Support indeterminate verification in Gitlab detector (#1613)

trufflesecurity · Aug 11, 2023 · c9f58b3 · c9f58b3
1 parent 62cbef5
commit c9f58b3
Show file tree

Hide file tree

Showing 2 changed files with 86 additions and 15 deletions.
diff --git a/pkg/detectors/gitlab/gitlab.go b/pkg/detectors/gitlab/gitlab.go
@@ -12,7 +12,11 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 )
 
-type Scanner struct{ detectors.EndpointSetter }
+type Scanner struct {
+	client *http.Client
+
+	detectors.EndpointSetter
+}
 
 // Ensure the Scanner satisfies the interfaces at compile time.
 var _ detectors.Detector = (*Scanner)(nil)
@@ -23,7 +27,8 @@ func (Scanner) Version() int            { return 1 }
 func (Scanner) DefaultEndpoint() string { return "https://gitlab.com" }
 
 var (
-	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"gitlab"}) + `\b((?:glpat|)[a-zA-Z0-9\-=_]{20,22})\b`)
+	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.
@@ -59,7 +64,10 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 			// 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 := common.SaneHttpClient()
+			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)
@@ -74,9 +82,19 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 					// 200 means good key and has `read_user` scope
 					// 403 means good key but not the right scope
 					// 401 is bad key
-					if res.StatusCode == http.StatusOK || res.StatusCode == http.StatusForbidden {
+					switch res.StatusCode {
+					case http.StatusOK:
+						secret.Verified = true
+					case http.StatusForbidden:
+						// Good key but not the right scope
 						secret.Verified = true
+					case http.StatusUnauthorized:
+						// Nothing to do; zero values are the ones we want
+					default:
+						secret.VerificationError = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
 					}
+				} else {
+					secret.VerificationError = err
 				}
 			}
 		}

diff --git a/pkg/detectors/gitlab/gitlab_test.go b/pkg/detectors/gitlab/gitlab_test.go
@@ -6,11 +6,11 @@ package gitlab
 import (
 	"context"
 	"fmt"
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
 	"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"
@@ -31,11 +31,12 @@ func TestGitlab_FromChunk(t *testing.T) {
 		verify bool
 	}
 	tests := []struct {
-		name    string
-		s       Scanner
-		args    args
-		want    []detectors.Result
-		wantErr bool
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
 	}{
 		{
 			name: "found",
@@ -85,6 +86,56 @@ func TestGitlab_FromChunk(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "found, would be verified but for timeout",
+			s:    Scanner{client: common.SaneHttpClientTimeOut(1 * time.Microsecond)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab super secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Gitlab,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "found and valid but unexpected api response",
+			s:    Scanner{client: common.ConstantResponseHttpClient(500, "")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab super secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Gitlab,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "found, good key but wrong scope",
+			s:    Scanner{client: common.ConstantResponseHttpClient(403, "")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a gitlab super secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Gitlab,
+					Verified:     true,
+				},
+			},
+			wantErr: false,
+		},
 		{
 			name: "not found",
 			s:    Scanner{},
@@ -99,8 +150,7 @@ func TestGitlab_FromChunk(t *testing.T) {
 	}
 	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)
+			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
@@ -109,9 +159,12 @@ func TestGitlab_FromChunk(t *testing.T) {
 				if len(got[i].Raw) == 0 {
 					t.Fatal("no raw secret present")
 				}
-				got[i].Raw = nil
+				if (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Fatalf(" wantVerificationError = %v, verification error = %v,", tt.wantVerificationErr, got[i].VerificationError)
+				}
 			}
-			if diff := pretty.Compare(got, tt.want); diff != "" {
+			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)
 			}
 		})