feat: cache changed files in Gitea provider

Implemented a caching mechanism for changed files in the Gitea provider
to prevent redundant API requests during the reconciliation process.

This mirror the implementation on PR #2317

Jira: https://issues.redhat.com/browse/SRVKP-10944
Signed-off-by: Chmouel Boudjnah <chmouel@redhat.com>

Author: chmouel

pkg/provider/gitea/gitea.go

@@ -61,12 +61,13 @@ type Provider struct {
 	Token            *string
 	giteaInstanceURL string
 	// only exposed for e2e tests
-	Password     string
-	repo         *v1alpha1.Repository
-	eventEmitter *events.EventEmitter
-	run          *params.Run
-	triggerEvent string
-	pacUserID    int64 // user login used by PAC
+	Password           string
+	repo               *v1alpha1.Repository
+	eventEmitter       *events.EventEmitter
+	run                *params.Run
+	triggerEvent       string
+	pacUserID          int64 // user login used by PAC
+	cachedChangedFiles *changedfiles.ChangedFiles
 }
 
 func (v *Provider) Client() *forgejo.Client {
@@ -535,7 +536,19 @@ type PushPayload struct {
 	Commits []forgejo.PayloadCommit `json:"commits,omitempty"`
 }
 
-func (v *Provider) GetFiles(_ context.Context, runevent *info.Event) (changedfiles.ChangedFiles, error) {
+// GetFiles gets and caches the list of files changed by a given event.
+func (v *Provider) GetFiles(ctx context.Context, runevent *info.Event) (changedfiles.ChangedFiles, error) {
+	if v.cachedChangedFiles == nil {
+		changes, err := v.fetchChangedFiles(ctx, runevent)
+		if err != nil {
+			return changedfiles.ChangedFiles{}, err
+		}
+		v.cachedChangedFiles = &changes
+	}
+	return *v.cachedChangedFiles, nil
+}
+
+func (v *Provider) fetchChangedFiles(_ context.Context, runevent *info.Event) (changedfiles.ChangedFiles, error) {
 	changedFiles := changedfiles.ChangedFiles{}
 
 	//nolint:exhaustive // we don't need to handle all cases

pkg/provider/gitea/gitea_test.go

@@ -25,10 +25,13 @@ import (
 	"github.com/openshift-pipelines/pipelines-as-code/pkg/provider"
 	tgitea "github.com/openshift-pipelines/pipelines-as-code/pkg/provider/gitea/test"
 	"github.com/openshift-pipelines/pipelines-as-code/pkg/provider/status"
+	metricsutils "github.com/openshift-pipelines/pipelines-as-code/pkg/test/metricstest"
 	"go.uber.org/zap"
 	zapobserver "go.uber.org/zap/zaptest/observer"
 	"gotest.tools/v3/assert"
 	"gotest.tools/v3/golden"
+	"knative.dev/pkg/metrics/metricstest"
+	_ "knative.dev/pkg/metrics/testing"
 	rtesting "knative.dev/pkg/reconciler/testing"
 )
 
@@ -277,16 +280,17 @@ func TestProvider_Validate(t *testing.T) {
 	}
 }
 
-func TestProvider_GetFiles(t *testing.T) {
+func TestProviderGetFiles(t *testing.T) {
 	type args struct {
 		runevent *info.Event
 	}
 	tests := []struct {
-		name         string
-		args         args
-		changedFiles string
-		want         changedfiles.ChangedFiles
-		wantErr      bool
+		name                string
+		args                args
+		changedFiles        string
+		want                changedfiles.ChangedFiles
+		wantErr             bool
+		wantAPIRequestCount int64
 	}{
 		{
 			name: "pull_request",
@@ -312,7 +316,8 @@ func TestProvider_GetFiles(t *testing.T) {
 				Modified: []string{"modified.txt"},
 				Renamed:  []string{"renamed.txt"},
 			},
-			changedFiles: `[{"filename":"added.txt","status":"added"},{"filename":"deleted.txt","status":"deleted"},{"filename":"modified.txt","status":"changed"},{"filename":"renamed.txt","status":"renamed"}]`,
+			changedFiles:        `[{"filename":"added.txt","status":"added"},{"filename":"deleted.txt","status":"deleted"},{"filename":"modified.txt","status":"changed"},{"filename":"renamed.txt","status":"renamed"}]`,
+			wantAPIRequestCount: 1,
 		},
 		{
 			name: "push",
@@ -342,12 +347,12 @@ func TestProvider_GetFiles(t *testing.T) {
 				},
 				Deleted:  []string{"deleted.txt"},
 				Modified: []string{"modified.txt"},
-				// Renamed:  []string{"renamed.txt"},
 			},
 		},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
+			metricsutils.ResetMetrics()
 			fakeclient, mux, teardown := tgitea.Setup(t)
 			defer teardown()
 
@@ -360,12 +365,18 @@ func TestProvider_GetFiles(t *testing.T) {
 			repo := &v1alpha1.Repository{Spec: v1alpha1.RepositorySpec{
 				Settings: &v1alpha1.Settings{},
 			}}
+			giteaInstanceURL := "https://gitea.example.com"
 			gprovider := Provider{
-				giteaClient: fakeclient,
-				repo:        repo,
-				Logger:      logger,
+				giteaClient:      fakeclient,
+				repo:             repo,
+				Logger:           logger,
+				giteaInstanceURL: giteaInstanceURL,
+				triggerEvent:     string(tt.args.runevent.TriggerTarget),
 			}
 
+			metricsTags := map[string]string{"provider": giteaInstanceURL, "event-type": string(tt.args.runevent.TriggerTarget)}
+			metricstest.CheckStatsNotReported(t, "pipelines_as_code_git_provider_api_request_count")
+
 			got, err := gprovider.GetFiles(ctx, tt.args.runevent)
 
 			if (err != nil) != tt.wantErr {
@@ -401,6 +412,24 @@ func TestProvider_GetFiles(t *testing.T) {
 			if !reflect.DeepEqual(got.Renamed, tt.want.Renamed) {
 				t.Errorf("Provider.GetFiles() Renamed = %v, want %v", got.Renamed, tt.want.Renamed)
 			}
+
+			// Verify metrics from first call
+			if tt.wantAPIRequestCount > 0 {
+				metricstest.CheckCountData(t, "pipelines_as_code_git_provider_api_request_count", metricsTags, tt.wantAPIRequestCount)
+			} else {
+				metricstest.CheckStatsNotReported(t, "pipelines_as_code_git_provider_api_request_count")
+			}
+
+			// Verify caching: second call should return cached result without additional API calls
+			got2, err2 := gprovider.GetFiles(ctx, tt.args.runevent)
+			assert.NilError(t, err2)
+			assert.DeepEqual(t, got, got2)
+
+			if tt.wantAPIRequestCount > 0 {
+				metricstest.CheckCountData(t, "pipelines_as_code_git_provider_api_request_count", metricsTags, tt.wantAPIRequestCount)
+			} else {
+				metricstest.CheckStatsNotReported(t, "pipelines_as_code_git_provider_api_request_count")
+			}
 		})
 	}
 }