:
 - Fresh data - new versions visible immediately
 - Metadata is small, upstream fetch is fast
 - Set `cache_metadata: true` or use the mirror command to enable metadata caching for offline use via the `metadata_cache` table
-- OCI manifests are the exception: they are cached automatically so previously fetched images remain pullable when the registry or token service is unavailable
+- OCI manifests and tag lists are exceptions: they are cached automatically so previously fetched images remain pullable and tag resolution works when the registry or token service is unavailable
 
 **Why stream artifacts?**
 - Memory efficient - don't load large files into RAM

docs/configuration.md

@@ -292,7 +292,7 @@ Note: Hex cooldown requires disabling registry signature verification since the
 
 By default the proxy fetches metadata fresh from upstream on every request. Enable `cache_metadata` to store metadata responses in the database and storage backend for offline fallback. When upstream is unreachable, the proxy serves the last cached copy. ETag-based revalidation avoids re-downloading unchanged metadata.
 
-OCI manifests are always cached because cached image blobs cannot be pulled without their manifests. Digest-addressed manifests are immutable and served directly from cache. Tag-addressed manifests follow `metadata_ttl`, revalidate when stale, and fall back to the last cached response when the registry is unavailable.
+OCI manifests and tag lists are always cached because cached image blobs cannot be pulled without their manifests and offline clients may need tag resolution. Digest-addressed manifests are immutable and served directly from cache. Tag-addressed manifests and tag lists follow `metadata_ttl`, revalidate when stale, and fall back to the last cached response when the registry is unavailable.
 
 ```yaml
 cache_metadata: true

internal/handler/container.go

@@ -4,7 +4,6 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"io"
 	"net/http"
 	"regexp"
 	"strings"
@@ -182,7 +181,7 @@ func (h *ContainerHandler) handleManifest(w http.ResponseWriter, r *http.Request
 	h.serveManifest(w, r, registryURL, upstreamName, reference)
 }
 
-// handleTagsList proxies tag list requests to upstream.
+// handleTagsList caches tag list responses for offline OCI pulls.
 func (h *ContainerHandler) handleTagsList(w http.ResponseWriter, r *http.Request, path string) {
 	if r.Method != http.MethodGet {
 		http.Error(w, "method not allowed", http.StatusMethodNotAllowed)
@@ -201,27 +200,7 @@ func (h *ContainerHandler) handleTagsList(w http.ResponseWriter, r *http.Request
 		return
 	}
 
-	upstreamURL := fmt.Sprintf("%s/v2/%s/tags/list", registryURL, upstreamName)
-	if r.URL.RawQuery != "" {
-		upstreamURL += "?" + r.URL.RawQuery
-	}
-
-	req, err := http.NewRequestWithContext(r.Context(), http.MethodGet, upstreamURL, nil)
-	if err != nil {
-		h.containerError(w, http.StatusInternalServerError, "INTERNAL_ERROR", "failed to create request")
-		return
-	}
-
-	resp, err := h.proxy.HTTPClient.Do(req)
-	if err != nil {
-		h.containerError(w, http.StatusBadGateway, "INTERNAL_ERROR", "failed to fetch from upstream")
-		return
-	}
-	defer func() { _ = resp.Body.Close() }()
-
-	w.Header().Set("Content-Type", "application/json")
-	w.WriteHeader(resp.StatusCode)
-	_, _ = io.Copy(w, resp.Body)
+	h.serveTagsList(w, r, registryURL, upstreamName)
 }
 
 // proxyBlobHead handles HEAD requests for blobs.

internal/handler/container_manifest.go

@@ -8,8 +8,10 @@ import (
 	"encoding/hex"
 	"fmt"
 	"io"
+	"mime"
 	"net/http"
 	"regexp"
+	"sort"
 	"strconv"
 	"strings"
 	"time"
@@ -20,6 +22,10 @@ import (
 const (
 	containerManifestCacheEcosystem = "oci-manifest"
 	containerStaleWarning           = `110 - "Response is Stale"`
+
+	containerAcceptWildcardSpecificity = iota
+	containerAcceptTypeWildcardSpecificity
+	containerAcceptExactSpecificity
 )
 
 var manifestDigestReferencePattern = regexp.MustCompile(`^[a-z0-9]+:[a-f0-9]+$`)
@@ -35,12 +41,9 @@ type cachedContainerManifest struct {
 
 func (h *ContainerHandler) serveManifest(w http.ResponseWriter, r *http.Request, registryURL, name, reference string) {
 	accept := containerManifestAccept(r)
-	cacheKey := h.containerManifestCacheKey(registryURL, name, reference, accept)
-	cached, err := h.loadContainerManifest(r.Context(), cacheKey)
-	if err != nil {
-		h.proxy.Logger.Warn("failed to read cached container manifest", "error", err)
-		cached = nil
-	}
+	cacheAccept := normalizeContainerManifestAccept(accept)
+	cacheKey := h.containerManifestCacheKey(registryURL, name, reference, cacheAccept)
+	cached := h.loadContainerManifestForAccept(r.Context(), registryURL, name, reference, accept, cacheKey)
 
 	immutable := manifestDigestReferencePattern.MatchString(reference)
 	if cached != nil && (immutable || h.containerManifestFresh(cached)) {
@@ -68,9 +71,7 @@ func (h *ContainerHandler) serveManifest(w http.ResponseWriter, r *http.Request,
 
 	if resp.StatusCode == http.StatusNotModified && cached != nil {
 		cached.fetchedAt = time.Now()
-		if err := h.storeContainerManifest(r.Context(), cacheKey, cached); err != nil {
-			h.proxy.Logger.Warn("failed to refresh cached container manifest", "error", err)
-		}
+		h.storeContainerManifestForAccept(r.Context(), registryURL, name, reference, accept, cacheAccept, cached)
 		writeContainerManifest(w, r.Method, cached, false)
 		return
 	}
@@ -107,14 +108,9 @@ func (h *ContainerHandler) serveManifest(w http.ResponseWriter, r *http.Request,
 	if manifest.contentDigest == "" {
 		manifest.contentDigest = sha256Digest(body)
 	}
-	if err := h.storeContainerManifest(r.Context(), cacheKey, manifest); err != nil {
-		h.proxy.Logger.Warn("failed to cache container manifest", "error", err)
-	}
+	h.storeContainerManifestForAccept(r.Context(), registryURL, name, reference, accept, cacheAccept, manifest)
 	if manifest.contentDigest != reference && manifestDigestReferencePattern.MatchString(manifest.contentDigest) {
-		digestKey := h.containerManifestCacheKey(registryURL, name, manifest.contentDigest, accept)
-		if err := h.storeContainerManifest(r.Context(), digestKey, manifest); err != nil {
-			h.proxy.Logger.Warn("failed to cache container manifest by digest", "error", err)
-		}
+		h.storeContainerManifestForAccept(r.Context(), registryURL, name, manifest.contentDigest, accept, cacheAccept, manifest)
 	}
 	writeContainerManifest(w, r.Method, manifest, false)
 }
@@ -139,6 +135,52 @@ func (h *ContainerHandler) containerManifestCacheKey(registryURL, name, referenc
 	return hex.EncodeToString(sum[:])
 }
 
+func (h *ContainerHandler) loadContainerManifestForAccept(ctx context.Context, registryURL, name, reference, accept, cacheKey string) *cachedContainerManifest {
+	cached, err := h.loadContainerManifest(ctx, cacheKey)
+	if err != nil {
+		h.proxy.Logger.Warn("failed to read cached container manifest", "error", err)
+		return nil
+	}
+	if cached != nil {
+		if containerManifestCacheCompatible(accept, cached) {
+			return cached
+		}
+		return nil
+	}
+
+	legacyCacheKey := h.containerManifestCacheKey(registryURL, name, reference, accept)
+	if legacyCacheKey == cacheKey {
+		return nil
+	}
+	cached, err = h.loadContainerManifest(ctx, legacyCacheKey)
+	if err != nil {
+		h.proxy.Logger.Warn("failed to read legacy cached container manifest", "error", err)
+		return nil
+	}
+	if cached == nil || !containerManifestCacheCompatible(accept, cached) {
+		return nil
+	}
+	if err := h.storeContainerManifest(ctx, cacheKey, cached); err != nil {
+		h.proxy.Logger.Warn("failed to migrate cached container manifest", "error", err)
+	}
+	return cached
+}
+
+func (h *ContainerHandler) storeContainerManifestForAccept(ctx context.Context, registryURL, name, reference, accept, cacheAccept string, manifest *cachedContainerManifest) {
+	cacheKey := h.containerManifestCacheKey(registryURL, name, reference, cacheAccept)
+	if err := h.storeContainerManifest(ctx, cacheKey, manifest); err != nil {
+		h.proxy.Logger.Warn("failed to cache container manifest", "error", err)
+	}
+
+	legacyCacheKey := h.containerManifestCacheKey(registryURL, name, reference, accept)
+	if legacyCacheKey == cacheKey {
+		return
+	}
+	if err := h.storeContainerManifest(ctx, legacyCacheKey, manifest); err != nil {
+		h.proxy.Logger.Warn("failed to cache legacy container manifest", "error", err)
+	}
+}
+
 func (h *ContainerHandler) loadContainerManifest(ctx context.Context, cacheKey string) (*cachedContainerManifest, error) {
 	if h.proxy.DB == nil || h.proxy.Storage == nil {
 		return nil, nil
@@ -233,6 +275,104 @@ func containerManifestAccept(r *http.Request) string {
 	}, ", ")
 }
 
+func normalizeContainerManifestAccept(accept string) string {
+	mediaTypes := make(map[string]struct{})
+	for _, value := range strings.Split(accept, ",") {
+		value = strings.TrimSpace(value)
+		if value == "" {
+			continue
+		}
+		mediaType, params, err := mime.ParseMediaType(value)
+		if err != nil {
+			mediaTypes[strings.ToLower(value)] = struct{}{}
+			continue
+		}
+		paramKeys := make([]string, 0, len(params))
+		for key := range params {
+			paramKeys = append(paramKeys, key)
+		}
+		sort.Strings(paramKeys)
+		canonical := strings.ToLower(mediaType)
+		for _, key := range paramKeys {
+			value := params[key]
+			if strings.EqualFold(key, "q") {
+				if quality, err := strconv.ParseFloat(value, 64); err == nil {
+					if quality == 1 {
+						continue
+					}
+					value = strconv.FormatFloat(quality, 'g', -1, 64)
+				}
+			}
+			canonical += ";" + strings.ToLower(key) + "=" + value
+		}
+		mediaTypes[canonical] = struct{}{}
+	}
+	canonicalMediaTypes := make([]string, 0, len(mediaTypes))
+	for mediaType := range mediaTypes {
+		canonicalMediaTypes = append(canonicalMediaTypes, mediaType)
+	}
+	sort.Strings(canonicalMediaTypes)
+	return strings.Join(canonicalMediaTypes, ",")
+}
+
+func containerManifestAccepts(accept, contentType string) bool {
+	contentType, _, err := mime.ParseMediaType(contentType)
+	if err != nil {
+		return false
+	}
+	contentType = strings.ToLower(contentType)
+	contentMajor, contentMinor, found := strings.Cut(contentType, "/")
+	if !found {
+		return false
+	}
+
+	bestSpecificity := -1
+	bestQuality := 0.0
+	for _, value := range strings.Split(accept, ",") {
+		mediaType, params, err := mime.ParseMediaType(strings.TrimSpace(value))
+		if err != nil {
+			continue
+		}
+		mediaType = strings.ToLower(mediaType)
+		major, minor, found := strings.Cut(mediaType, "/")
+		if found && (major == "*" || major == contentMajor) && (minor == "*" || minor == contentMinor) {
+			specificity := containerAcceptSpecificity(major, minor)
+			if specificity > bestSpecificity {
+				bestSpecificity = specificity
+				bestQuality = containerAcceptQuality(params)
+			}
+		}
+	}
+	return bestQuality > 0
+}
+
+func containerManifestCacheCompatible(accept string, manifest *cachedContainerManifest) bool {
+	return manifest.contentType == "" || containerManifestAccepts(accept, manifest.contentType)
+}
+
+func containerAcceptSpecificity(major, minor string) int {
+	switch {
+	case major == "*" && minor == "*":
+		return containerAcceptWildcardSpecificity
+	case major == "*" || minor == "*":
+		return containerAcceptTypeWildcardSpecificity
+	default:
+		return containerAcceptExactSpecificity
+	}
+}
+
+func containerAcceptQuality(params map[string]string) float64 {
+	value, ok := params["q"]
+	if !ok {
+		return 1
+	}
+	quality, err := strconv.ParseFloat(value, 64)
+	if err != nil || quality < 0 || quality > 1 {
+		return 0
+	}
+	return quality
+}
+
 func copyContainerManifestHeaders(destination, source http.Header) {
 	for _, header := range []string{"Content-Type", "Content-Length", "Docker-Content-Digest", "ETag", "WWW-Authenticate"} {
 		if value := source.Get(header); value != "" {