add salesforce detector (#1608)

* setup

* update time out case to return detector result

* fix

* remove unneeded comment

* remove debug print

* cleanup

* more robust error handling

* reflect new detector template changes

* fixes

* mark response body check err as indeterminate

pkg/detectors/salesforce/salesforce.go

@@ -0,0 +1,113 @@
+package salesforce
+
+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"
+)
+
+const (
+	currentVersion = "58.0" // current Salesforce version
+)
+
+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.
+	accessTokenPat = regexp.MustCompile(`\b00[a-zA-Z0-9]{13}![a-zA-Z0-9_.]{96}\b`)
+	instancePat    = regexp.MustCompile(`\bhttps://[0-9a-zA-Z-\.]{1,100}\.my\.salesforce\.com\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{"salesforce"}
+}
+
+// FromData will find and optionally verify Salesforce 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)
+
+	instanceMatches := instancePat.FindAllStringSubmatch(dataStr, -1)
+	tokenMatches := accessTokenPat.FindAllStringSubmatch(dataStr, -1)
+
+	fmt.Printf("instanceMatches: %v\n", instanceMatches)
+
+	for _, instance := range instanceMatches {
+		if len(instance) != 1 {
+			continue
+		}
+
+		instanceMatch := strings.TrimSpace(instance[0])
+
+		for _, token := range tokenMatches {
+			if len(token) != 1 {
+				continue
+			}
+
+			tokenMatch := strings.TrimSpace(token[0])
+
+			s1 := detectors.Result{
+				DetectorType: detectorspb.DetectorType_Salesforce,
+				Raw:          []byte(tokenMatch),
+			}
+
+			if verify {
+				client := s.client
+				if client == nil {
+					client = defaultClient
+				}
+
+				req, err := http.NewRequestWithContext(ctx, "GET", instanceMatch+"/services/data/v"+currentVersion+"/query?q=SELECT+name+from+Account", nil)
+				if err != nil {
+					continue
+				}
+				req.Header.Set("Authorization", "Bearer "+tokenMatch)
+
+				res, err := client.Do(req)
+
+				if err != nil {
+					// End execution, append Detector Result if request fails to prevent panic on response body checks
+					s1.VerificationError = err
+					results = append(results, s1)
+					continue
+				}
+
+				verifiedBodyResponse, err := common.ResponseContainsSubstring(res.Body, "records")
+
+				defer res.Body.Close()
+				if res.StatusCode >= 200 && res.StatusCode < 300 && verifiedBodyResponse {
+					s1.Verified = true
+				} else if res.StatusCode >= 200 && res.StatusCode < 300 && !verifiedBodyResponse {
+					s1.Verified = false
+				} else if res.StatusCode == 401 {
+					s1.Verified = false
+				} else {
+					s1.VerificationError = fmt.Errorf("request to %v returned status %d with error %+v", res.Request.URL, res.StatusCode, err)
+				}
+
+			}
+
+			results = append(results, s1)
+
+		}
+	}
+
+	return results, nil
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_Salesforce
+}

pkg/detectors/salesforce/salesforce_test.go

@@ -0,0 +1,167 @@
+//go:build detectors
+// +build detectors
+
+package salesforce
+
+import (
+	"context"
+	"fmt"
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
+	"testing"
+	"time"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestSalesforce_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)
+	}
+	instance := testSecrets.MustGetField("SALESFORCE_INSTANCE")
+	// Salesforce secrets are not valid for long, so we may need to regenerate them.
+	// Steps to regenerate:
+	// 1. Install Salesforce CLI
+	// 2. Authenticate with `sfdx org web login`. This will open a browser window. Use the credentials in the detection test accounts 1Pass vault.
+	// 3. Run `sfdx org display --targetusername [email protected]`to obtain new token.
+	// Source: https://developer.salesforce.com/docs/atlas.en-us.api_rest.meta/api_rest/quickstart_oauth.htm
+	token := testSecrets.MustGetField("SALESFORCE_TOKEN")
+	inactiveToken := testSecrets.MustGetField("SALESFORCE_INACTIVE_TOKEN")
+
+	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 salesforce secret within %s for %s", token, instance)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Salesforce,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a salesforce secret %s within but not valid for %s", inactiveToken, instance)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Salesforce,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: 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,
+			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 salesforce secret within %s for %s", token, instance)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Salesforce,
+					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 salesforce secret within %s for %s", token, instance)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Salesforce,
+					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("Salesforce.FromData() error = %v, wantErr %v", err, tt.wantErr)
+			}
+			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")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Salesforce.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) {
+			for n := 0; n < b.N; n++ {
+				_, err := s.FromData(ctx, false, data)
+				if err != nil {
+					b.Fatal(err)
+				}
+			}
+		})
+	}
+}

proto/detectors.proto

@@ -933,7 +933,7 @@ enum DetectorType {
   TrufflehogEnterprise = 922;
   EnvoyApiKey = 923;
   GitHubOauth2 = 924;
-  // 925 reserved for Salesforce
+  Salesforce = 925;
   HuggingFace = 926;
 }