From 702f879db4106bf89408787286e24447791985ba Mon Sep 17 00:00:00 2001
From: Richard Gomez <rmgomez368@gmail.com>
Date: Fri, 1 Sep 2023 00:12:39 -0400
Subject: [PATCH] feat: optional detectors

---
 pkg/detectors/detectors.go       |  5 +++++
 pkg/detectors/parseur/parseur.go | 33 ++++++++++++++++++++++++++++++++
 pkg/engine/engine.go             |  5 +++++
 3 files changed, 43 insertions(+)

diff --git a/pkg/detectors/detectors.go b/pkg/detectors/detectors.go
index 0a2356c0e791..13aa14b0d4f9 100644
--- a/pkg/detectors/detectors.go
+++ b/pkg/detectors/detectors.go
@@ -25,6 +25,11 @@ type Detector interface {
 	Type() detectorspb.DetectorType
 }
 
+type OptionalDetector interface {
+	// ShouldScanChunk determines whether the detector should run.
+	ShouldScanChunk(chunk sources.Chunk) bool
+}
+
 // Versioner is an optional interface that a detector can implement to
 // differentiate instances of the same detector type.
 type Versioner interface {
diff --git a/pkg/detectors/parseur/parseur.go b/pkg/detectors/parseur/parseur.go
index 4cad1da9dfcd..1007e26b41c4 100644
--- a/pkg/detectors/parseur/parseur.go
+++ b/pkg/detectors/parseur/parseur.go
@@ -3,6 +3,8 @@ package parseur
 import (
 	"context"
 	"fmt"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 	"net/http"
 	"regexp"
 	"strings"
@@ -20,6 +22,10 @@ var _ detectors.Detector = (*Scanner)(nil)
 var (
 	client = common.SaneHttpClient()
 
+	// Node.js lock files are a common source of false-positives.
+	// https://github.com/trufflesecurity/trufflehog/issues/1460
+	ignoreFilesPat = regexp.MustCompile(`(^|/)(package(-lock)?\.json|yarn\.lock)$`)
+
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives
 	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"parseur"}) + `\b([a-f0-9]{40})\b`)
 )
@@ -30,6 +36,33 @@ func (s Scanner) Keywords() []string {
 	return []string{"parseur"}
 }
 
+func (s Scanner) ShouldScanChunk(chunk sources.Chunk) bool {
+	m := chunk.SourceMetadata
+	var file string
+	switch chunk.SourceType {
+	case sourcespb.SourceType_SOURCE_TYPE_BITBUCKET:
+		file = m.GetBitbucket().File
+	case sourcespb.SourceType_SOURCE_TYPE_DOCKER:
+		file = m.GetDocker().File
+	case sourcespb.SourceType_SOURCE_TYPE_GITHUB:
+		file = m.GetGithub().File
+	case sourcespb.SourceType_SOURCE_TYPE_PUBLIC_GIT:
+		// Here be dragons
+		//file = m.Get
+		return true
+	case sourcespb.SourceType_SOURCE_TYPE_GITLAB:
+		file = m.GetGitlab().File
+	case sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM:
+		file = m.GetFilesystem().File
+	case sourcespb.SourceType_SOURCE_TYPE_GIT:
+		file = m.GetGit().File
+	default:
+		return true
+	}
+
+	return ignoreFilesPat.MatchString(file)
+}
+
 // FromData will find and optionally verify Parseur 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)
diff --git a/pkg/engine/engine.go b/pkg/engine/engine.go
index 3489d980d375..0eeefd0722d9 100644
--- a/pkg/engine/engine.go
+++ b/pkg/engine/engine.go
@@ -464,6 +464,11 @@ func (e *Engine) detectorWorker(ctx context.Context) {
 
 				for verify, detectorsSet := range e.detectors {
 					for _, detector := range detectorsSet {
+						d, ok := detector.(detectors.OptionalDetector)
+						if ok && !d.ShouldScanChunk(*chunk) {
+							continue
+						}
+
 						chunkContainsKeyword := false
 						for _, kw := range detector.Keywords() {
 							if _, ok := matchedKeywords[strings.ToLower(kw)]; ok {