Merge branch 'main' into tri-state-mongodb

go.mod

@@ -34,6 +34,7 @@ require (
 	github.com/go-sql-driver/mysql v1.7.1
 	github.com/gobwas/glob v0.2.3
 	github.com/golang-jwt/jwt v3.2.2+incompatible
+	github.com/golang/mock v1.4.3
 	github.com/google/go-cmp v0.5.9
 	github.com/google/go-containerregistry v0.15.2
 	github.com/google/go-github/v42 v42.0.0
@@ -56,6 +57,7 @@ require (
 	github.com/tailscale/depaware v0.0.0-20210622194025-720c4b409502
 	github.com/xanzy/go-gitlab v0.88.0
 	go.mongodb.org/mongo-driver v1.12.0
+	go.uber.org/mock v0.2.0
 	go.uber.org/zap v1.24.0
 	golang.org/x/crypto v0.11.0
 	golang.org/x/exp v0.0.0-20221018205818-5c77f4b2bbd7

go.sum

@@ -486,6 +486,8 @@ go.uber.org/atomic v1.7.0/go.mod h1:fEN4uk6kAWBTFdckzkM89CLk9XfWZrxpCo0nPH17wJc=
 go.uber.org/goleak v1.1.11/go.mod h1:cwTWslyiVhfpKIDGSZEM2HlOvcqm+tG4zioyIeLoqMQ=
 go.uber.org/goleak v1.2.1 h1:NBol2c7O1ZokfZ0LEU9K6Whx/KnwvepVetCUhtKja4A=
 go.uber.org/goleak v1.2.1/go.mod h1:qlT2yGI9QafXHhZZLxlSuNsMw3FFLxBr+tBRlmO1xH4=
+go.uber.org/mock v0.2.0 h1:TaP3xedm7JaAgScZO7tlvlKrqT0p7I6OsdGB5YNSMDU=
+go.uber.org/mock v0.2.0/go.mod h1:J0y0rp9L3xiff1+ZBfKxlC1fz2+aO16tw0tsDOixfuM=
 go.uber.org/multierr v1.6.0 h1:y6IPFStTAIT5Ytl7/XYmHvzXQ7S3g/IeZW9hyZ5thw4=
 go.uber.org/multierr v1.6.0/go.mod h1:cdWPpRnG4AhwMwsgIHip0KRBQjJy5kYEpYjJxpXp9iU=
 go.uber.org/zap v1.24.0 h1:FiJd5l1UOLj0wCgbSE0rwwXHzEdAZS6hiiSnxJN/D60=

pkg/common/http.go

@@ -100,14 +100,14 @@ func NewCustomTransport(T http.RoundTripper) *CustomTransport {
 	return &CustomTransport{T}
 }
 
-func ConstantStatusHttpClient(statusCode int) *http.Client {
+func ConstantResponseHttpClient(statusCode int, body string) *http.Client {
 	return &http.Client{
 		Timeout: DefaultResponseTimeout,
 		Transport: FakeTransport{
 			CreateResponse: func(req *http.Request) (*http.Response, error) {
 				return &http.Response{
 					Request:    req,
-					Body:       io.NopCloser(strings.NewReader("")),
+					Body:       io.NopCloser(strings.NewReader(body)),
 					StatusCode: statusCode,
 				}, nil
 			},

pkg/context/context.go

@@ -3,18 +3,26 @@ package context
 import (
 	"context"
 	"fmt"
+	"os"
 	"runtime/debug"
 	"sync"
 	"time"
 
 	"github.com/go-logr/logr"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/log"
 )
 
 var (
 	// defaultLogger can be set via SetDefaultLogger.
-	defaultLogger logr.Logger = logr.Discard()
+	// It is initialized to write to stderr. To disable, you can call
+	// SetDefaultLogger with logr.Discard().
+	defaultLogger logr.Logger
 )
 
+func init() {
+	defaultLogger, _ = log.New("context", log.WithConsoleSink(os.Stderr))
+}
+
 // Context wraps context.Context and includes an additional Logger() method.
 type Context interface {
 	context.Context
@@ -136,7 +144,9 @@ func AddLogger(parent context.Context) Context {
 }
 
 // SetupDefaultLogger sets the package-level global default logger that will be
-// used for Background and TODO contexts.
+// used for Background and TODO contexts. On startup, the default logger will
+// be configured to output logs to stderr. Use logr.Discard() to disable all
+// logs from Contexts.
 func SetDefaultLogger(l logr.Logger) {
 	defaultLogger = l
 }

pkg/context/context_test.go

@@ -158,7 +158,7 @@ func TestWithValues(t *testing.T) {
 	assert.NotContains(t, logs[6], `what does this do?`)
 }
 
-func TestDiscardLogger(t *testing.T) {
+func TestDefaultLogger(t *testing.T) {
 	var panicked bool
 	defer func() {
 		if r := recover(); r != nil {

pkg/detectors/alchemy/alchemy_test.go

@@ -104,7 +104,7 @@ func TestAlchemy_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, verified but unexpected api surface",
-			s:    Scanner{client: common.ConstantStatusHttpClient(404)},
+			s:    Scanner{client: common.ConstantResponseHttpClient(404, "")},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a alchemy secret %s within", secret)),

pkg/detectors/aws/aws.go

@@ -18,7 +18,8 @@ import (
 )
 
 type scanner struct {
-	skipIDs map[string]struct{}
+	verificationClient *http.Client
+	skipIDs            map[string]struct{}
 }
 
 func New(opts ...func(*scanner)) *scanner {
@@ -48,7 +49,7 @@ func WithSkipIDs(skipIDs []string) func(*scanner) {
 var _ detectors.Detector = (*scanner)(nil)
 
 var (
-	client = common.SaneHttpClient()
+	defaultVerificationClient = common.SaneHttpClient()
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
 	// Key types are from this list https://docs.aws.amazon.com/IAM/latest/UserGuide/reference_identifiers.html#identifiers-unique-ids
@@ -123,8 +124,9 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 				host := "sts.amazonaws.com"
 				region := "us-east-1"
 				endpoint := "https://sts.amazonaws.com"
-				datestamp := time.Now().UTC().Format("20060102")
-				amzDate := time.Now().UTC().Format("20060102T150405Z0700")
+				now := time.Now().UTC()
+				datestamp := now.Format("20060102")
+				amzDate := now.Format("20060102T150405Z0700")
 
 				req, err := http.NewRequestWithContext(ctx, method, endpoint, nil)
 				if err != nil {
@@ -171,6 +173,10 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 				req.Header.Add("Content-type", "application/x-www-form-urlencoded; charset=utf-8")
 				req.URL.RawQuery = params.Encode()
 
+				client := s.verificationClient
+				if client == nil {
+					client = defaultVerificationClient
+				}
 				res, err := client.Do(req)
 				if err == nil {
 
@@ -194,7 +200,25 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 							continue
 						}
 
-						if res.StatusCode != 403 {
+						if res.StatusCode == 403 {
+							var body awsErrorResponseBody
+							err = json.NewDecoder(res.Body).Decode(&body)
+							res.Body.Close()
+							if err == nil {
+								// All instances of the code I've seen in the wild are PascalCased but this check is
+								// case-insensitive out of an abundance of caution
+								if strings.EqualFold(body.Error.Code, "InvalidClientTokenId") {
+									// determinate failure - nothing to do
+								} else {
+									// We see a surprising number of false-negative signature mismatch errors here
+									// (The official SDK somehow elicits even more than just making the request
+									// ourselves)
+									s1.VerificationError = fmt.Errorf("request to %v returned status %d with an unexpected reason (%s: %s)", res.Request.URL, res.StatusCode, body.Error.Code, body.Error.Message)
+								}
+							} else {
+								s1.VerificationError = fmt.Errorf("couldn't parse the sts response body (%v)", err)
+							}
+						} else {
 							s1.VerificationError = fmt.Errorf("request to %v returned unexpected status %d", res.Request.URL, res.StatusCode)
 						}
 					}
@@ -245,6 +269,15 @@ func awsCustomCleanResults(results []detectors.Result) []detectors.Result {
 	return out
 }
 
+type awsError struct {
+	Code    string `json:"Code"`
+	Message string `json:"Message"`
+}
+
+type awsErrorResponseBody struct {
+	Error awsError `json:"Error"`
+}
+
 type identityRes struct {
 	GetCallerIdentityResponse struct {
 		GetCallerIdentityResult struct {

pkg/detectors/aws/aws_test.go

@@ -17,6 +17,8 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 )
 
+var unverifiedSecretClient = common.ConstantResponseHttpClient(403, `{"Error": {"Code": "InvalidClientTokenId"} }`)
+
 func TestAWS_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
@@ -69,7 +71,7 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, unverified",
-			s:    scanner{},
+			s:    scanner{verificationClient: unverifiedSecretClient},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s but not valid", inactiveSecret, id)), // the secret would satisfy the regex but not pass validation
@@ -163,7 +165,7 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, unverified, with leading +",
-			s:    scanner{},
+			s:    scanner{verificationClient: unverifiedSecretClient},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s but not valid", "+HaNv9cTwheDKGJaws/+BMF2GgybQgBWdhcOOdfF", id)), // the secret would satisfy the regex but not pass validation
@@ -194,9 +196,45 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, would be verified if not for http timeout",
-			s:    scanner{},
+			s:    scanner{verificationClient: common.SaneHttpClientTimeOut(1 * time.Microsecond)},
 			args: args{
-				ctx:    timeoutContext(1 * time.Microsecond),
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AWS,
+					Verified:     false,
+					Redacted:     "AKIASP2TPHJSQH3FJRUX",
+				},
+			},
+			wantErr:               false,
+			wantVerificationError: true,
+		},
+		{
+			name: "found, unverified due to unexpected http response status",
+			s:    scanner{verificationClient: common.ConstantResponseHttpClient(500, "internal server error")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AWS,
+					Verified:     false,
+					Redacted:     "AKIASP2TPHJSQH3FJRUX",
+				},
+			},
+			wantErr:               false,
+			wantVerificationError: true,
+		},
+		{
+			name: "found, unverified due to unexpected 403 response reason",
+			s:    scanner{verificationClient: common.ConstantResponseHttpClient(403, `{"Error": {"Code": "SignatureDoesNotMatch"} }`)},
+			args: args{
+				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
 				verify: true,
 			},
@@ -224,7 +262,7 @@ func TestAWS_FromChunk(t *testing.T) {
 					t.Fatalf("no raw secret present: \n %+v", got[i])
 				}
 				if (got[i].VerificationError != nil) != tt.wantVerificationError {
-					t.Fatalf("verification error = %v, wantVerificationError %v", got[i].VerificationError, tt.wantVerificationError)
+					t.Fatalf("wantVerificationError %v, verification error = %v", tt.wantVerificationError, got[i].VerificationError)
 				}
 			}
 			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "RawV2", "Raw", "VerificationError")
@@ -249,8 +287,3 @@ func BenchmarkFromData(benchmark *testing.B) {
 		})
 	}
 }
-
-func timeoutContext(timeout time.Duration) context.Context {
-	c, _ := context.WithTimeout(context.Background(), timeout)
-	return c
-}

pkg/detectors/pubnubpublishkey/pubnubpublishkey.go

@@ -21,7 +21,7 @@ var (
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
 	pubPat = regexp.MustCompile(`\b(pub-c-[0-9a-z]{8}-[0-9a-z]{4}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{12})\b`)
-	subPat = regexp.MustCompile(`\b(sub-c-[0-9a-z]{8}-[a-z]{4}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{12})\b`)
+	subPat = regexp.MustCompile(`\b(sub-c-[0-9a-z]{8}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{12})\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.

pkg/detectors/pubnubsubscriptionkey/pubnubsubscriptionkey.go

@@ -20,7 +20,7 @@ var (
 	client = common.SaneHttpClient()
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
-	keyPat = regexp.MustCompile(`\b(sub-c-[0-9a-z]{8}-[a-z]{4}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{12})\b`)
+	keyPat = regexp.MustCompile(`\b(sub-c-[0-9a-z]{8}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{4}-[a-z0-9]{12})\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.

pkg/engine/engine.go

@@ -126,7 +126,7 @@ func Start(ctx context.Context, options ...EngineOption) *Engine {
 		ctx.Logger().Info("No concurrency specified, defaulting to max", "cpu", numCPU)
 		e.concurrency = numCPU
 	}
-	ctx.Logger().V(2).Info("engine started", "workers", e.concurrency)
+	ctx.Logger().V(3).Info("engine started", "workers", e.concurrency)
 
 	// Limit number of concurrent goroutines dedicated to chunking a source.
 	e.sourcesWg.SetLimit(e.concurrency)
@@ -156,8 +156,8 @@ func Start(ctx context.Context, options ...EngineOption) *Engine {
 	}
 	e.prefilter = *ahocorasick.NewTrieBuilder().AddStrings(keywords).Build()
 
-	ctx.Logger().Info("loaded decoders", "count", len(e.decoders))
-	ctx.Logger().Info("loaded detectors",
+	ctx.Logger().V(3).Info("loaded decoders", "count", len(e.decoders))
+	ctx.Logger().V(3).Info("loaded detectors",
 		"total", len(e.detectors[true])+len(e.detectors[false]),
 		"verification_enabled", len(e.detectors[true]),
 		"verification_disabled", len(e.detectors[false]),