cleanup nesting

delete test file (note: these are not real secrets)

pkg/detectors/gitlab/gitlab.go

@@ -14,21 +14,20 @@ import (
 
 type Scanner struct {
 	client *http.Client
-
 	detectors.EndpointSetter
 }
 
 // Ensure the Scanner satisfies the interfaces at compile time.
 var _ detectors.Detector = (*Scanner)(nil)
-var _ detectors.Versioner = (*Scanner)(nil)
 var _ detectors.EndpointCustomizer = (*Scanner)(nil)
+var _ detectors.Versioner = (*Scanner)(nil)
 
 func (Scanner) Version() int            { return 1 }
 func (Scanner) DefaultEndpoint() string { return "https://gitlab.com" }
 
 var (
 	defaultClient = common.SaneHttpClient()
-	keyPat        = regexp.MustCompile(detectors.PrefixRegex([]string{"gitlab"}) + `\b((?:glpat|)[a-zA-Z0-9\-=_]{20,22})\b`)
+	keyPat        = regexp.MustCompile(detectors.PrefixRegex([]string{"gitlab"}) + `\b[^glpat-]([a-zA-Z0-9\-=_]{20,22})\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.
@@ -42,61 +41,20 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 	dataStr := string(data)
 
 	matches := keyPat.FindAllStringSubmatch(dataStr, -1)
-
 	for _, match := range matches {
 		if len(match) != 2 {
 			continue
 		}
 		resMatch := strings.TrimSpace(match[1])
-		if strings.Contains(match[0], "glpat") {
-			keyString := strings.Split(match[0], " ")
-			resMatch = keyString[len(keyString)-1]
-		}
-
 		s1 := detectors.Result{
 			DetectorType: detectorspb.DetectorType_Gitlab,
-			Raw:          []byte(match[1]),
+			Raw:          []byte(resMatch),
 		}
 
 		if verify {
-			// there are 4 read 'scopes' for a gitlab token: api, read_user, read_repo, and read_registry
-			// they all grant access to different parts of the API. I couldn't find an endpoint that every
-			// one of these scopes has access to, so we just check an example endpoint for each scope. If any
-			// of them contain data, we know we have a valid key, but if they all fail, we don't
-
-			client := s.client
-			if client == nil {
-				client = defaultClient
-			}
-			for _, baseURL := range s.Endpoints(s.DefaultEndpoint()) {
-				// test `read_user` scope
-				req, err := http.NewRequestWithContext(ctx, "GET", baseURL+"/api/v4/user", nil)
-				if err != nil {
-					continue
-				}
-				req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", resMatch))
-				res, err := client.Do(req)
-				if err == nil {
-					res.Body.Close() // The request body is unused.
-
-					// 200 means good key and has `read_user` scope
-					// 403 means good key but not the right scope
-					// 401 is bad key
-					switch res.StatusCode {
-					case http.StatusOK:
-						s1.Verified = true
-					case http.StatusForbidden:
-						// Good key but not the right scope
-						s1.Verified = true
-					case http.StatusUnauthorized:
-						// Nothing to do; zero values are the ones we want
-					default:
-						s1.VerificationError = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
-					}
-				} else {
-					s1.VerificationError = err
-				}
-			}
+			isVerified, verificationErr := s.verifyGitlab(ctx, resMatch)
+			s1.Verified = isVerified
+			s1.VerificationError = verificationErr
 		}
 
 		if !s1.Verified && detectors.IsKnownFalsePositive(string(s1.Raw), detectors.DefaultFalsePositives, true) {
@@ -109,6 +67,49 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 	return results, nil
 }
 
+func (s Scanner) verifyGitlab(ctx context.Context, resMatch string) (bool, error) {
+	// there are 4 read 'scopes' for a gitlab token: api, read_user, read_repo, and read_registry
+	// they all grant access to different parts of the API. I couldn't find an endpoint that every
+	// one of these scopes has access to, so we just check an example endpoint for each scope. If any
+	// of them contain data, we know we have a valid key, but if they all fail, we don't
+
+	client := s.client
+	if client == nil {
+		client = defaultClient
+	}
+	for _, baseURL := range s.Endpoints(s.DefaultEndpoint()) {
+		// test `read_user` scope
+		req, err := http.NewRequestWithContext(ctx, "GET", baseURL+"/api/v4/user", nil)
+		if err != nil {
+			continue
+		}
+		req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", resMatch))
+		res, err := client.Do(req)
+		if err != nil {
+			return false, err
+		}
+		defer res.Body.Close() // The request body is unused.
+
+		// 200 means good key and has `read_user` scope
+		// 403 means good key but not the right scope
+		// 401 is bad key
+		switch res.StatusCode {
+		case http.StatusOK:
+			return true, nil
+		case http.StatusForbidden:
+			// Good key but not the right scope
+			return true, nil
+		case http.StatusUnauthorized:
+			// Nothing to do; zero values are the ones we want
+			return false, nil
+		default:
+			return false, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+		}
+
+	}
+	return false, nil
+}
+
 func (s Scanner) Type() detectorspb.DetectorType {
 	return detectorspb.DetectorType_Gitlab
 }

pkg/detectors/gitlab/gitlab_v1_test.go

@@ -17,7 +17,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 )
 
-func TestGitlab_FromChunk(t *testing.T) {
+func TestGitlab_FromChunk_WithV2Secrets(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
 	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")

pkg/detectors/gitlab/gitlab_v2_test.go

@@ -0,0 +1,113 @@
+//go:build detectors
+// +build detectors
+
+package gitlab
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+)
+
+// This test ensures gitlab v1 detector does not work on gitlab v2 secrets
+func TestGitlab_FromChunk(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+	secret := testSecrets.MustGetField("GITLABV2")
+	secretInactive := testSecrets.MustGetField("GITLABV2_INACTIVE")
+
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+	tests := []struct {
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
+	}{
+		{
+			name: "verified secret, not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab secret %s within", secret)),
+				verify: true,
+			},
+			want:    nil,
+			wantErr: false,
+		},
+		{
+			name: "unverified secret, not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab secret %s within", secretInactive)),
+				verify: true,
+			},
+			want:    nil,
+			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) {
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("Gitlab.FromData() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			for i := range got {
+				if len(got[i].Raw) == 0 {
+					t.Fatal("no raw secret present")
+				}
+				if (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Fatalf(" wantVerificationError = %v, verification error = %v,", tt.wantVerificationErr, got[i].VerificationError)
+				}
+			}
+			opts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "VerificationError")
+			if diff := cmp.Diff(got, tt.want, opts); diff != "" {
+				t.Errorf("Gitlab.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
+			}
+		})
+	}
+}
+
+func BenchmarkV2FromData(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/gitlabV2/gitlab_v1_test.go

@@ -0,0 +1,124 @@
+//go:build detectors
+// +build detectors
+
+package gitlabV2
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+)
+
+// This test ensures gitlab v2 detector does not work on gitlab v1 secrets
+func TestGitlabV2_FromChunk_WithV1Secrets(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+	secret := testSecrets.MustGetField("GITLAB")
+	secretInactive := testSecrets.MustGetField("GITLAB_INACTIVE")
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+	tests := []struct {
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
+	}{
+		{
+			name: "verified v1 secret, not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab super secret %s within", secret)),
+				verify: true,
+			},
+			want:    nil,
+			wantErr: false,
+		},
+		{
+			name: "verified v1 secret, not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("gitlab %s", secret)),
+				verify: true,
+			},
+			want:    nil,
+			wantErr: false,
+		},
+		{
+			name: "unverified v1 secret, not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab secret %s within", secretInactive)),
+				verify: true,
+			},
+			want:    nil,
+			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) {
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("Gitlab.FromData() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			for i := range got {
+				if len(got[i].Raw) == 0 {
+					t.Fatal("no raw secret present")
+				}
+				if (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Fatalf(" wantVerificationError = %v, verification error = %v,", tt.wantVerificationErr, got[i].VerificationError)
+				}
+			}
+			opts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "VerificationError")
+			if diff := cmp.Diff(got, tt.want, opts); diff != "" {
+				t.Errorf("Gitlab.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/gitlabV2/gitlab_v2.go

@@ -0,0 +1,115 @@
+package gitlabV2
+
+import (
+	"context"
+	"fmt"
+	"net/http"
+	"regexp"
+	"strings"
+
+	"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 {
+	client *http.Client
+	detectors.EndpointSetter
+}
+
+// Ensure the Scanner satisfies the interfaces at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+var _ detectors.EndpointCustomizer = (*Scanner)(nil)
+var _ detectors.Versioner = (*Scanner)(nil)
+
+func (Scanner) Version() int            { return 2 }
+func (Scanner) DefaultEndpoint() string { return "https://gitlab.com" }
+
+var (
+	defaultClient = common.SaneHttpClient()
+	keyPat        = regexp.MustCompile(detectors.PrefixRegex([]string{"gitlab"}) + `\b(glpat-[a-zA-Z0-9\-=_]{20,22})\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{"gitlab"}
+}
+
+// FromData will find and optionally verify Gitlab 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)
+	for _, match := range matches {
+		if len(match) != 2 {
+			continue
+		}
+		resMatch := strings.TrimSpace(match[1])
+		s1 := detectors.Result{
+			DetectorType: detectorspb.DetectorType_Gitlab,
+			Raw:          []byte(resMatch),
+		}
+
+		if verify {
+			isVerified, verificationErr := s.verifyGitlab(ctx, resMatch)
+			s1.Verified = isVerified
+			s1.VerificationError = verificationErr
+		}
+
+		if !s1.Verified && detectors.IsKnownFalsePositive(string(s1.Raw), detectors.DefaultFalsePositives, true) {
+			continue
+		}
+
+		results = append(results, s1)
+	}
+
+	return results, nil
+}
+
+func (s Scanner) verifyGitlab(ctx context.Context, resMatch string) (bool, error) {
+	// there are 4 read 'scopes' for a gitlab token: api, read_user, read_repo, and read_registry
+	// they all grant access to different parts of the API. I couldn't find an endpoint that every
+	// one of these scopes has access to, so we just check an example endpoint for each scope. If any
+	// of them contain data, we know we have a valid key, but if they all fail, we don't
+
+	client := s.client
+	if client == nil {
+		client = defaultClient
+	}
+	for _, baseURL := range s.Endpoints(s.DefaultEndpoint()) {
+		// test `read_user` scope
+		req, err := http.NewRequestWithContext(ctx, "GET", baseURL+"/api/v4/user", nil)
+		if err != nil {
+			continue
+		}
+		req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", resMatch))
+		res, err := client.Do(req)
+		if err != nil {
+			return false, err
+		}
+		defer res.Body.Close() // The request body is unused.
+
+		// 200 means good key and has `read_user` scope
+		// 403 means good key but not the right scope
+		// 401 is bad key
+		switch res.StatusCode {
+		case http.StatusOK:
+			return true, nil
+		case http.StatusForbidden:
+			// Good key but not the right scope
+			return true, nil
+		case http.StatusUnauthorized:
+			// Nothing to do; zero values are the ones we want
+			return false, nil
+		default:
+			return false, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+		}
+
+	}
+	return false, nil
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_Gitlab
+}

pkg/detectors/gitlabV2/gitlab_v2_test.go

@@ -1,7 +1,7 @@
 //go:build detectors
 // +build detectors
 
-package gitlabv2
+package gitlabV2
 
 import (
 	"context"
@@ -16,7 +16,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 )
 
-func TestGitlab_FromChunk(t *testing.T) {
+func TestGitlabV2_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
 	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")
@@ -156,7 +156,7 @@ func TestGitlab_FromChunk(t *testing.T) {
 	}
 }
 
-func BenchmarkFromData(benchmark *testing.B) {
+func BenchmarkV2FromData(benchmark *testing.B) {
 	ctx := context.Background()
 	s := Scanner{}
 	for name, data := range detectors.MustGetBenchmarkData() {

pkg/detectors/gitlabv2/gitlab.go

@@ -1,107 +0,0 @@
-package gitlabv2
-
-import (
-	"context"
-	"fmt"
-	"net/http"
-	"regexp"
-
-	"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 {
-	client *http.Client
-	detectors.EndpointSetter
-}
-
-// Ensure the Scanner satisfies the interfaces at compile time.
-var _ detectors.Detector = (*Scanner)(nil)
-var _ detectors.Versioner = (*Scanner)(nil)
-var _ detectors.EndpointCustomizer = (*Scanner)(nil)
-
-func (Scanner) Version() int            { return 2 }
-func (Scanner) DefaultEndpoint() string { return "https://gitlab.com" }
-
-var (
-	defaultClient = common.SaneHttpClient()
-	keyPat        = regexp.MustCompile(`\b(glpat-[a-zA-Z0-9\-=_]{20,22})\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{"glpat-"}
-}
-
-// FromData will find and optionally verify Gitlab 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)
-
-	for _, match := range matches {
-		if len(match) != 2 {
-			continue
-		}
-
-		s1 := detectors.Result{
-			DetectorType: detectorspb.DetectorType_Gitlab,
-			Raw:          []byte(match[1]),
-		}
-
-		if verify {
-			// there are 4 read 'scopes' for a gitlab token: api, read_user, read_repo, and read_registry
-			// they all grant access to different parts of the API. I couldn't find an endpoint that every
-			// one of these scopes has access to, so we just check an example endpoint for each scope. If any
-			// of them contain data, we know we have a valid key, but if they all fail, we don't
-			client := s.client
-			if client == nil {
-				client = defaultClient
-			}
-			for _, baseURL := range s.Endpoints(s.DefaultEndpoint()) {
-				// test `read_user` scope
-				req, err := http.NewRequestWithContext(ctx, "GET", baseURL+"/api/v4/user", nil)
-				if err != nil {
-					continue
-				}
-				req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", match[1]))
-				res, err := client.Do(req)
-				if err == nil {
-					res.Body.Close() // The request body is unused.
-
-					// 200 means good key and has `read_user` scope
-					// 403 means good key but not the right scope
-					// 401 is bad key
-					switch res.StatusCode {
-					case http.StatusOK:
-						s1.Verified = true
-					case http.StatusForbidden:
-						// Good key but not the right scope
-						s1.Verified = true
-					case http.StatusUnauthorized:
-						// Nothing to do; zero values are the ones we want
-					default:
-						s1.VerificationError = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
-					}
-				} else {
-					s1.VerificationError = err
-				}
-			}
-
-		}
-
-		if !s1.Verified && detectors.IsKnownFalsePositive(string(s1.Raw), detectors.DefaultFalsePositives, true) {
-			continue
-		}
-
-		results = append(results, s1)
-	}
-
-	return
-}
-
-func (s Scanner) Type() detectorspb.DetectorType {
-	return detectorspb.DetectorType_Gitlab
-}

pkg/engine/defaults.go

@@ -297,7 +297,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/github_old"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/githubapp"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/gitlab"
-	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/gitlabv2"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/gitlabV2"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/gitter"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/glassnode"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/glitterlyapi"
@@ -769,8 +769,8 @@ func DefaultDetectors() []detectors.Detector {
 		aws.New(),
 		&azure.Scanner{},
 		&slack.Scanner{}, // has 4 secret types
-		&gitlabv2.Scanner{},
 		&gitlab.Scanner{},
+		&gitlabV2.Scanner{},
 		&sendgrid.Scanner{},
 		&mailchimp.Scanner{},
 		&okta.Scanner{},