Enhanced the easyinsight detector (#3384)

* Enhanced the easyinsight detector

* restructured verification code and resolved comments

* resolved comments

* added basic auth

* updated statuscode logic

pkg/detectors/easyinsight/easyinsight.go

@@ -2,18 +2,17 @@ package easyinsight
 
 import (
 	"context"
-	b64 "encoding/base64"
 	"fmt"
 	regexp "github.com/wasilibs/go-re2"
+	"io"
 	"net/http"
-	"strings"
 
 	"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 {
 	detectors.DefaultMultiPartCredentialProvider
 }
 
@@ -24,8 +23,8 @@ var (
 	client = 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{"easyinsight", "easy-insight"}) + `\b([0-9Aa-zA-Z]{20})\b`)
-	idPat  = regexp.MustCompile(detectors.PrefixRegex([]string{"easyinsight", "easy-insight"}) + `\b([a-zA-Z0-9]{20})\b`)
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"easyinsight", "easy-insight", "key"}) + `\b([0-9a-zA-Z]{20})\b`)
+	idPat  = regexp.MustCompile(detectors.PrefixRegex([]string{"easyinsight", "easy-insight", "id"}) + `\b([a-zA-Z0-9]{20})\b`)
 )
 
 // Keywords are used for efficiently pre-filtering chunks.
@@ -38,49 +37,45 @@ 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)
+	var keyMatches, idMatches = make(map[string]struct{}), make(map[string]struct{})
 
-	idmatches := idPat.FindAllStringSubmatch(dataStr, -1)
+	// get unique key and id matches
+	for _, matches := range keyPat.FindAllStringSubmatch(dataStr, -1) {
+		keyMatches[matches[1]] = struct{}{}
+	}
 
-	for _, match := range matches {
-		if len(match) != 2 {
-			continue
-		}
-		resMatch := strings.TrimSpace(match[1])
-		for _, idmatch := range idmatches {
-			if len(idmatch) != 2 {
+	for _, matches := range idPat.FindAllStringSubmatch(dataStr, -1) {
+		idMatches[matches[1]] = struct{}{}
+	}
+
+	for keyMatch := range keyMatches {
+		for idMatch := range idMatches {
+			//as key and id regex are same, the strings captured by both regex will be same.
+			//avoid processing when key is same as id. This will allow detector to process only different combinations
+			if keyMatch == idMatch {
 				continue
 			}
-			resIdMatch := strings.TrimSpace(idmatch[1])
 
 			s1 := detectors.Result{
 				DetectorType: detectorspb.DetectorType_EasyInsight,
-				Raw:          []byte(resMatch),
-				RawV2:        []byte(resMatch + resIdMatch),
+				Raw:          []byte(keyMatch),
+				RawV2:        []byte(keyMatch + idMatch),
 			}
 
 			if verify {
-				data := fmt.Sprintf("%s:%s", resIdMatch, resMatch)
-				sEnc := b64.StdEncoding.EncodeToString([]byte(data))
-				req, err := http.NewRequestWithContext(ctx, "GET", "https://www.easy-insight.com/app/api/users.json", nil)
-				if err != nil {
-					continue
-				}
-				req.Header.Add("Content-Type", "application/json")
-				req.Header.Add("Accept", "application/json")
-				req.Header.Add("Authorization", fmt.Sprintf("Basic %s", sEnc))
-				res, err := client.Do(req)
-				if err == nil {
-					defer res.Body.Close()
-					if res.StatusCode >= 200 && res.StatusCode < 300 {
-						s1.Verified = true
-					}
+				verified, verificationErr := verifyEasyInsight(ctx, idMatch, keyMatch)
+				s1.Verified = verified
+				if verificationErr != nil {
+					s1.SetVerificationError(verificationErr)
 				}
 			}
 
 			results = append(results, s1)
+			// if key id combination is verified, skip other idMatches for that key
+			if s1.Verified {
+				break
+			}
 		}
-
 	}
 
 	return results, nil
@@ -93,3 +88,38 @@ func (s Scanner) Type() detectorspb.DetectorType {
 func (s Scanner) Description() string {
 	return "EasyInsight is a business intelligence tool that provides data visualization and reporting. EasyInsight API keys can be used to access and manage data within the platform."
 }
+
+func verifyEasyInsight(ctx context.Context, id, key string) (bool, error) {
+	// docs: https://www.easy-insight.com/api/users.html
+	req, err := http.NewRequestWithContext(ctx, "GET", "https://www.easy-insight.com/app/api/users.json", nil)
+	if err != nil {
+		return false, err
+	}
+
+	// add required headers to the request
+	req.Header.Add("Content-Type", "application/json")
+	req.Header.Add("Accept", "application/json")
+	// set basic auth for the request
+	req.SetBasicAuth(id, key)
+
+	res, reqErr := client.Do(req)
+	if reqErr != nil {
+		return false, reqErr
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	switch res.StatusCode {
+	// id, key verified
+	case http.StatusOK:
+		return true, nil
+	// id, key unverified
+	case http.StatusUnauthorized:
+		return false, nil
+	// something invalid
+	default:
+		return false, fmt.Errorf("unexpected status code: %d", res.StatusCode)
+	}
+}

pkg/detectors/easyinsight/easyinsight_integration_test.go

@@ -0,0 +1,121 @@
+//go:build detectors
+// +build detectors
+
+package easyinsight
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/kylelemons/godebug/pretty"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestEasyInsight_FromChunk(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors1")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+	secret := testSecrets.MustGetField("EASYINSIGHT")
+	inactiveSecret := testSecrets.MustGetField("EASYINSIGHT_INACTIVE")
+	id := testSecrets.MustGetField("EASYINSIGHT_ID")
+
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+	tests := []struct {
+		name    string
+		s       Scanner
+		args    args
+		want    []detectors.Result
+		wantErr bool
+	}{
+		{
+			name: "found, verified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a easyinsight secret %s within easyid %s", secret, id)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_EasyInsight,
+					Verified:     true,
+				},
+			},
+			wantErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a easyinsight secret %s within easyid %s but not valid", inactiveSecret, id)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_EasyInsight,
+					Verified:     false,
+				},
+			},
+			wantErr: 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,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			s := Scanner{}
+			got, err := s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("EasyInsight.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])
+				}
+				got[i].Raw = nil
+			}
+			if diff := pretty.Compare(got, tt.want); diff != "" {
+				t.Errorf("EasyInsight.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/detectors/easyinsight/easyinsight_test.go

@@ -1,120 +1,105 @@
-//go:build detectors
-// +build detectors
-
 package easyinsight
 
 import (
 	"context"
 	"fmt"
 	"testing"
-	"time"
 
-	"github.com/kylelemons/godebug/pretty"
+	"github.com/google/go-cmp/cmp"
 
-	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
-	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick"
 )
 
-func TestEasyInsight_FromChunk(t *testing.T) {
-	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
-	defer cancel()
-	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors1")
-	if err != nil {
-		t.Fatalf("could not get test secrets from GCP: %s", err)
-	}
-	secret := testSecrets.MustGetField("EASYINSIGHT")
-	inactiveSecret := testSecrets.MustGetField("EASYINSIGHT_INACTIVE")
-	id := testSecrets.MustGetField("EASYINSIGHT_ID")
+var (
+	validKeyPattern = "987ahjjdasgUcaaraAdd"
+	validIDPattern  = "poiuy76RaEf90ertgh0K"
+	// this should result in 4 combinations
+	complexPattern = `easyinsight credentials
+						these credentials are for testing a pattern
+						key: A876AcaraTsaAKcae09a
+						id: chECk12345ChecK12345
+						-------------------------
+						second credentials:
+						key: B874CDaraTsaAKVBe08A
+						id: CHECK12345ChecK09876`
+	invalidPattern = "poiuy76=a_$90ertgh0K"
+)
+
+func TestEasyInsight_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
 
-	type args struct {
-		ctx    context.Context
-		data   []byte
-		verify bool
-	}
 	tests := []struct {
-		name    string
-		s       Scanner
-		args    args
-		want    []detectors.Result
-		wantErr bool
+		name  string
+		input string
+		want  []string
 	}{
 		{
-			name: "found, verified",
-			s:    Scanner{},
-			args: args{
-				ctx:    context.Background(),
-				data:   []byte(fmt.Sprintf("You can find a easyinsight secret %s within easyid %s", secret, id)),
-				verify: true,
-			},
-			want: []detectors.Result{
-				{
-					DetectorType: detectorspb.DetectorType_EasyInsight,
-					Verified:     true,
-				},
-			},
-			wantErr: false,
+			name:  "valid pattern",
+			input: fmt.Sprintf("easyinsight key = '%s' easy-insight id = '%s", validKeyPattern, validIDPattern),
+			want:  []string{validKeyPattern + validIDPattern, validIDPattern + validKeyPattern},
 		},
 		{
-			name: "found, unverified",
-			s:    Scanner{},
-			args: args{
-				ctx:    context.Background(),
-				data:   []byte(fmt.Sprintf("You can find a easyinsight secret %s within easyid %s but not valid", inactiveSecret, id)), // the secret would satisfy the regex but not pass validation
-				verify: true,
+			name:  "valid pattern - complex",
+			input: fmt.Sprintf("easyinsight token = '%s'", complexPattern),
+			want: []string{
+				"A876AcaraTsaAKcae09achECk12345ChecK12345",
+				"A876AcaraTsaAKcae09aCHECK12345ChecK09876",
+				"B874CDaraTsaAKVBe08ACHECK12345ChecK09876",
+				"B874CDaraTsaAKVBe08AchECk12345ChecK12345",
 			},
-			want: []detectors.Result{
-				{
-					DetectorType: detectorspb.DetectorType_EasyInsight,
-					Verified:     false,
-				},
-			},
-			wantErr: 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,
+			name:  "valid pattern - out of prefix range",
+			input: fmt.Sprintf("easyinsight key and id keyword is not close to the real token = '%s|%s'", validKeyPattern, validIDPattern),
+			want:  nil,
+		},
+		{
+			name:  "invalid pattern",
+			input: fmt.Sprintf("easyinsight = '%s|%s'", invalidPattern, invalidPattern),
+			want:  nil,
 		},
 	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			s := Scanner{}
-			got, err := s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
-			if (err != nil) != tt.wantErr {
-				t.Errorf("EasyInsight.FromData() error = %v, wantErr %v", err, tt.wantErr)
+
+	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
 			}
-			for i := range got {
-				if len(got[i].Raw) == 0 {
-					t.Fatalf("no raw secret present: \n %+v", got[i])
-				}
-				got[i].Raw = nil
-			}
-			if diff := pretty.Compare(got, tt.want); diff != "" {
-				t.Errorf("EasyInsight.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)
+			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)
 			}
 		})
 	}