updated storyblok detector (#3981)

* updated storyblok detector

* resolved abdul basit comments

pkg/detectors/storyblok/storyblok.go

@@ -2,9 +2,11 @@ package storyblok
 
 import (
 	"context"
-	regexp "github.com/wasilibs/go-re2"
+	"fmt"
+	"io"
 	"net/http"
-	"strings"
+
+	regexp "github.com/wasilibs/go-re2"
 
 	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
@@ -20,7 +22,7 @@ var (
 	client = common.SaneHttpClient()
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
-	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"storyblok"}) + `\b([0-9A-Za-z]{22}t{2})\b`)
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"storyblok"}) + `\b([0-9A-Za-z]{22}tt)\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.
@@ -29,33 +31,35 @@ func (s Scanner) Keywords() []string {
 	return []string{"storyblok"}
 }
 
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_StoryblokAccessToken
+}
+
+func (s Scanner) Description() string {
+	return "Storyblok is a headless CMS that allows developers to build flexible and powerful content management solutions. Storyblok tokens can be used to access and modify content within a Storyblok space."
+}
+
 // FromData will find and optionally verify Storyblok secrets in a given set of bytes.
 func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (results []detectors.Result, err error) {
 	dataStr := string(data)
 
-	matches := keyPat.FindAllStringSubmatch(dataStr, -1)
+	var uniqueAccessTokens = make(map[string]struct{})
 
-	for _, match := range matches {
-		resMatch := strings.TrimSpace(match[1])
+	for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueAccessTokens[match[1]] = struct{}{}
+	}
+
+	for accessToken := range uniqueAccessTokens {
 
 		s1 := detectors.Result{
-			DetectorType: detectorspb.DetectorType_Storyblok,
-			Raw:          []byte(resMatch),
+			DetectorType: detectorspb.DetectorType_StoryblokAccessToken,
+			Raw:          []byte(accessToken),
 		}
 
 		if verify {
-			req, err := http.NewRequestWithContext(ctx, "GET", "https://api.storyblok.com/v1/cdn/spaces/me/?token="+resMatch, nil)
-			if err != nil {
-				continue
-			}
-
-			res, err := client.Do(req)
-			if err == nil {
-				defer res.Body.Close()
-				if res.StatusCode >= 200 && res.StatusCode < 300 {
-					s1.Verified = true
-				}
-			}
+			isVerified, verificationErr := verifyStoryBlokAccessToken(ctx, client, accessToken)
+			s1.Verified = isVerified
+			s1.SetVerificationError(verificationErr)
 		}
 
 		results = append(results, s1)
@@ -64,10 +68,29 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 	return results, nil
 }
 
-func (s Scanner) Type() detectorspb.DetectorType {
-	return detectorspb.DetectorType_Storyblok
-}
+// docs: https://www.storyblok.com/docs/api/content-delivery/v2/getting-started/authentication
+func verifyStoryBlokAccessToken(ctx context.Context, client *http.Client, token string) (bool, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://api.storyblok.com/v1/cdn/spaces/me/?token="+token, nil)
+	if err != nil {
+		return false, err
+	}
 
-func (s Scanner) Description() string {
-	return "Storyblok is a headless CMS that allows developers to build flexible and powerful content management solutions. Storyblok tokens can be used to access and modify content within a Storyblok space."
+	resp, err := client.Do(req)
+	if err != nil {
+		return false, err
+	}
+
+	defer func() {
+		_, _ = io.Copy(io.Discard, resp.Body)
+		_ = resp.Body.Close()
+	}()
+
+	switch resp.StatusCode {
+	case http.StatusOK:
+		return true, nil
+	case http.StatusUnauthorized:
+		return false, nil
+	default:
+		return false, fmt.Errorf("unexpected status code: %d", resp.StatusCode)
+	}
 }

pkg/detectors/storyblok/storyblok_integration_test.go

@@ -48,7 +48,7 @@ func TestStoryblok_FromChunk(t *testing.T) {
 			},
 			want: []detectors.Result{
 				{
-					DetectorType: detectorspb.DetectorType_Storyblok,
+					DetectorType: detectorspb.DetectorType_StoryblokAccessToken,
 					Verified:     true,
 				},
 			},
@@ -64,7 +64,7 @@ func TestStoryblok_FromChunk(t *testing.T) {
 			},
 			want: []detectors.Result{
 				{
-					DetectorType: detectorspb.DetectorType_Storyblok,
+					DetectorType: detectorspb.DetectorType_StoryblokAccessToken,
 					Verified:     false,
 				},
 			},

pkg/detectors/storyblokpersonalaccesstoken/storyblok.go

@@ -0,0 +1,100 @@
+package storyblokpersonalaccesstoken
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+
+	regexp "github.com/wasilibs/go-re2"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+type Scanner struct{}
+
+// Ensure the Scanner satisfies the interface at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+
+var (
+	client = common.SaneHttpClient()
+
+	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"storyblok"}) + `\b([0-9A-Za-z]{22}tt-[0-9]{6}-[A-Za-z0-9_-]{20})\b`)
+)
+
+// Keywords are used for efficiently pre-filtering chunks.
+// Use identifiers in the secret preferably, or the provider name.
+func (s Scanner) Keywords() []string {
+	return []string{"storyblok"}
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_StoryblokPersonalAccessToken
+}
+
+func (s Scanner) Description() string {
+	return `Storyblok is a headless CMS that allows developers to build flexible and powerful content management solutions.
+			Storyblok personal access tokens can be used with management APIs.
+			The Storyblok Management API allows you to create, edit, update, and delete content using a common interface`
+}
+
+// FromData will find and optionally verify Storyblok secrets in a given set of bytes.
+func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (results []detectors.Result, err error) {
+	dataStr := string(data)
+
+	var uniquePATs = make(map[string]struct{})
+
+	for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) {
+		uniquePATs[match[1]] = struct{}{}
+	}
+
+	for pat := range uniquePATs {
+		s1 := detectors.Result{
+			DetectorType: detectorspb.DetectorType_StoryblokPersonalAccessToken,
+			Raw:          []byte(pat),
+		}
+
+		if verify {
+			isVerified, verificationErr := verifyStoryBlokPersonalAccessToken(ctx, client, pat)
+			s1.Verified = isVerified
+			s1.SetVerificationError(verificationErr)
+		}
+
+		results = append(results, s1)
+	}
+
+	return results, nil
+}
+
+// docs: http://storyblok.com/docs/api/management/core-resources/spaces/retrieve-multiple-spaces
+func verifyStoryBlokPersonalAccessToken(ctx context.Context, client *http.Client, token string) (bool, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://mapi.storyblok.com/v1/spaces/", nil)
+	if err != nil {
+		return false, err
+	}
+
+	// docs: https://www.storyblok.com/docs/api/management/getting-started/authentication
+	req.Header.Set("Authorization", token)
+
+	resp, err := client.Do(req)
+	if err != nil {
+		return false, err
+	}
+
+	defer func() {
+		_, _ = io.Copy(io.Discard, resp.Body)
+		_ = resp.Body.Close()
+	}()
+
+	switch resp.StatusCode {
+	case http.StatusOK:
+		return true, nil
+	case http.StatusUnauthorized:
+		return false, nil
+	default:
+		return false, fmt.Errorf("unexpected status code: %d", resp.StatusCode)
+	}
+}

pkg/detectors/storyblokpersonalaccesstoken/storyblok_integration_test.go

@@ -0,0 +1,121 @@
+//go:build detectors
+// +build detectors
+
+package storyblokpersonalaccesstoken
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/kylelemons/godebug/pretty"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestStoryblokPersonalAccessToken_FromChunk(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors5")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+
+	secret := testSecrets.MustGetField("STORYBLOK_PAT")
+	inactiveSecret := testSecrets.MustGetField("STORYBLOK_PAT_INACTIVE")
+
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+	tests := []struct {
+		name    string
+		s       Scanner
+		args    args
+		want    []detectors.Result
+		wantErr bool
+	}{
+		{
+			name: "found, verified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a storyblok secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_StoryblokPersonalAccessToken,
+					Verified:     true,
+				},
+			},
+			wantErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a storyblok secret %s within but not valid", inactiveSecret)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_StoryblokPersonalAccessToken,
+					Verified:     false,
+				},
+			},
+			wantErr: false,
+		},
+		{
+			name: "not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte("You cannot find the secret within"),
+				verify: true,
+			},
+			want:    nil,
+			wantErr: false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			s := Scanner{}
+			got, err := s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("Storyblok.FromData() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			for i := range got {
+				if len(got[i].Raw) == 0 {
+					t.Fatalf("no raw secret present: \n %+v", got[i])
+				}
+				got[i].Raw = nil
+			}
+			if diff := pretty.Compare(got, tt.want); diff != "" {
+				t.Errorf("Storyblok.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
+			}
+		})
+	}
+}
+
+func BenchmarkFromData(benchmark *testing.B) {
+	ctx := context.Background()
+	s := Scanner{}
+	for name, data := range detectors.MustGetBenchmarkData() {
+		benchmark.Run(name, func(b *testing.B) {
+			b.ResetTimer()
+			for n := 0; n < b.N; n++ {
+				_, err := s.FromData(ctx, false, data)
+				if err != nil {
+					b.Fatal(err)
+				}
+			}
+		})
+	}
+}

pkg/detectors/storyblokpersonalaccesstoken/storyblok_test.go

@@ -0,0 +1,91 @@
+package storyblokpersonalaccesstoken
+
+import (
+	"context"
+	"fmt"
+	"testing"
+
+	"github.com/google/go-cmp/cmp"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick"
+)
+
+var (
+	validPattern   = "5r7EgNfakeXi6ZoEls1twAtt-001100-Q5E2fKfakeRqsUjwmsJn"
+	invalidPattern = "5r7EgNfakeXi6ZoEls1twAt-001100-Q5E2fKfakeRqsUjwmsJn"
+	keyword        = "storyblok"
+)
+
+func TestStoryblokPersonalAccessToken_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name:  "valid pattern - with keyword storyblok",
+			input: fmt.Sprintf("%s token = '%s'", keyword, validPattern),
+			want:  []string{validPattern},
+		},
+		{
+			name:  "valid pattern - ignore duplicate",
+			input: fmt.Sprintf("%s token = '%s' | '%s'", keyword, validPattern, validPattern),
+			want:  []string{validPattern},
+		},
+		{
+			name:  "valid pattern - key out of prefix range",
+			input: fmt.Sprintf("%s keyword is not close to the real key in the data\n = '%s'", keyword, validPattern),
+			want:  []string{},
+		},
+		{
+			name:  "invalid pattern",
+			input: fmt.Sprintf("%s = '%s'", keyword, invalidPattern),
+			want:  []string{},
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			matchedDetectors := ahoCorasickCore.FindDetectorMatches([]byte(test.input))
+			if len(matchedDetectors) == 0 {
+				t.Errorf("keywords '%v' not matched by: %s", d.Keywords(), test.input)
+				return
+			}
+
+			results, err := d.FromData(context.Background(), false, []byte(test.input))
+			if err != nil {
+				t.Errorf("error = %v", err)
+				return
+			}
+
+			if len(results) != len(test.want) {
+				if len(results) == 0 {
+					t.Errorf("did not receive result")
+				} else {
+					t.Errorf("expected %d results, only received %d", len(test.want), len(results))
+				}
+				return
+			}
+
+			actual := make(map[string]struct{}, len(results))
+			for _, r := range results {
+				if len(r.RawV2) > 0 {
+					actual[string(r.RawV2)] = struct{}{}
+				} else {
+					actual[string(r.Raw)] = struct{}{}
+				}
+			}
+			expected := make(map[string]struct{}, len(test.want))
+			for _, v := range test.want {
+				expected[v] = struct{}{}
+			}
+
+			if diff := cmp.Diff(expected, actual); diff != "" {
+				t.Errorf("%s diff: (-want +got)\n%s", test.name, diff)
+			}
+		})
+	}
+}

pkg/engine/defaults/defaults.go

@@ -686,6 +686,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/stormboard"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/stormglass"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/storyblok"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/storyblokpersonalaccesstoken"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/storychief"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/strava"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/streak"
@@ -1536,6 +1537,7 @@ func buildDetectorList() []detectors.Detector {
 		&stormboard.Scanner{},
 		&stormglass.Scanner{},
 		&storyblok.Scanner{},
+		&storyblokpersonalaccesstoken.Scanner{},
 		&storychief.Scanner{},
 		&strava.Scanner{},
 		&streak.Scanner{},

proto/detectors.proto

@@ -232,7 +232,7 @@ enum DetectorType {
   Duda = 219;
   Yext = 220;
   ContentStack = 221;
-  Storyblok = 222;
+  StoryblokAccessToken = 222;
   GraphCMS = 223;
   Checkmarket = 224;
   Convertkit = 225;
@@ -1025,6 +1025,7 @@ enum DetectorType {
   AzureRefreshToken = 1013;
   AirtableOAuth = 1014;
   AirtablePersonalAccessToken = 1015;
+  StoryblokPersonalAccessToken = 1016;
 }
 
 message Result {