Followup to PR #4208 : LambdaTest secret detector

pkg/detectors/lambdatest/lambdatest.go

@@ -0,0 +1,140 @@
+package lambdatest
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"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 {
+	client *http.Client
+}
+
+// Ensure the Scanner satisfies the interface at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+
+var (
+	defaultClient = common.SaneHttpClient()
+	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
+	usernamePat = regexp.MustCompile(detectors.PrefixRegex([]string{
+		"hub.lambdatest.com",
+		"userName",
+		"\"username\":",
+		"USER_NAME",
+		"user",
+		"lambdatestUser",
+		"LT_USERNAME",
+		"LAMBDATEST_USERNAME",
+	}) + `\b([a-zA-Z0-9]+)\b`)
+
+	accessKeyPat = regexp.MustCompile(detectors.PrefixRegex([]string{
+		"hub.lambdatest.com",
+		"accessKey",
+		"\"access_Key\":",
+		"ACCESS_KEY",
+		"key",
+		"lambdatestKey",
+		"LT_AUTHKEY",
+		"LT_ACCESS_KEY",
+	}) + `\b(LT_[a-zA-Z0-9]{47})\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{"lambdatest"}
+}
+
+// FromData will find and optionally verify Lambdatest 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)
+	uniqueUsernameMatches := make(map[string]struct{})
+	for _, match := range usernamePat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueUsernameMatches[match[1]] = struct{}{}
+	}
+
+	uniqueAccessKeyMatches := make(map[string]struct{})
+	for _, match := range accessKeyPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueAccessKeyMatches[match[1]] = struct{}{}
+	}
+	for usernameMatch := range uniqueUsernameMatches {
+		for accessKeyMatch := range uniqueAccessKeyMatches {
+
+			s1 := detectors.Result{
+				DetectorType: detectorspb.DetectorType_LambdaTest,
+				Raw:          []byte(fmt.Sprintf("%s:%s", usernameMatch, accessKeyMatch)),
+				RawV2:        []byte(fmt.Sprintf("%s:%s", usernameMatch, accessKeyMatch)),
+			}
+
+			if verify {
+				client := s.client
+				if client == nil {
+					client = defaultClient
+				}
+
+				isVerified, verificationErr := verifyMatch(ctx, client, usernameMatch, accessKeyMatch)
+				s1.Verified = isVerified
+				s1.SetVerificationError(verificationErr, usernameMatch)
+			}
+
+			results = append(results, s1)
+		}
+	}
+
+	return results, nil
+}
+
+func verifyMatch(ctx context.Context, client *http.Client, usernameMatch string, accessKeyMatch string) (bool, error) {
+	body := map[string]string{
+		"username": usernameMatch,
+		"token":    accessKeyMatch,
+	}
+
+	// encode the body as JSON
+	jsonBody, err := json.Marshal(body)
+	if err != nil {
+		return false, err
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, "https://auth.lambdatest.com/api/user/token/auth", bytes.NewBuffer(jsonBody))
+	if err != nil {
+		return false, nil
+	}
+	req.Header.Set("Content-Type", "application/json")
+
+	res, err := client.Do(req)
+	if err != nil {
+		return false, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	switch res.StatusCode {
+	case http.StatusOK:
+		// If the endpoint returns useful information, we can return it as a map.
+		return true, nil
+	case http.StatusUnauthorized:
+		// The secret is determinately not verified (nothing to do)
+		return false, nil
+	default:
+		return false, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+	}
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_LambdaTest
+}
+
+func (s Scanner) Description() string {
+	return "LambdaTest is a cloud-based cross-browser testing platform that allows developers to test their web applications across various browsers and devices."
+}

pkg/detectors/lambdatest/lambdatest_integration_test.go

@@ -0,0 +1,163 @@
+//go:build detectors
+// +build detectors
+
+package lambdatest
+
+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"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestLambdatest_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)
+	}
+	username := testSecrets.MustGetField("LAMBDATEST_USERNAME")
+	accessKey := testSecrets.MustGetField("LAMBDATEST_ACCESSKEY")
+	invalidUsername := testSecrets.MustGetField("LAMBDATEST_INVALID_USERNAME")
+	invalidAccessKey := testSecrets.MustGetField("LAMBDATEST_INVALID_ACCESSKEY")
+
+	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: "found, verified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a lambdatest username %s and lambdatest accesskey %s within", username, accessKey)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_LambdaTest,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a lambdatest username %s and lambdatest accesskey %s within but not valid", invalidUsername, invalidAccessKey)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_LambdaTest,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte("You cannot find the secret within"),
+				verify: true,
+			},
+			want:                nil,
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, would be verified if not for timeout",
+			s:    Scanner{client: common.SaneHttpClientTimeOut(1 * time.Microsecond)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a lambdatest username %s and lambdatest accesskey %s within", username, accessKey)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_LambdaTest,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "found, verified but unexpected api surface",
+			s:    Scanner{client: common.ConstantResponseHttpClient(404, "")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a lambdatest username %s and lambdatest accesskey %s within", username, accessKey)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_LambdaTest,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+	}
+	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("Lambdatest.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])
+				}
+				if (got[i].VerificationError() != nil) != tt.wantVerificationErr {
+					t.Fatalf("wantVerificationError = %v, verification error = %v", tt.wantVerificationErr, got[i].VerificationError())
+				}
+			}
+			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "verificationError", "primarySecret")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Lambdatest.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/lambdatest/lambdatest_test.go

@@ -0,0 +1,76 @@
+package lambdatest
+
+import (
+	"context"
+	"testing"
+
+	"github.com/google/go-cmp/cmp"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick"
+)
+
+func TestLambdatest_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name: "typical pattern",
+			input: `lambdatest_username = sampleUserName123
+					lambdatest_accessKey = LT_depBE6n7BUl4itF6aWhB0KVPUirnGtp3wu15I2yec60DByD`,
+			want: []string{"sampleUserName123" + ":" + "LT_depBE6n7BUl4itF6aWhB0KVPUirnGtp3wu15I2yec60DByD"},
+		},
+		{
+			name: "invalid pattern",
+			input: `lambdatest_username = sampleUserName123
+					lambdatest_token = 1a2b3c4d`,
+			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

@@ -409,6 +409,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/kraken"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/kucoin"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/kylas"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/lambdatest"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/langfuse"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/langsmith"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/languagelayer"
@@ -1271,6 +1272,7 @@ func buildDetectorList() []detectors.Detector {
 		&kraken.Scanner{},
 		&kucoin.Scanner{},
 		&kylas.Scanner{},
+		&lambdatest.Scanner{},
 		&langfuse.Scanner{},
 		&langsmith.Scanner{},
 		&languagelayer.Scanner{},