Feature: Airtable OAuth Detector (#3868)

* added airtable oauth detector

* updated regex pattern for airtable oauth

* added comment on airtable oauth token regex

pkg/detectors/airtableoauth/airtableoauth.go

@@ -0,0 +1,104 @@
+package airtableoauth
+
+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 {
+	client *http.Client
+}
+
+// Ensure the Scanner satisfies the interface at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+
+var (
+	defaultClient = common.SaneHttpClient()
+	// The detector will attempt to match access tokens generated through the Airtable OAuth flow
+	// Airtable OAuth does not support generating access tokens using client ID and key
+	// Reference: https://airtable.com/developers/web/api/oauth-reference
+	tokenPat = regexp.MustCompile(`\b([[:alnum:]]+\.v1\.[a-zA-Z0-9_-]+\.[a-f0-9]+)\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{"airtable"}
+}
+
+// FromData will find and optionally verify AirtableOAuth 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)
+
+	uniqueMatches := make(map[string]struct{})
+	for _, match := range tokenPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueMatches[match[1]] = struct{}{}
+	}
+
+	for match := range uniqueMatches {
+		s1 := detectors.Result{
+			DetectorType: detectorspb.DetectorType_AirtableOAuth,
+			Raw:          []byte(match),
+		}
+
+		if verify {
+			client := s.client
+			if client == nil {
+				client = defaultClient
+			}
+
+			isVerified, extraData, verificationErr := verifyMatch(ctx, client, match)
+			s1.Verified = isVerified
+			s1.ExtraData = extraData
+			s1.SetVerificationError(verificationErr, match)
+		}
+
+		results = append(results, s1)
+	}
+
+	return
+}
+
+func verifyMatch(ctx context.Context, client *http.Client, token string) (bool, map[string]string, error) {
+	endpoint := "https://api.airtable.com/v0/meta/whoami"
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
+	if err != nil {
+		return false, nil, nil
+	}
+
+	req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", token))
+	res, err := client.Do(req)
+	if err != nil {
+		return false, nil, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	switch res.StatusCode {
+	case http.StatusOK:
+		return true, nil, nil
+	case http.StatusUnauthorized:
+		// The secret is determinately not verified (nothing to do)
+		return false, nil, nil
+	default:
+		return false, nil, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+	}
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_AirtableOAuth
+}
+
+func (s Scanner) Description() string {
+	return "Airtable is a cloud collaboration service that offers database-like features. Airtable OAuth tokens can be used to access and modify data within Airtable bases."
+}

pkg/detectors/airtableoauth/airtableoauth_integration_test.go

@@ -0,0 +1,167 @@
+//go:build detectors
+// +build detectors
+
+package airtableoauth
+
+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"
+)
+
+// TestAirtableoauth_FromChunk verifies the validity of an Airtable OAuth token
+// Note: The token validity test relies on an access token stored in the GCP secret manager.
+// Since Airtable OAuth tokens expire after 60 minutes, this test will eventually fail once the token becomes invalid.
+// The official guide linked below can be followed in order to generate a new valid access token:
+// https://airtable.com/developers/web/api/oauth-reference
+
+func TestAirtableoauth_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("AIRTABLEOAUTH")
+	inactiveSecret := testSecrets.MustGetField("AIRTABLEOAUTH_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: "found, verified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a airtableoauth secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AirtableOAuth,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a airtableoauth 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_AirtableOAuth,
+					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 airtableoauth secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AirtableOAuth,
+					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 airtableoauth secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AirtableOAuth,
+					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("Airtableoauth.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")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Airtableoauth.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/airtableoauth/airtableoauth_test.go

@@ -0,0 +1,87 @@
+package airtableoauth
+
+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 (
+	validPattern1  = "oaajtCy2lVMUN1Cm5.v1.eyJ1c2VySWQiOiJ1c3JjQ09QVlJudGlrU1lzdyIsImV4cGlyZXNBdCI6IjIwMjUtMDItMDNUMTk6NTY6MzcuMDAwWiIsIm9hdXRoQXBwbGljYXRpb25JZCI6Im9hcG14aXcyUlRrVGlzcHJIIiwic2VjcmV0IjoiMzczNThlNzdlZjlhMjljY2Q5MWIwNmNlNTdkZDYxNDg0MWVmNmIyOWYwYjQ5ZWE0MTMxZGI4NzBkNTAzYTE1NyJ9.0d67c8b334048135a93615610445e4aa90c6af6222392b49eea9419e1d6717d0"
+	validPattern2  = "oaaRYiYSlTFXZzxDM.v1.eyJ1c2VySWQiOiJ1c3JjQ09QVlJudGlrU1lzdyIsIm9hdXRoQXBwbGljYXRpb25JZCI6Im9hcG14aXcyUlRrVGlzcHJIIiwiZXhwaXJlc0F0IjoiMjAyNS0wMS0yOVQwMDowMTo0NC4wMDBaIiwic2VjcmV0IjoiZjYyOWE1MWVkM2M0ZjU5ODlmOTcyMDU1ZjkwODk3NDA4NmU0NjQxY2JhODU5Y2FhZTJkZjliMWQwODg0ZjIzMiJ9.27a8998029ac9bdd599b435572821dcb63c60cbf62b9cb2ba2a73511e5553d66"
+	invalidPattern = "oaaRYiYSlTFXZzxDM.v2.eyJ1c2VySWQiOiJ1c3JjQ09QVlJudGlrU1lzdyIsIm9hdXRoQXBwbGljYXRpb25JZCI6Im9hcG14aXcyUlRrVGlzcHJIIiwiZXhwaXJlc0F0IjoiMjAyNS0wMS0yOVQwMDowMTo0NC4wMDBaIiwic2VjcmV0IjoiZjYyOWE1MWVkM2M0ZjU5ODlmOTcyMDU1ZjkwODk3NDA4NmU0NjQxY2JhODU5Y2FhZTJkZjliMWQwODg0ZjIzMiJ9.27a8998029ac9bdd599b435572821dcb63c60cbf62b9cb2ba2a73511e5553d66"
+)
+
+func TestAirtableoauth_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name:  "typical pattern",
+			input: fmt.Sprintf("airtable token = '%s'", validPattern1),
+			want:  []string{validPattern1},
+		},
+		{
+			name: "finds all matches",
+			input: fmt.Sprintf(`airtable token 1 = '%s'
+			airtabl token 2 = '%s'`, validPattern1, validPattern2),
+			want: []string{validPattern1, validPattern2},
+		},
+		{
+			name:  "invalid pattern",
+			input: fmt.Sprintf("airtable token = '%s'", 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

@@ -14,6 +14,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/airbrakeuserkey"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/airship"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/airtableapikey"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/airtableoauth"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/airvisual"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/aiven"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/alchemy"
@@ -837,6 +838,7 @@ func buildDetectorList() []detectors.Detector {
 		&airbrakeuserkey.Scanner{},
 		&airship.Scanner{},
 		&airtableapikey.Scanner{},
+		&airtableoauth.Scanner{},
 		&airvisual.Scanner{},
 		&aiven.Scanner{},
 		&alchemy.Scanner{},

pkg/pb/detectorspb/detectors.pb.go

@@ -1118,6 +1118,7 @@ const (
 	DetectorType_TwilioApiKey                            DetectorType = 1011
 	DetectorType_Sanity                                  DetectorType = 1012
 	DetectorType_AzureRefreshToken                       DetectorType = 1013
+	DetectorType_AirtableOAuth                           DetectorType = 1014
 )
 
 // Enum value maps for DetectorType.
@@ -2133,6 +2134,7 @@ var (
 		1011: "TwilioApiKey",
 		1012: "Sanity",
 		1013: "AzureRefreshToken",
+		1014: "AirtableOAuth",
 	}
 	DetectorType_value = map[string]int32{
 		"Alibaba":                               0,
@@ -3145,6 +3147,7 @@ var (
 		"TwilioApiKey":                     1011,
 		"Sanity":                           1012,
 		"AzureRefreshToken":                1013,
+		"AirtableOAuth":                    1014,
 	}
 )
 
@@ -3598,7 +3601,7 @@ var file_detectors_proto_rawDesc = []byte{
 	0x4c, 0x41, 0x49, 0x4e, 0x10, 0x01, 0x12, 0x0a, 0x0a, 0x06, 0x42, 0x41, 0x53, 0x45, 0x36, 0x34,
 	0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x55, 0x54, 0x46, 0x31, 0x36, 0x10, 0x03, 0x12, 0x13, 0x0a,
 	0x0f, 0x45, 0x53, 0x43, 0x41, 0x50, 0x45, 0x44, 0x5f, 0x55, 0x4e, 0x49, 0x43, 0x4f, 0x44, 0x45,
-	0x10, 0x04, 0x2a, 0xd4, 0x81, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72,
+	0x10, 0x04, 0x2a, 0xe8, 0x81, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72,
 	0x54, 0x79, 0x70, 0x65, 0x12, 0x0b, 0x0a, 0x07, 0x41, 0x6c, 0x69, 0x62, 0x61, 0x62, 0x61, 0x10,
 	0x00, 0x12, 0x08, 0x0a, 0x04, 0x41, 0x4d, 0x51, 0x50, 0x10, 0x01, 0x12, 0x07, 0x0a, 0x03, 0x41,
 	0x57, 0x53, 0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x10, 0x03, 0x12,
@@ -4635,12 +4638,13 @@ var file_detectors_proto_rawDesc = []byte{
 	0xf2, 0x07, 0x12, 0x11, 0x0a, 0x0c, 0x54, 0x77, 0x69, 0x6c, 0x69, 0x6f, 0x41, 0x70, 0x69, 0x4b,
 	0x65, 0x79, 0x10, 0xf3, 0x07, 0x12, 0x0b, 0x0a, 0x06, 0x53, 0x61, 0x6e, 0x69, 0x74, 0x79, 0x10,
 	0xf4, 0x07, 0x12, 0x16, 0x0a, 0x11, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x52, 0x65, 0x66, 0x72, 0x65,
-	0x73, 0x68, 0x54, 0x6f, 0x6b, 0x65, 0x6e, 0x10, 0xf5, 0x07, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69,
+	0x73, 0x68, 0x54, 0x6f, 0x6b, 0x65, 0x6e, 0x10, 0xf5, 0x07, 0x12, 0x12, 0x0a, 0x0d, 0x41, 0x69,
-	0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65,
+	0x72, 0x74, 0x61, 0x62, 0x6c, 0x65, 0x4f, 0x41, 0x75, 0x74, 0x68, 0x10, 0xf6, 0x07, 0x42, 0x3d,
-	0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65,
+	0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74, 0x72, 0x75,
-	0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b, 0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65,
+	0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f, 0x74, 0x72, 0x75,
-	0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f,
+	0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b, 0x67, 0x2f, 0x70,
-	0x33,
+	0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62, 0x62, 0x06, 0x70,
+	0x72, 0x6f, 0x74, 0x6f, 0x33,
 }
 
 var (

proto/detectors.proto

@@ -1023,6 +1023,7 @@ enum DetectorType {
   TwilioApiKey = 1011;
   Sanity = 1012;
   AzureRefreshToken = 1013;
+  AirtableOAuth = 1014;
 }
 
 message Result {