Add Wiz detector (#2691)

* Implement wiz detector

* Fix tests

* Update false positive logic

pkg/detectors/wiz/wiz.go

@@ -0,0 +1,130 @@
+package wiz
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+
+	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.
+	idPat     = regexp.MustCompile(detectors.PrefixRegex([]string{"wiz"}) + `\b([a-zA-Z0-9]{53})\b`)
+	secretPat = regexp.MustCompile(detectors.PrefixRegex([]string{"wiz"}) + `\b([a-zA-Z0-9]{64})\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{"wiz"}
+}
+
+// FromData will find and optionally verify Wiz 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)
+
+	idMatches := make(map[string]struct{})
+	for _, match := range idPat.FindAllStringSubmatch(dataStr, -1) {
+		idMatches[match[1]] = struct{}{}
+	}
+
+	secretMatches := make(map[string]struct{})
+	for _, match := range secretPat.FindAllStringSubmatch(dataStr, -1) {
+		secretMatches[match[1]] = struct{}{}
+	}
+
+	for idMatch := range idMatches {
+		for secretMatch := range secretMatches {
+			s1 := detectors.Result{
+				DetectorType: detectorspb.DetectorType_Wiz,
+				Raw:          []byte(idMatch),
+				RawV2:        []byte(idMatch + secretMatch),
+			}
+
+			if verify {
+				client := s.client
+				if client == nil {
+					client = defaultClient
+				}
+
+				isVerified, extraData, verificationErr := verifyMatch(ctx, client, idMatch, secretMatch)
+				s1.Verified = isVerified
+				s1.ExtraData = extraData
+				s1.SetVerificationError(verificationErr, idMatch, secretMatch)
+			}
+
+			// This function will check false positives for common test words, but also it will make sure the key appears 'random' enough to be a real key.
+			if !s1.Verified {
+				if detectors.IsKnownFalsePositive(idMatch, detectors.DefaultFalsePositives, true) ||
+					detectors.IsKnownFalsePositive(secretMatch, detectors.DefaultFalsePositives, true) {
+					continue
+				}
+			}
+
+			results = append(results, s1)
+
+			// If we've found a verified match with this ID, we don't need to look for anymore. So move on to the next ID.
+			if s1.Verified {
+				break
+			}
+		}
+	}
+
+	return
+}
+
+func verifyMatch(ctx context.Context, client *http.Client, clientID, clientSecret string) (bool, map[string]string, error) {
+	authData := url.Values{}
+	authData.Set("grant_type", "client_credentials")
+	authData.Set("audience", "wiz-api")
+	authData.Set("client_id", clientID)
+	authData.Set("client_secret", clientSecret)
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, "https://auth.app.wiz.io/oauth/token", strings.NewReader(authData.Encode()))
+	if err != nil {
+		return false, nil, nil
+	}
+
+	req.Header.Add("Encoding", "UTF-8")
+	req.Header.Add("Content-Type", "application/x-www-form-urlencoded")
+
+	res, err := client.Do(req)
+	if err != nil {
+		return false, nil, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	if res.StatusCode >= 200 && res.StatusCode < 300 {
+		// If the endpoint returns useful information, we can return it as a map.
+		return true, nil, nil
+	} else if res.StatusCode == 401 {
+		// The secret is determinately not verified (nothing to do)
+		return false, nil, nil
+	} else {
+		err = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+		return false, nil, err
+	}
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_Wiz
+}

pkg/detectors/wiz/wiz_test.go

@@ -0,0 +1,227 @@
+//go:build detectors
+// +build detectors
+
+package wiz
+
+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/engine/ahocorasick"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestWiz_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name: "typical pattern",
+			input: `
+wiz_client_id = 'lmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDblmSlx1fe6yCfwAbDa8pMp'
+wiz_client_secret = 'lmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDblmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDb'
+`,
+			want: []string{
+				"lmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDblmSlx1fe6yCfwAbDa8pMp" +
+					"lmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDblmSlx1fe6yCfwAbDa8pMp9sJDM9rZzDb",
+			},
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			chunkSpecificDetectors := make(map[ahocorasick.DetectorKey]detectors.Detector, 2)
+			ahoCorasickCore.PopulateMatchingDetectors(test.input, chunkSpecificDetectors)
+			if len(chunkSpecificDetectors) == 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)
+			}
+		})
+	}
+}
+
+func TestWiz_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("WIZ")
+	inactiveSecret := testSecrets.MustGetField("WIZ_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 wiz secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Wiz,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a wiz 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_Wiz,
+					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 wiz secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Wiz,
+					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 wiz secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Wiz,
+					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("Wiz.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", "RawV2", "verificationError")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Wiz.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/engine/defaults.go

@@ -759,6 +759,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/websitepulse"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/whoxy"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/wistia"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/wiz"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/worksnaps"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/workstack"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/worldcoinindex"
@@ -1597,6 +1598,7 @@ func DefaultDetectors() []detectors.Detector {
 		dockerhubv2.Scanner{},
 		&jupiterone.Scanner{},
 		gcpapplicationdefaultcredentials.Scanner{},
+		wiz.Scanner{},
 	}
 }