Merge branch 'main' into tests/pattern-test-cases-c

pkg/detectors/falsepositives.go

@@ -13,7 +13,12 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 )
 
-var DefaultFalsePositives = []FalsePositive{"example", "xxxxxx", "aaaaaa", "abcde", "00000", "sample", "*****"}
+var (
+	DefaultFalsePositives = map[FalsePositive]struct{}{
+		"example": {}, "xxxxxx": {}, "aaaaaa": {}, "abcde": {}, "00000": {}, "sample": {}, "*****": {},
+	}
+	UuidFalsePositives map[FalsePositive]struct{}
+)
 
 type FalsePositive string
 
@@ -24,18 +29,21 @@ type CustomFalsePositiveChecker interface {
 	IsFalsePositive(result Result) (bool, string)
 }
 
-//go:embed "badlist.txt"
-var badList []byte
-
-//go:embed "words.txt"
-var wordList []byte
-
-//go:embed "programmingbooks.txt"
-var programmingBookWords []byte
-
-var filter *ahocorasick.Trie
+var (
+	filter *ahocorasick.Trie
+
+	//go:embed "fp_badlist.txt"
+	badList []byte
+	//go:embed "fp_words.txt"
+	wordList []byte
+	//go:embed "fp_programmingbooks.txt"
+	programmingBookWords []byte
+	//go:embed "fp_uuids.txt"
+	uuidList []byte
+)
 
 func init() {
+	// Populate trie.
 	builder := ahocorasick.NewTrieBuilder()
 
 	wordList := bytesToCleanWordList(wordList)
@@ -47,7 +55,16 @@ func init() {
 	programmingBookWords := bytesToCleanWordList(programmingBookWords)
 	builder.AddStrings(programmingBookWords)
 
+	uuidList := bytesToCleanWordList(uuidList)
+	builder.AddStrings(uuidList)
+
 	filter = builder.Build()
+
+	// Populate custom FalsePositive list
+	UuidFalsePositives = make(map[FalsePositive]struct{}, len(uuidList))
+	for _, uuid := range uuidList {
+		UuidFalsePositives[FalsePositive(uuid)] = struct{}{}
+	}
 }
 
 func GetFalsePositiveCheck(detector Detector) func(Result) (bool, string) {
@@ -65,15 +82,20 @@ func GetFalsePositiveCheck(detector Detector) func(Result) (bool, string) {
 //
 // Currently, this includes: english word in key or matches common example patterns.
 // Only the secret key material should be passed into this function
-func IsKnownFalsePositive(match string, falsePositives []FalsePositive, wordCheck bool) (bool, string) {
+func IsKnownFalsePositive(match string, falsePositives map[FalsePositive]struct{}, wordCheck bool) (bool, string) {
 	if !utf8.ValidString(match) {
 		return true, "invalid utf8"
 	}
 	lower := strings.ToLower(match)
-	for _, fp := range falsePositives {
+
+	if _, exists := falsePositives[FalsePositive(lower)]; exists {
+		return true, "matches term: " + lower
+	}
+
+	for fp := range falsePositives {
 		fps := string(fp)
 		if strings.Contains(lower, fps) {
-			return true, "matches term: " + fps
+			return true, "contains term: " + fps
 		}
 	}
 

pkg/detectors/falsepositives_test.go

@@ -1,6 +1,3 @@
-//go:build detectors
-// +build detectors
-
 package detectors
 
 import (
@@ -32,17 +29,23 @@ func (d fakeDetector) Type() detectorspb.DetectorType {
 func (f fakeDetector) Description() string { return "" }
 
 func (d customFalsePositiveChecker) IsFalsePositive(result Result) (bool, string) {
-	return IsKnownFalsePositive(string(result.Raw), []FalsePositive{"a specific magic string"}, false)
+	return IsKnownFalsePositive(string(result.Raw), map[FalsePositive]struct{}{"a specific magic string": {}}, false)
 }
 
 func TestFilterKnownFalsePositives_DefaultLogic(t *testing.T) {
 	results := []Result{
-		{Raw: []byte("00000")},           // "default" false positive list
-		{Raw: []byte("number")},          // from wordlist
-		{Raw: []byte("hga8adshla3434g")}, // real secret
+		{Raw: []byte("00000")},  // "default" false positive list
+		{Raw: []byte("number")}, // from wordlist
+		// from uuid list
+		{Raw: []byte("00000000-0000-0000-0000-000000000000")},
+		{Raw: []byte("xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx")},
+		// real secrets
+		{Raw: []byte("hga8adshla3434g")},
+		{Raw: []byte("f795f7db-2dfe-4095-96f3-8f8370c735f9")},
 	}
 	expected := []Result{
 		{Raw: []byte("hga8adshla3434g")},
+		{Raw: []byte("f795f7db-2dfe-4095-96f3-8f8370c735f9")},
 	}
 	filtered := FilterKnownFalsePositives(logContext.Background(), fakeDetector{}, results)
 	assert.ElementsMatch(t, expected, filtered)
@@ -67,7 +70,7 @@ func TestFilterKnownFalsePositives_CustomLogic(t *testing.T) {
 func TestIsFalsePositive(t *testing.T) {
 	type args struct {
 		match          string
-		falsePositives []FalsePositive
+		falsePositives map[FalsePositive]struct{}
 		useWordlist    bool
 	}
 	tests := []struct {

pkg/detectors/fetchrss/fetchrss.go

@@ -2,26 +2,30 @@ package fetchrss
 
 import (
 	"context"
-	regexp "github.com/wasilibs/go-re2"
+	"encoding/json"
+	"fmt"
 	"io"
 	"net/http"
-	"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{}
+type Scanner struct {
+	client *http.Client
+}
 
 // Ensure the Scanner satisfies the interface at compile time.
 var _ detectors.Detector = (*Scanner)(nil)
 
 var (
-	client = common.SaneHttpClient()
+	defaultClient = common.SaneHttpClient()
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
-	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"fetchrss"}) + `\b([0-9A-Za-z.]{40})\b`)
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"fetchrss"}) + `\b([a-zA-Z0-9.]{40})\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.
@@ -34,37 +38,26 @@ func (s Scanner) Keywords() []string {
 func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (results []detectors.Result, err error) {
 	dataStr := string(data)
 
-	matches := keyPat.FindAllStringSubmatch(dataStr, -1)
-
-	for _, match := range matches {
-		if len(match) != 2 {
-			continue
-		}
-		resMatch := strings.TrimSpace(match[1])
+	uniqueMatches := make(map[string]struct{})
+	for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueMatches[match[1]] = struct{}{}
+	}
 
+	for token := range uniqueMatches {
 		s1 := detectors.Result{
 			DetectorType: detectorspb.DetectorType_Fetchrss,
-			Raw:          []byte(resMatch),
+			Raw:          []byte(token),
 		}
 
 		if verify {
-			req, err := http.NewRequestWithContext(ctx, "GET", "https://fetchrss.com/api/v1/feed/list?auth="+resMatch, nil)
-			if err != nil {
-				continue
-			}
-			res, err := client.Do(req)
-			if err == nil {
-				defer res.Body.Close()
-				bodyBytes, err := io.ReadAll(res.Body)
-				if err != nil {
-					continue
-				}
-				body := string(bodyBytes)
-
-				if !strings.Contains(body, "Not authorised") {
-					s1.Verified = true
-				}
+			client := s.client
+			if client == nil {
+				client = defaultClient
 			}
+
+			verified, verificationErr := verifyToken(ctx, client, token)
+			s1.Verified = verified
+			s1.SetVerificationError(verificationErr)
 		}
 
 		results = append(results, s1)
@@ -73,6 +66,51 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 	return results, nil
 }
 
+func verifyToken(ctx context.Context, client *http.Client, token string) (bool, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://fetchrss.com/api/v1/feed/list?auth="+token, nil)
+	if err != nil {
+		return false, err
+	}
+
+	res, err := client.Do(req)
+	if err != nil {
+		return false, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	// The API seems to always return a 200 status code.
+	// See: https://fetchrss.com/developers
+	if res.StatusCode != http.StatusOK {
+		return false, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+	}
+
+	var apiRes response
+	if err := json.NewDecoder(res.Body).Decode(&apiRes); err != nil {
+		return false, err
+	}
+
+	if apiRes.Success {
+		// The key is valid.
+		return true, nil
+	} else if apiRes.Error.Code == 401 {
+		// The key is invalid.
+		return false, nil
+	} else {
+		return false, fmt.Errorf("unexpected error: [code=%d, message=%s]", apiRes.Error.Code, apiRes.Error.Message)
+	}
+}
+
+type response struct {
+	Success bool `json:"success"`
+	Error   struct {
+		Message string `json:"message"`
+		Code    int    `json:"code"`
+	} `json:"error"`
+}
+
 func (s Scanner) Type() detectorspb.DetectorType {
 	return detectorspb.DetectorType_Fetchrss
 }

pkg/detectors/fp_uuids.txt

@@ -0,0 +1,37 @@
+00000000-0000-0000-0000-000000000000
+11111111-1111-1111-1111-111111111111
+22222222-2222-2222-2222-222222222222
+33333333-3333-3333-3333-333333333333
+44444444-4444-4444-4444-444444444444
+55555555-5555-5555-5555-555555555555
+66666666-6666-6666-6666-666666666666
+77777777-7777-7777-7777-777777777777
+88888888-8888-8888-8888-888888888888
+99999999-9999-9999-9999-999999999999
+12345678-1234-1234-1234-123456789abc
+23456789-2345-2345-2345-23456789abcd
+34567890-3456-3456-3456-34567890bcde
+45678901-4567-4567-4567-45678901cdef
+56789012-5678-5678-5678-56789012def0
+aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa
+bbbbbbbb-bbbb-bbbb-bbbb-bbbbbbbbbbbb
+cccccccc-cccc-cccc-cccc-cccccccccccc
+dddddddd-dddd-dddd-dddd-dddddddddddd
+eeeeeeee-eeee-eeee-eeee-eeeeeeeeeeee
+ffffffff-ffff-ffff-ffff-ffffffffffff
+deadbeef-dead-beef-dead-beefdeadbeef
+cafebabe-cafe-babe-cafe-babecafebabe
+badc0ffee-badc-0ffe-badc-0ffeebadc0f
+deadface-dead-face-dead-facedeadface
+feedface-feed-face-feed-facefeedface
+a1b2c3d4-a1b2-c3d4-a1b2-c3d4a1b2c3d4
+98765432-9876-5432-9876-543298765432
+abcdefab-cdef-abcd-efab-cdefabcdefab
+a0a0a0a0-a0a0-a0a0-a0a0-a0a0a0a0a0a0
+b0b0b0b0-b0b0-b0b0-b0b0-b0b0b0b0b0b0
+c0c0c0c0-c0c0-c0c0-c0c0-c0c0c0c0c0c0
+d0d0d0d0-d0d0-d0d0-d0d0-d0d0d0d0d0d0
+e0e0e0e0-e0e0-e0e0-e0e0-e0e0e0e0e0e0
+f0f0f0f0-f0f0-f0f0-f0f0-f0f0f0f0f0f0
+xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
+-xxxx-xxxx-xxxx-xxxxxxxxxxxx

pkg/detectors/ftp/ftp.go

@@ -103,8 +103,12 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 	return results, nil
 }
 
+var ftpFalsePositives = map[detectors.FalsePositive]struct{}{
+	detectors.FalsePositive("@ftp.freebsd.org"): {},
+}
+
 func (s Scanner) IsFalsePositive(result detectors.Result) (bool, string) {
-	return detectors.IsKnownFalsePositive(string(result.Raw), []detectors.FalsePositive{"@ftp.freebsd.org"}, false)
+	return detectors.IsKnownFalsePositive(string(result.Raw), ftpFalsePositives, false)
 }
 
 func isErrDeterminate(e error) bool {

pkg/detectors/github/v1/github_old.go

@@ -58,6 +58,10 @@ func (s Scanner) Keywords() []string {
 	return []string{"github", "gh", "pat", "token"}
 }
 
+var ghFalsePositives = map[detectors.FalsePositive]struct{}{
+	detectors.FalsePositive("github commit"): {},
+}
+
 // FromData will find and optionally verify GitHub 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)
@@ -74,8 +78,7 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 
 		// Note that this false positive check happens **before** verification! I don't know why it's written this way
 		// but that's why this logic wasn't moved into a CustomFalsePositiveChecker implementation.
-		specificFPs := []detectors.FalsePositive{"github commit"}
-		if isFp, _ := detectors.IsKnownFalsePositive(token, specificFPs, false); isFp {
+		if isFp, _ := detectors.IsKnownFalsePositive(token, ghFalsePositives, false); isFp {
 			continue
 		}