Add ansible Galaxy remote type

- add ansible package type proxying the Galaxy v3 API
- rewrite href/links/download_url back to the remote so tarballs flow through it
- forward the query string for providers that opt in, keeping Galaxy pagination working
- add a UI usage snippet and an ansible e2e caching case

internal/api/v1/proxy.go

@@ -92,6 +92,10 @@ func (h *ProxyHandler) handleProxy(w http.ResponseWriter, r *http.Request) {
 		}
 	}
 
+	if _, ok := prov.(provider.QueryForwarder); ok && r.URL.RawQuery != "" {
+		path += "?" + r.URL.RawQuery
+	}
+
 	result, err := h.engine.Fetch(r.Context(), *remote, path, prov, r.Header)
 	if err != nil {
 		var proxyErr *proxy.ProxyError

internal/provider/ansible/ansible.go

@@ -0,0 +1,65 @@
+// Package ansible proxies an Ansible Galaxy server (v3 collections API):
+// mutable JSON metadata whose links are rewritten back to the remote, plus
+// immutable collection tarballs.
+package ansible
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"strings"
+
+	"git.unkin.net/unkin/artifactapi/internal/auth"
+	"git.unkin.net/unkin/artifactapi/internal/provider"
+	"git.unkin.net/unkin/artifactapi/pkg/models"
+)
+
+func init() {
+	provider.Register(&Provider{})
+}
+
+type Provider struct{}
+
+func (p *Provider) Type() models.PackageType { return models.PackageAnsible }
+
+func (p *Provider) Classify(path string) provider.Mutability {
+	if isArtifact(path) {
+		return provider.Immutable
+	}
+	return provider.Mutable
+}
+
+func (p *Provider) ContentType(path string) string {
+	if isArtifact(path) {
+		return "application/gzip"
+	}
+	return "application/json"
+}
+
+func (p *Provider) UpstreamURL(remote models.Remote, path string) string {
+	return strings.TrimRight(remote.BaseURL, "/") + "/" + strings.TrimLeft(path, "/")
+}
+
+// ForwardQuery keeps limit/offset pagination in the proxied path.
+func (p *Provider) ForwardQuery() {}
+
+// RewriteResponse points root-relative hrefs/links and absolute download_url
+// values at the remote. Links stay root-relative, which ansible-galaxy joins
+// onto the server URL, so the cached body is independent of the request host.
+func (p *Provider) RewriteResponse(body []byte, remote models.Remote, _ string) ([]byte, error) {
+	if !json.Valid(body) {
+		return nil, nil
+	}
+	prefix := `"/api/v1/remote/` + remote.Name + `/`
+	content := strings.ReplaceAll(string(body), `"/api/`, prefix+`api/`)
+	content = strings.ReplaceAll(content, `"`+strings.TrimRight(remote.BaseURL, "/")+`/`, prefix)
+	return []byte(content), nil
+}
+
+func (p *Provider) AuthHeaders(_ context.Context, remote models.Remote) (http.Header, error) {
+	return auth.BasicHeaders(remote), nil
+}
+
+func isArtifact(path string) bool {
+	return strings.HasSuffix(path, ".tar.gz")
+}

internal/provider/ansible/ansible_test.go

@@ -0,0 +1,85 @@
+package ansible
+
+import (
+	"encoding/json"
+	"testing"
+
+	"git.unkin.net/unkin/artifactapi/internal/provider"
+	"git.unkin.net/unkin/artifactapi/pkg/models"
+)
+
+func TestClassify(t *testing.T) {
+	p := &Provider{}
+	tests := []struct {
+		path string
+		want provider.Mutability
+	}{
+		{"api/", provider.Mutable},
+		{"api/v3/collections/community/general/versions/?limit=100", provider.Mutable},
+		{"api/v3/collections/ansible/posix/versions/2.0.0/", provider.Mutable},
+		{"api/v3/plugin/ansible/content/published/collections/artifacts/ansible-posix-2.0.0.tar.gz", provider.Immutable},
+	}
+	for _, tt := range tests {
+		if got := p.Classify(tt.path); got != tt.want {
+			t.Errorf("Classify(%q) = %v, want %v", tt.path, got, tt.want)
+		}
+	}
+}
+
+func TestUpstreamURLKeepsQuery(t *testing.T) {
+	p := &Provider{}
+	got := p.UpstreamURL(models.Remote{BaseURL: "https://galaxy.ansible.com/"}, "api/v3/collections/a/b/versions/?limit=100&offset=100")
+	want := "https://galaxy.ansible.com/api/v3/collections/a/b/versions/?limit=100&offset=100"
+	if got != want {
+		t.Errorf("UpstreamURL = %q, want %q", got, want)
+	}
+}
+
+func TestRewriteResponse(t *testing.T) {
+	p := &Provider{}
+	remote := models.Remote{Name: "galaxy", BaseURL: "https://galaxy.ansible.com/"}
+	body := `{"links":{"next":"/api/v3/plugin/ansible/content/published/collections/index/a/b/versions/?limit=10&offset=10","previous":null},` +
+		`"href":"/api/content/published/v3/plugin/ansible/content/published/collections/index/a/b/versions/1.0.0/",` +
+		`"download_url":"https://galaxy.ansible.com/api/v3/plugin/ansible/content/published/collections/artifacts/a-b-1.0.0.tar.gz",` +
+		`"metadata":{"repository":"https://github.com/a/b"}}`
+
+	out, err := p.RewriteResponse([]byte(body), remote, "")
+	if err != nil {
+		t.Fatal(err)
+	}
+	var got struct {
+		Links       map[string]any `json:"links"`
+		Href        string         `json:"href"`
+		DownloadURL string         `json:"download_url"`
+		Metadata    map[string]any `json:"metadata"`
+	}
+	if err := json.Unmarshal(out, &got); err != nil {
+		t.Fatalf("rewritten body is not JSON: %v", err)
+	}
+	pre := "/api/v1/remote/galaxy/"
+	if want := pre + "api/v3/plugin/ansible/content/published/collections/index/a/b/versions/?limit=10&offset=10"; got.Links["next"] != want {
+		t.Errorf("next = %v, want %q", got.Links["next"], want)
+	}
+	if want := pre + "api/content/published/v3/plugin/ansible/content/published/collections/index/a/b/versions/1.0.0/"; got.Href != want {
+		t.Errorf("href = %q, want %q", got.Href, want)
+	}
+	if want := pre + "api/v3/plugin/ansible/content/published/collections/artifacts/a-b-1.0.0.tar.gz"; got.DownloadURL != want {
+		t.Errorf("download_url = %q, want %q", got.DownloadURL, want)
+	}
+	if got.Metadata["repository"] != "https://github.com/a/b" {
+		t.Errorf("unrelated URL rewritten: %v", got.Metadata["repository"])
+	}
+}
+
+func TestRewriteResponseSkipsNonJSON(t *testing.T) {
+	out, err := (&Provider{}).RewriteResponse([]byte("<html>"), models.Remote{Name: "g"}, "")
+	if err != nil || out != nil {
+		t.Errorf("got %q, %v; want nil, nil", out, err)
+	}
+}
+
+func TestForwardsQuery(t *testing.T) {
+	if _, ok := any(&Provider{}).(provider.QueryForwarder); !ok {
+		t.Error("ansible provider must forward the query string for pagination")
+	}
+}

internal/provider/provider.go

@@ -65,6 +65,13 @@ type MetadataStore interface {
 	InsertDebMetadata(ctx context.Context, meta *DebMetadata) error
 }
 
+// QueryForwarder marks a remote provider whose upstream answers depend on the
+// query string (e.g. paginated JSON APIs); the query is kept as part of the
+// proxied and cached path.
+type QueryForwarder interface {
+	ForwardQuery()
+}
+
 // RemoteServer lets a remote provider fully answer a request itself instead of
 // going through the byte-proxy engine. It is the remote-side analog of
 // LocalIndexer: a metadata-only remote (e.g. github_rpm) uses it to synthesize

internal/server/server.go

@@ -22,6 +22,7 @@ import (
 	"git.unkin.net/unkin/artifactapi/internal/gc"
 	"git.unkin.net/unkin/artifactapi/internal/githubauth"
 	"git.unkin.net/unkin/artifactapi/internal/provider/alpine"
+	_ "git.unkin.net/unkin/artifactapi/internal/provider/ansible"
 	_ "git.unkin.net/unkin/artifactapi/internal/provider/cargo"
 	"git.unkin.net/unkin/artifactapi/internal/provider/deb"
 	_ "git.unkin.net/unkin/artifactapi/internal/provider/docker"