Enable storage retention for pub

docs/configuration.md

@@ -104,7 +104,7 @@ A package rule wins over an ecosystem rule, which wins over `default`. Package r
 | `npm` | yes | `pkg:npm/lodash`, `pkg:npm/@babel/core`; the name has to match exactly, case included, so `pkg:npm/jsonstream` doesn't cover `JSONStream`. |
 | `nuget` | no | - |
 | `oci` | yes | `pkg:oci/library/nginx`, the repository name of the pull. A named upstream keeps its prefix, also when containerd reaches it through `ns`: `ghcr.io/owner/app` pulled that way is `pkg:oci/upstream/ghcr/owner/app`. Homebrew bottles use Homebrew's image name, with `@` written as `/` and `+` as `x` (`openssl@3` is `pkg:oci/homebrew/core/openssl/3`). Helm charts pulled over OCI are stored like images. Only blobs are evicted; manifests stay in the metadata cache. |
-| `pub` | no | - |
+| `pub` | yes | `pkg:pub/http`, the package name as published. |
 | `pypi` | yes | `pkg:pypi/requests`; case and single `_`/`-` don't matter. Dots and repeated separators are kept as written, so a project like `zope.interface` can be stored as `zope.interface` and `zope-interface`; add a rule for each. |
 | `rpm` | no | - |
 | `swift` | no | - |

internal/handler/pub.go

@@ -7,13 +7,22 @@ import (
 	"net/http"
 	"strings"
 	"time"
+
+	"github.com/git-pkgs/proxy/internal/retention"
 )
 
 const (
 	pubUpstream  = "https://pub.dev"
 	pubPathParts = 2 // name + version in path split by /versions/
 )
 
+// Packages are stored under pkg:pub/{name} with the name from the download
+// path, and purl keeps pub names as they are, so package rules use the
+// default key mapping.
+func init() {
+	retention.Default.Register(retention.Spec{Key: "pub", CanonicalPackage: retention.DefaultCanonical("pub")})
+}
+
 // PubHandler handles pub.dev registry protocol requests.
 type PubHandler struct {
 	proxy       *Proxy

internal/server/retention_pub_test.go

@@ -0,0 +1,234 @@
+package server
+
+import (
+	"context"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"net/url"
+	"strings"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/git-pkgs/proxy/internal/config"
+	"github.com/git-pkgs/proxy/internal/database"
+	"github.com/git-pkgs/proxy/internal/handler"
+	"github.com/git-pkgs/proxy/internal/mirror"
+	"github.com/git-pkgs/proxy/internal/retention"
+	"github.com/git-pkgs/registries/fetch"
+)
+
+const (
+	pubHTTPPath = "/packages/http/versions/1.2.0.tar.gz"
+	pubPathPath = "/packages/path/versions/1.9.0.tar.gz"
+)
+
+// pubRetentionUpstream serves package archives the way pub.dev does,
+// counting downloads.
+func pubRetentionUpstream(t *testing.T) (*httptest.Server, *atomic.Int64) {
+	t.Helper()
+	var downloads atomic.Int64
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		switch r.URL.Path {
+		case pubHTTPPath, pubPathPath:
+			downloads.Add(1)
+			_, _ = io.WriteString(w, "pub archive bytes")
+		default:
+			http.NotFound(w, r)
+		}
+	}))
+	t.Cleanup(srv.Close)
+	return srv, &downloads
+}
+
+func newPubRetentionProxy(t *testing.T, upstream *httptest.Server) (*httptest.Server, *database.DB) {
+	t.Helper()
+	db, store := setupEvictionTest(t)
+	// The upstream is a plain-HTTP loopback server; the default fetcher
+	// refuses loopback, a stock client reaches it.
+	fetcher := fetch.NewFetcher(fetch.WithHTTPClient(upstream.Client()), fetch.WithMaxRetries(0))
+	t.Cleanup(func() { _ = fetcher.Close() })
+	proxy := handler.NewProxy(db, store, fetcher, fetch.NewResolver(), discardLogger())
+	proxy.HTTPClient = upstream.Client()
+	h := handler.NewPubHandlerWithUpstream(proxy, "http://proxy.test", upstream.URL)
+	srv := httptest.NewServer(h.Routes())
+	t.Cleanup(srv.Close)
+	return srv, db
+}
+
+func pubGet(t *testing.T, url string) {
+	t.Helper()
+	resp, err := http.Get(url)
+	if err != nil {
+		t.Fatalf("GET %s: %v", url, err)
+	}
+	body, _ := io.ReadAll(resp.Body)
+	_ = resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		t.Fatalf("GET %s = %d %s, want 200", url, resp.StatusCode, body)
+	}
+}
+
+// pubArtifact returns the record of the archive named filename.
+func pubArtifact(t *testing.T, db *database.DB, filename string) database.Artifact {
+	t.Helper()
+	var art database.Artifact
+	if err := db.Get(&art, db.Rebind(`SELECT * FROM artifacts WHERE filename = ?`), filename); err != nil {
+		t.Fatalf("reading artifact %s: %v", filename, err)
+	}
+	return art
+}
+
+func pubBackdate(t *testing.T, db *database.DB, filename string, fetched, accessed time.Time) {
+	t.Helper()
+	query := db.Rebind(`UPDATE artifacts SET fetched_at = ?, last_accessed_at = ? WHERE filename = ?`)
+	if _, err := db.Exec(query, fetched, accessed, filename); err != nil {
+		t.Fatalf("backdating %s: %v", filename, err)
+	}
+}
+
+func TestPubRegistersForRetention(t *testing.T) {
+	if _, ok := retention.Default.Lookup("pub"); !ok {
+		t.Fatal("pub is not registered for retention")
+	}
+	if key, ok := retention.Default.KeyForDBEcosystem("pub"); !ok || key != "pub" {
+		t.Errorf("KeyForDBEcosystem(pub) = %q, %v, want pub, true", key, ok)
+	}
+}
+
+// An archive served from the cache stays while it is used, is evicted once
+// its rule runs out, comes back from upstream on the next download, and
+// stays after that even with the evicted copy's old access time on the
+// record.
+func TestPubRetentionEvictsAndRefetches(t *testing.T) {
+	upstream, downloads := pubRetentionUpstream(t)
+	srv, db := newPubRetentionProxy(t, upstream)
+	const archive = "http-1.2.0.tar.gz"
+
+	pubGet(t, srv.URL+pubHTTPPath)
+	pubGet(t, srv.URL+pubHTTPPath)
+	if got := downloads.Load(); got != 1 {
+		t.Fatalf("upstream downloads = %d, want 1", got)
+	}
+	if !pubArtifact(t, db, archive).LastAccessedAt.Valid {
+		t.Fatal("cache hit did not record an access")
+	}
+
+	rules := retention.Rules{Ecosystems: map[string]time.Duration{"pub": retentionDay}}
+	now := time.Now()
+	sweepRetention(context.Background(), db, discardLogger(), retention.Default, rules, now)
+	if !pubArtifact(t, db, archive).StoragePath.Valid {
+		t.Fatal("archive used today was evicted")
+	}
+
+	old := now.Add(-2 * retentionDay)
+	pubBackdate(t, db, archive, old, old)
+	sweepRetention(context.Background(), db, discardLogger(), retention.Default, rules, now)
+	if pubArtifact(t, db, archive).StoragePath.Valid {
+		t.Fatal("archive unused for two days kept under a one-day rule")
+	}
+
+	pubGet(t, srv.URL+pubHTTPPath)
+	if got := downloads.Load(); got != 2 {
+		t.Errorf("upstream downloads = %d, want 2 after the eviction", got)
+	}
+	art := pubArtifact(t, db, archive)
+	pubBackdate(t, db, archive, art.FetchedAt.Time, now.Add(-30*retentionDay))
+	sweepRetention(context.Background(), db, discardLogger(), retention.Default, rules, time.Now())
+	if !pubArtifact(t, db, archive).StoragePath.Valid {
+		t.Error("refetched archive evicted again by the next sweep")
+	}
+}
+
+// A key with a namespace parses as a pub PURL but can never match a stored
+// package, so it is refused at validation. A plain key is accepted.
+func TestPubRetentionRefusesNamespacedKeys(t *testing.T) {
+	cfg := config.Default()
+	cfg.Storage.Retention.Packages = map[string]string{"pkg:pub/http": "0"}
+	if _, err := cfg.RetentionRules(retention.Default); err != nil {
+		t.Fatalf("RetentionRules(pkg:pub/http): %v", err)
+	}
+	cfg.Storage.Retention.Packages = map[string]string{"pkg:pub/dart-lang/http": "0"}
+	_, err := cfg.RetentionRules(retention.Default)
+	if err == nil || !strings.Contains(err.Error(), "does not match how pub packages are stored") {
+		t.Errorf("RetentionRules(pkg:pub/dart-lang/http) error = %v, want the key refused", err)
+	}
+}
+
+// A package rule written as a pub PURL reaches the stored package.
+func TestPubRetentionPackageRule(t *testing.T) {
+	upstream, _ := pubRetentionUpstream(t)
+	srv, db := newPubRetentionProxy(t, upstream)
+	const httpArchive, pathArchive = "http-1.2.0.tar.gz", "path-1.9.0.tar.gz"
+	pubGet(t, srv.URL+pubHTTPPath)
+	pubGet(t, srv.URL+pubPathPath)
+	old := time.Now().Add(-30 * retentionDay)
+	pubBackdate(t, db, httpArchive, old, old)
+	pubBackdate(t, db, pathArchive, old, old)
+
+	cfg := config.Default()
+	cfg.Storage.Retention.Ecosystems = map[string]string{"pub": "1d"}
+	cfg.Storage.Retention.Packages = map[string]string{"pkg:pub/http": "0"}
+	rules, err := cfg.RetentionRules(retention.Default)
+	if err != nil {
+		t.Fatalf("RetentionRules: %v", err)
+	}
+	sweepRetention(context.Background(), db, discardLogger(), retention.Default, rules, time.Now())
+
+	if !pubArtifact(t, db, httpArchive).StoragePath.Valid {
+		t.Error("http evicted despite its never-evict rule")
+	}
+	if pubArtifact(t, db, pathArchive).StoragePath.Valid {
+		t.Error("path kept past the pub rule")
+	}
+}
+
+// pubRewriteTransport sends every request to the test upstream, so the
+// mirror command's fixed pub.dev URLs reach it.
+type pubRewriteTransport struct {
+	target *url.URL
+	base   http.RoundTripper
+}
+
+func (r pubRewriteTransport) RoundTrip(req *http.Request) (*http.Response, error) {
+	req = req.Clone(req.Context())
+	req.URL.Scheme = r.target.Scheme
+	req.URL.Host = r.target.Host
+	req.Host = r.target.Host
+	return r.base.RoundTrip(req)
+}
+
+// A package preloaded with the mirror command falls under the pub rule.
+func TestPubRetentionCoversMirroredPackages(t *testing.T) {
+	upstream, downloads := pubRetentionUpstream(t)
+	target, err := url.Parse(upstream.URL)
+	if err != nil {
+		t.Fatal(err)
+	}
+	db, store := setupEvictionTest(t)
+	client := &http.Client{Transport: pubRewriteTransport{target: target, base: upstream.Client().Transport}}
+	fetcher := fetch.NewFetcher(fetch.WithHTTPClient(client), fetch.WithMaxRetries(0))
+	t.Cleanup(func() { _ = fetcher.Close() })
+	proxy := handler.NewProxy(db, store, fetcher, fetch.NewResolver(), discardLogger())
+
+	m := mirror.New(proxy, db, store, discardLogger(), 1)
+	progress, err := m.Run(context.Background(), &mirror.PURLSource{PURLs: []string{"pkg:pub/http@1.2.0"}})
+	if err != nil || progress.Failed != 0 || downloads.Load() != 1 {
+		t.Fatalf("mirror run = %+v, %v, downloads %d; want one successful download", progress, err, downloads.Load())
+	}
+
+	const archive = "http-1.2.0.tar.gz"
+	if pubArtifact(t, db, archive).LastAccessedAt.Valid {
+		t.Fatal("mirrored archive has an access recorded before anyone downloaded it")
+	}
+	query := db.Rebind(`UPDATE artifacts SET fetched_at = ? WHERE filename = ?`)
+	if _, err := db.Exec(query, time.Now().Add(-30*retentionDay), archive); err != nil {
+		t.Fatalf("backdating %s: %v", archive, err)
+	}
+	rules := retention.Rules{Ecosystems: map[string]time.Duration{"pub": retentionDay}}
+	sweepRetention(context.Background(), db, discardLogger(), retention.Default, rules, time.Now())
+	if pubArtifact(t, db, archive).StoragePath.Valid {
+		t.Error("never-served mirrored archive kept past the pub rule")
+	}
+}