adding pypi detector (#3287)

* adding pypi detector

* update test and use helper

---------

Co-authored-by: Dylan Ayrey <dxa4481@rit.edu>
Co-authored-by: Dustin Decker <dustin@trufflesec.com>

pkg/detectors/pypi/pypi.go

@@ -0,0 +1,125 @@
+package pypi
+
+import (
+	"bytes"
+	"context"
+	"fmt"
+	"io"
+	"mime/multipart"
+	"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()
+	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
+	keyPat = regexp.MustCompile("(pypi-AgEIcHlwaS5vcmcCJ[a-zA-Z0-9-_]{157})")
+)
+
+// 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{"pypi-AgEIcHlwaS5vcmcCJ"}
+}
+
+// FromData will find and optionally verify Pypi 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 keyPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueMatches[match[1]] = struct{}{}
+	}
+
+	for match := range uniqueMatches {
+		s1 := detectors.Result{
+			DetectorType: detectorspb.DetectorType_PyPI,
+			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) {
+	// Create a buffer to hold the multipart form data
+	var body bytes.Buffer
+	writer := multipart.NewWriter(&body)
+
+	// Add the form fields like in the curl request
+	_ = writer.WriteField(":action", "file_upload")
+	_ = writer.WriteField("name", "dummy-package")
+	_ = writer.WriteField("version", "0.0.1")
+	_ = writer.WriteField("content", "dummy-content")
+
+	// Close the writer to finalize the form
+	writer.Close()
+
+	// Create a new POST request to the PyPI legacy upload URL
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, "https://upload.pypi.org/legacy/", &body)
+	if err != nil {
+		return false, nil, err
+	}
+
+	// Add the Authorization header with the PyPI API token
+	req.Header.Add("Authorization", "token "+token)
+	// Set the Content-Type to the multipart form boundary
+	req.Header.Set("Content-Type", writer.FormDataContentType())
+
+	// Execute the HTTP request
+	res, err := client.Do(req)
+	if err != nil {
+		return false, nil, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	// Check for expected status codes for verification
+	if res.StatusCode == http.StatusBadRequest {
+		verified, err := common.ResponseContainsSubstring(res.Body, "Include at least one message digest.")
+		if err != nil {
+			return false, nil, err
+		}
+		if verified {
+			return true, nil, nil
+		}
+	} else if res.StatusCode == http.StatusForbidden {
+		// If we get a 403 status, the key is invalid
+		return false, nil, nil
+	}
+
+	// For all other status codes, return an error
+	return false, nil, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_PyPI
+}

pkg/detectors/pypi/pypi_test.go

@@ -0,0 +1,220 @@
+//go:build detectors
+// +build detectors
+
+package pypi
+
+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 TestPypi_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name:  "typical pattern",
+			input: "pypi_token = 'pypi-AgEIcHlwaS5vcmcCJDQyM2M0Yjg4LWUyNDnnnnhhMy1hNigyLWI2ZWUyMTMwYzI2MgACKlszLCJhOWQwMWE0MS01Nzk4LTQyOWYtOTk4MS1lYzE5NTJhM2E3YzgiXQAABiBeGtDnnnnnV32VpiyeU-YUDKplSv0E5ngmwsnHaV2jGg'",
+			want:  []string{"pypi-AgEIcHlwaS5vcmcCJDQyM2M0Yjg4LWUyNDnnnnhhMy1hNigyLWI2ZWUyMTMwYzI2MgACKlszLCJhOWQwMWE0MS01Nzk4LTQyOWYtOTk4MS1lYzE5NTJhM2E3YzgiXQAABiBeGtDnnnnnV32VpiyeU-YUDKplSv0E5ngmwsnHaV2jGg"},
+		},
+	}
+
+	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)
+			}
+		})
+	}
+}
+
+func TestPypi_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("PYPI")
+	inactiveSecret := testSecrets.MustGetField("PYPI_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 pypi secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_PyPI,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a pypi 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_PyPI,
+					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 pypi secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_PyPI,
+					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 pypi secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_PyPI,
+					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("Pypi.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("Pypi.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

@@ -551,6 +551,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/purestake"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/pushbulletapikey"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/pusherchannelkey"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/pypi"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/qase"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/qualaroo"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/qubole"
@@ -806,6 +807,7 @@ import (
 func DefaultDetectors() []detectors.Detector {
 	return []detectors.Detector{
 		&heroku.Scanner{},
+		&pypi.Scanner{},
 		&linearapi.Scanner{},
 		&alibaba.Scanner{},
 		aws.New(),

pkg/pb/detectorspb/detectors.pb.go

@@ -1099,6 +1099,7 @@ const (
 	DetectorType_Netsuite                                DetectorType = 995
 	DetectorType_RobinhoodCrypto                         DetectorType = 996
 	DetectorType_NVAPI                                   DetectorType = 997
+	DetectorType_PyPI                                    DetectorType = 998
 )
 
 // Enum value maps for DetectorType.
@@ -2098,6 +2099,7 @@ var (
 		995: "Netsuite",
 		996: "RobinhoodCrypto",
 		997: "NVAPI",
+		998: "PyPI",
 	}
 	DetectorType_value = map[string]int32{
 		"Alibaba":                               0,
@@ -3094,6 +3096,7 @@ var (
 		"Netsuite":                         995,
 		"RobinhoodCrypto":                  996,
 		"NVAPI":                            997,
+		"PyPI":                             998,
 	}
 )
 
@@ -3547,7 +3550,7 @@ 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, 0xba, 0x7f, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x54,
+	0x10, 0x04, 0x2a, 0xc5, 0x7f, 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, 0x0a,
@@ -4566,12 +4569,13 @@ var file_detectors_proto_rawDesc = []byte{
 	0x0a, 0x45, 0x6c, 0x65, 0x76, 0x65, 0x6e, 0x4c, 0x61, 0x62, 0x73, 0x10, 0xe2, 0x07, 0x12, 0x0d,
 	0x0a, 0x08, 0x4e, 0x65, 0x74, 0x73, 0x75, 0x69, 0x74, 0x65, 0x10, 0xe3, 0x07, 0x12, 0x14, 0x0a,
 	0x0f, 0x52, 0x6f, 0x62, 0x69, 0x6e, 0x68, 0x6f, 0x6f, 0x64, 0x43, 0x72, 0x79, 0x70, 0x74, 0x6f,
-	0x10, 0xe4, 0x07, 0x12, 0x0a, 0x0a, 0x05, 0x4e, 0x56, 0x41, 0x50, 0x49, 0x10, 0xe5, 0x07, 0x42,
+	0x10, 0xe4, 0x07, 0x12, 0x0a, 0x0a, 0x05, 0x4e, 0x56, 0x41, 0x50, 0x49, 0x10, 0xe5, 0x07, 0x12,
-	0x3d, 0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74, 0x72,
+	0x09, 0x0a, 0x04, 0x50, 0x79, 0x50, 0x49, 0x10, 0xe6, 0x07, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69,
-	0x75, 0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f, 0x74, 0x72,
+	0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65,
-	0x75, 0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b, 0x67, 0x2f,
+	0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65,
-	0x70, 0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62, 0x62, 0x06,
+	0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b, 0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65,
-	0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+	0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f,
+	0x33,
 }
 
 var (

proto/detectors.proto

@@ -1007,6 +1007,7 @@ enum DetectorType {
   Netsuite = 995;
   RobinhoodCrypto = 996;
   NVAPI = 997;
+  PyPI = 998;
 }
 
 message Result {