Including the Scorecard API

- Fixes #1892 - Updated tests - Added a README for certifier/scorecard Signed-off-by: neilnaveen <42328488+neilnaveen@users.noreply.github.com>
guacsec · Jun 17, 2024 · 2db7b2b · 2db7b2b
1 parent ff4c8af
commit 2db7b2b
Show file tree

Hide file tree

Showing 16 changed files with 183 additions and 99 deletions.
diff --git a/cmd/guaccollect/cmd/osv.go b/cmd/guaccollect/cmd/osv.go
@@ -129,10 +129,14 @@ func validateOSVFlags(
 	interval string,
 	poll bool,
 	pubToQueue bool,
+<<<<<<< HEAD
 	daysSince int,
 	certifierLatencyStr string,
 	batchSize int) (osvOptions, error) {
 
+=======
+) (osvOptions, error) {
+>>>>>>> 080dbd51 (Including the Scorecard API)
 	var opts osvOptions
 
 	opts.graphqlEndpoint = graphqlEndpoint
@@ -178,8 +182,8 @@ func getPackageQuery(client graphql.Client, daysSinceLastScan int, batchSize int
 }
 
 func initializeNATsandCertifier(ctx context.Context, blobAddr, pubsubAddr string,
-	poll, publishToQueue bool, interval time.Duration, query certifier.QueryComponents) {
-
+	poll, publishToQueue bool, interval time.Duration, query certifier.QueryComponents,
+) {
 	logger := logging.FromContext(ctx)
 
 	blobStore, err := blob.NewBlobStore(ctx, blobAddr)
@@ -236,7 +240,11 @@ func initializeNATsandCertifier(ctx context.Context, blobAddr, pubsubAddr string
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
+<<<<<<< HEAD
 		if err := certify.Certify(ctx, query, emit, errHandler, poll, interval); err != nil {
+=======
+		if err := certify.Certify(ctx, query, emit, errHandler, poll, time.Minute*time.Duration(interval), false); err != nil {
+>>>>>>> 080dbd51 (Including the Scorecard API)
 			logger.Fatal(err)
 		}
 		done <- true

diff --git a/cmd/guacone/cmd/osv.go b/cmd/guacone/cmd/osv.go
@@ -187,7 +187,7 @@ var osvCmd = &cobra.Command{
 		wg.Add(1)
 		go func() {
 			defer wg.Done()
-			if err := certify.Certify(ctx, packageQuery, emit, errHandler, opts.poll, opts.interval); err != nil {
+			if err := certify.Certify(ctx, packageQuery, emit, errHandler, opts.poll, opts.interval, false); err != nil {
 				logger.Errorf("Unhandled error in the certifier: %s", err)
 			}
 			done <- true
@@ -258,8 +258,10 @@ func validateOSVFlags(
 }
 
 func init() {
-	set, err := cli.BuildFlags([]string{"certifier-latency",
-		"certifier-batch-size"})
+	set, err := cli.BuildFlags([]string{
+		"certifier-latency",
+		"certifier-batch-size",
+	})
 	if err != nil {
 		fmt.Fprintf(os.Stderr, "failed to setup flag: %v", err)
 		os.Exit(1)

diff --git a/cmd/guacone/cmd/scorecard.go b/cmd/guacone/cmd/scorecard.go
@@ -26,16 +26,14 @@ import (
 	"time"
 
 	"github.com/Khan/genqlient/graphql"
+	"github.com/guacsec/guac/pkg/certifier"
+	"github.com/guacsec/guac/pkg/certifier/certify"
 	sc "github.com/guacsec/guac/pkg/certifier/components/source"
+	"github.com/guacsec/guac/pkg/certifier/scorecard"
 	"github.com/guacsec/guac/pkg/cli"
 	csub_client "github.com/guacsec/guac/pkg/collectsub/client"
-	"github.com/guacsec/guac/pkg/ingestor"
-
-	"github.com/guacsec/guac/pkg/certifier"
-	"github.com/guacsec/guac/pkg/certifier/scorecard"
-
-	"github.com/guacsec/guac/pkg/certifier/certify"
 	"github.com/guacsec/guac/pkg/handler/processor"
+	"github.com/guacsec/guac/pkg/ingestor"
 	"github.com/guacsec/guac/pkg/logging"
 	"github.com/spf13/cobra"
 	"github.com/spf13/viper"
@@ -51,7 +49,8 @@ type scorecardOptions struct {
 	// sets artificial latency on the certifier (default to nil)
 	addedLatency *time.Duration
 	// sets the batch size for pagination query for the certifier
-	batchSize int
+	batchSize       int
+	useScorecardAPI bool
 }
 
 var scorecardCmd = &cobra.Command{
@@ -69,6 +68,7 @@ var scorecardCmd = &cobra.Command{
 			viper.GetBool("add-vuln-on-ingest"),
 			viper.GetString("certifier-latency"),
 			viper.GetInt("certifier-batch-size"),
+			viper.GetBool("use-scorecard-api"),
 		)
 		if err != nil {
 			fmt.Printf("unable to validate flags: %v\n", err)
@@ -102,7 +102,6 @@ var scorecardCmd = &cobra.Command{
 
 		// running and getting the scorecard checks
 		scorecardCertifier, err := scorecard.NewScorecardCertifier(scorecardRunner)
-
 		if err != nil {
 			fmt.Printf("unable to create scorecard certifier: %v\n", err)
 			_ = cmd.Help()
@@ -112,7 +111,6 @@ var scorecardCmd = &cobra.Command{
 		// scorecard certifier is the certifier that gets the scorecard data graphQL
 		// setting "daysSinceLastScan" to 0 does not check the timestamp on the scorecard that exist
 		query, err := sc.NewCertifier(gqlclient, 0, opts.batchSize, opts.addedLatency)
-
 		if err != nil {
 			fmt.Printf("unable to create scorecard certifier: %v\n", err)
 			_ = cmd.Help()
@@ -132,7 +130,6 @@ var scorecardCmd = &cobra.Command{
 		emit := func(d *processor.Document) error {
 			totalNum += 1
 			err := ingestor.Ingest(ctx, d, opts.graphqlEndpoint, transport, csubClient, opts.queryVulnOnIngestion)
-
 			if err != nil {
 				return fmt.Errorf("unable to ingest document: %v", err)
 			}
@@ -156,7 +153,7 @@ var scorecardCmd = &cobra.Command{
 		wg.Add(1)
 		go func() {
 			defer wg.Done()
-			if err := certify.Certify(ctx, query, emit, errHandler, opts.poll, opts.interval); err != nil {
+			if err := certify.Certify(ctx, query, emit, errHandler, opts.poll, opts.interval, opts.useScorecardAPI); err != nil {
 				logger.Errorf("Unhandled error in the certifier: %s", err)
 			}
 			done <- true
@@ -191,6 +188,7 @@ func validateScorecardFlags(
 	queryVulnIngestion bool,
 	certifierLatencyStr string,
 	batchSize int,
+	useScorecardAPI bool,
 ) (scorecardOptions, error) {
 	var opts scorecardOptions
 	opts.graphqlEndpoint = graphqlEndpoint
@@ -215,6 +213,7 @@ func validateScorecardFlags(
 	opts.csubClientOptions = csubOpts
 
 	opts.poll = poll
+	opts.useScorecardAPI = useScorecardAPI
 	i, err := time.ParseDuration(interval)
 	if err != nil {
 		return opts, err
@@ -225,8 +224,10 @@ func validateScorecardFlags(
 }
 
 func init() {
-	set, err := cli.BuildFlags([]string{"certifier-latency",
-		"certifier-batch-size"})
+	set, err := cli.BuildFlags([]string{
+		"certifier-latency",
+		"certifier-batch-size",
+	})
 	if err != nil {
 		fmt.Fprintf(os.Stderr, "failed to setup flag: %v", err)
 		os.Exit(1)
@@ -237,4 +238,6 @@ func init() {
 		os.Exit(1)
 	}
 	certifierCmd.AddCommand(scorecardCmd)
+	scorecardCmd.Flags().Bool("use-scorecard-api", false, "use the scorecard API")
+	viper.BindPFlag("use-scorecard-api", scorecardCmd.Flags().Lookup("use-scorecard-api"))
 }
diff --git a/internal/testing/cmd/pubsub_test/cmd/osv.go b/internal/testing/cmd/pubsub_test/cmd/osv.go
@@ -56,7 +56,6 @@ var osvCmd = &cobra.Command{
 			viper.GetBool("poll"),
 			viper.GetInt("interval"),
 		)
-
 		if err != nil {
 			fmt.Printf("unable to validate flags: %v\n", err)
 			_ = cmd.Help()
@@ -158,7 +157,7 @@ func initializeNATsandCertifier(ctx context.Context, opts options) {
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
-		if err := certify.Certify(ctx, packageQueryFunc(), emit, errHandler, opts.poll, time.Minute*time.Duration(opts.interval)); err != nil {
+		if err := certify.Certify(ctx, packageQueryFunc(), emit, errHandler, opts.poll, time.Minute*time.Duration(opts.interval), false); err != nil {
 			logger.Fatal(err)
 		}
 		done <- true

diff --git a/internal/testing/mocks/scorecard.go b/internal/testing/mocks/scorecard.go
diff --git a/pkg/certifier/certifier.go b/pkg/certifier/certifier.go
@@ -26,7 +26,7 @@ type Certifier interface {
 	// push to the docChannel to be ingested.
 	// Note: there is an implicit contract with "QueryComponents" where the compChan type must be the same as
 	// the one used by "components"
-	CertifyComponent(ctx context.Context, components interface{}, docChannel chan<- *processor.Document) error
+	CertifyComponent(ctx context.Context, components interface{}, docChannel chan<- *processor.Document, useScorecardAPI bool) error
 }
 
 type QueryComponents interface {

diff --git a/pkg/certifier/certify/certify.go b/pkg/certifier/certify/certify.go
@@ -52,8 +52,7 @@ func RegisterCertifier(c func() certifier.Certifier, certifierType certifier.Cer
 
 // Certify queries the graph DB to get the components to scan. Utilizing the registered certifiers,
 // it generates new nodes and attestations.
-func Certify(ctx context.Context, query certifier.QueryComponents, emitter certifier.Emitter, handleErr certifier.ErrHandler, poll bool, interval time.Duration) error {
-
+func Certify(ctx context.Context, query certifier.QueryComponents, emitter certifier.Emitter, handleErr certifier.ErrHandler, poll bool, interval time.Duration, useScorecardAPI bool) error {
 	runCertifier := func() error {
 		// compChan to collect query components
 		compChan := make(chan interface{}, BufferChannelSize)
@@ -70,7 +69,7 @@ func Certify(ctx context.Context, query certifier.QueryComponents, emitter certi
 		for !componentsCaptured {
 			select {
 			case d := <-compChan:
-				if err := generateDocuments(ctx, d, emitter, handleErr); err != nil {
+				if err := generateDocuments(ctx, d, emitter, handleErr, useScorecardAPI); err != nil {
 					return fmt.Errorf("generate certifier documents error: %w", err)
 				}
 			case err := <-errChan:
@@ -84,7 +83,7 @@ func Certify(ctx context.Context, query certifier.QueryComponents, emitter certi
 		}
 		for len(compChan) > 0 {
 			d := <-compChan
-			if err := generateDocuments(ctx, d, emitter, handleErr); err != nil {
+			if err := generateDocuments(ctx, d, emitter, handleErr, useScorecardAPI); err != nil {
 				logger.Errorf("generate certifier documents error: %v", err)
 			}
 		}
@@ -118,7 +117,7 @@ func Certify(ctx context.Context, query certifier.QueryComponents, emitter certi
 
 // generateDocuments runs CertifyVulns as a goroutine to scan and generates attestations that
 // are emitted as processor documents to be ingested
-func generateDocuments(ctx context.Context, collectedComponent interface{}, emitter certifier.Emitter, handleErr certifier.ErrHandler) error {
+func generateDocuments(ctx context.Context, collectedComponent interface{}, emitter certifier.Emitter, handleErr certifier.ErrHandler, useScorecardAPI bool) error {
 	// docChan to collect artifacts
 	docChan := make(chan *processor.Document, BufferChannelSize)
 	// errChan to receive error from collectors
@@ -129,7 +128,7 @@ func generateDocuments(ctx context.Context, collectedComponent interface{}, emit
 	for _, certifier := range documentCertifier {
 		c := certifier()
 		go func() {
-			errChan <- c.CertifyComponent(ctx, collectedComponent, docChan)
+			errChan <- c.CertifyComponent(ctx, collectedComponent, docChan, useScorecardAPI)
 		}()
 	}
 

diff --git a/pkg/certifier/certify/certify_test.go b/pkg/certifier/certify/certify_test.go
@@ -30,8 +30,7 @@ import (
 	"github.com/guacsec/guac/pkg/logging"
 )
 
-type mockQuery struct {
-}
+type mockQuery struct{}
 
 // NewMockQuery initializes the mockQuery to query for tests
 func newMockQuery() certifier.QueryComponents {
@@ -44,8 +43,7 @@ func (q *mockQuery) GetComponents(ctx context.Context, compChan chan<- interface
 	return nil
 }
 
-type mockUnknownQuery struct {
-}
+type mockUnknownQuery struct{}
 
 // NewMockQuery initializes the mockQuery to query for tests
 func newMockUnknownQuery() certifier.QueryComponents {
@@ -170,7 +168,6 @@ func TestCertify(t *testing.T) {
 	}}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-
 			ctx := logging.WithLogger(context.Background())
 			if tt.poll {
 				var cancel context.CancelFunc
@@ -184,7 +181,7 @@ func TestCertify(t *testing.T) {
 				return nil
 			}
 
-			err := Certify(ctx, tt.query, emit, errHandler, tt.poll, time.Second*1)
+			err := Certify(ctx, tt.query, emit, errHandler, tt.poll, time.Second*1, false)
 			if (err != nil) != tt.wantErr {
 				t.Errorf("Certify() error = %v, wantErr %v", err, tt.wantErr)
 			}

diff --git a/pkg/certifier/osv/osv.go b/pkg/certifier/osv/osv.go
@@ -22,17 +22,15 @@ import (
 	"net/http"
 	"time"
 
-	jsoniter "github.com/json-iterator/go"
-
 	osv_scanner "github.com/google/osv-scanner/pkg/osv"
-	intoto "github.com/in-toto/in-toto-golang/in_toto"
-
 	"github.com/guacsec/guac/pkg/certifier"
 	attestation_vuln "github.com/guacsec/guac/pkg/certifier/attestation"
 	"github.com/guacsec/guac/pkg/certifier/components/root_package"
 	"github.com/guacsec/guac/pkg/events"
 	"github.com/guacsec/guac/pkg/handler/processor"
 	"github.com/guacsec/guac/pkg/version"
+	intoto "github.com/in-toto/in-toto-golang/in_toto"
+	jsoniter "github.com/json-iterator/go"
 )
 
 var json = jsoniter.ConfigCompatibleWithStandardLibrary
@@ -62,7 +60,7 @@ func NewOSVCertificationParser() certifier.Certifier {
 
 // CertifyComponent takes in the root component from the gauc database and does a recursive scan
 // to generate vulnerability attestations
-func (o *osvCertifier) CertifyComponent(ctx context.Context, rootComponent interface{}, docChannel chan<- *processor.Document) error {
+func (o *osvCertifier) CertifyComponent(ctx context.Context, rootComponent interface{}, docChannel chan<- *processor.Document, _ bool) error {
 	packageNodes, ok := rootComponent.([]*root_package.PackageNode)
 	if !ok {
 		return ErrOSVComponenetTypeMismatch

diff --git a/pkg/certifier/osv/osv_test.go b/pkg/certifier/osv/osv_test.go
@@ -23,14 +23,13 @@ import (
 	"time"
 
 	osv_scanner "github.com/google/osv-scanner/pkg/osv"
-	attestation_vuln "github.com/guacsec/guac/pkg/certifier/attestation"
-	"github.com/guacsec/guac/pkg/certifier/components/root_package"
-	intoto "github.com/in-toto/in-toto-golang/in_toto"
-
 	"github.com/guacsec/guac/internal/testing/dochelper"
 	"github.com/guacsec/guac/internal/testing/testdata"
+	attestation_vuln "github.com/guacsec/guac/pkg/certifier/attestation"
+	"github.com/guacsec/guac/pkg/certifier/components/root_package"
 	"github.com/guacsec/guac/pkg/handler/processor"
 	"github.com/guacsec/guac/pkg/logging"
+	intoto "github.com/in-toto/in-toto-golang/in_toto"
 )
 
 func TestOSVCertifier_CertifyVulns(t *testing.T) {
@@ -150,7 +149,7 @@ func TestOSVCertifier_CertifyVulns(t *testing.T) {
 			defer close(docChan)
 			defer close(errChan)
 			go func() {
-				errChan <- o.CertifyComponent(ctx, tt.rootComponent, docChan)
+				errChan <- o.CertifyComponent(ctx, tt.rootComponent, docChan, false)
 			}()
 			numCollectors := 1
 			certifiersDone := 0

diff --git a/pkg/certifier/scorecard/README.md b/pkg/certifier/scorecard/README.md
@@ -0,0 +1,29 @@
+# Scorecard Certifier
+
+The Scorecard Certifier is a component that generates scorecard attestations for repositories. It uses the [OpenSSF Scorecard](https://github.com/ossf/scorecard) to evaluate the security posture of a repository.
+
+## How It Works
+
+### Initialization
+
+The `NewScorecardCertifier` function initializes the scorecard certifier. It checks if the `GITHUB_AUTH_TOKEN` is set in the environment. If not, it returns an error. The token is used to access the GitHub API.
+
+### Certifying Components
+
+The `CertifyComponent` function takes a `source.SourceNode` as input and generates a scorecard attestation. It uses the `GetScore` function to retrieve the scorecard data for the repository.
+
+### Using the Scorecard Library and GitHub Auth Token
+
+The `GetScore` function first checks if the `useScorecardAPI` flag is set to `true`. If it is, it calls the Scorecard API to retrieve the scorecard data. If the API call fails, it uses the Scorecard library and the GitHub auth token to retrieve the scorecard data.
+
+### Using the Scorecard API
+
+The Scorecard API is a public API that provides access to scorecard data. It can be used to retrieve scorecard data for any repository, regardless of whether the user has access to the GitHub repository. However, the API might fail if the repository does not exist in its database.
+
+### Differences
+
+The main difference between using the GitHub auth token/Scorecard library and the Scorecard API is that the GitHub auth token/Scorecard library requires access to the GitHub repository, while the Scorecard API does not.
+
+The Scorecard API is also more efficient than using the GitHub auth token/Scorecard library, as it does not need to download the entire repository.
+
+If the `useScorecardAPI` flag is not set, or the Scorecard API call fails, the certifier will default to using the GitHub auth token/Scorecard library.