Detector-Competition-Feat: Adding Azure Batch keys (#1956)

* adding azure batch

* fmt

* fix lint

---------

Co-authored-by: Chair <[email protected]>
Co-authored-by: Zachary Rice <[email protected]>

pkg/detectors/azurebatch/azurebatch.go

@@ -0,0 +1,114 @@
+package azurebatch
+
+import (
+	"context"
+	"crypto/hmac"
+	"crypto/sha256"
+	"encoding/base64"
+	"fmt"
+	"net/http"
+	"regexp"
+	"strings"
+	"time"
+
+	"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.
+	urlPat    = regexp.MustCompile(`https://(.{1,50})\.(.{1,50})\.batch\.azure\.com`)
+	secretPat = regexp.MustCompile(`[A-Za-z0-9+/=]{88}`)
+)
+
+// 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{".batch.azure.com"}
+}
+
+// FromData will find and optionally verify Azurebatch 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)
+
+	urlMatches := urlPat.FindAllStringSubmatch(dataStr, -1)
+	secretMatches := secretPat.FindAllStringSubmatch(dataStr, -1)
+
+	for _, urlMatch := range urlMatches {
+
+		for _, secretMatch := range secretMatches {
+
+			endpoint := urlMatch[0]
+			accountName := urlMatch[1]
+			accountKey := secretMatch[0]
+
+			s1 := detectors.Result{
+				DetectorType: detectorspb.DetectorType_AzureBatch,
+				Raw:          []byte(endpoint),
+				Redacted:     endpoint,
+				RawV2:        []byte(endpoint + accountKey),
+			}
+
+			if verify {
+				client := s.client
+				if client == nil {
+					client = defaultClient
+				}
+				url := fmt.Sprintf("%s/applications?api-version=2020-09-01.12.0", endpoint)
+				date := time.Now().UTC().Format(http.TimeFormat)
+				stringToSign := fmt.Sprintf(
+					"GET\n\n\n\n\napplication/json\n%s\n\n\n\n\n\n%s\napi-version:%s",
+					date,
+					strings.ToLower(fmt.Sprintf("/%s/applications", accountName)),
+					"2020-09-01.12.0",
+				)
+				key, _ := base64.StdEncoding.DecodeString(accountKey)
+				h := hmac.New(sha256.New, key)
+				h.Write([]byte(stringToSign))
+				signature := base64.StdEncoding.EncodeToString(h.Sum(nil))
+				req, err := http.NewRequest("GET", url, nil)
+				if err != nil {
+					continue
+				}
+				req.Header.Set("Content-Type", "application/json")
+				req.Header.Set("Authorization", fmt.Sprintf("SharedKey %s:%s", accountName, signature))
+				req.Header.Set("Date", date)
+				resp, err := client.Do(req)
+				if err != nil {
+					continue
+				}
+				defer resp.Body.Close()
+
+				if resp.StatusCode == http.StatusOK {
+					s1.Verified = true
+				}
+
+			}
+
+			// This function will check false positives for common test words, but also it will make sure the key appears 'random' enough to be a real key.
+			if !s1.Verified && detectors.IsKnownFalsePositive(accountKey, detectors.DefaultFalsePositives, true) {
+				continue
+			}
+
+			results = append(results, s1)
+			if s1.Verified {
+				break
+			}
+		}
+	}
+
+	return results, nil
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_AzureBatch
+}

pkg/detectors/azurebatch/azurebatch_test.go

@@ -0,0 +1,130 @@
+//go:build detectors
+// +build detectors
+
+package azurebatch
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestAzurebatch_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)
+	}
+	url := testSecrets.MustGetField("AZUREBATCH_URL")
+	secret := testSecrets.MustGetField("AZUREBATCH_KEY")
+	inactiveSecret := testSecrets.MustGetField("AZUREBATCH_KEY_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 azurebatch secret %s and %s within", url, secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AzureBatch,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a azurebatch secret %s and %s within but not valid", url, inactiveSecret)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AzureBatch,
+					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,
+		},
+	}
+	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("AzureBatch.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{}, "RawV2", "Raw", "Redacted", "VerificationError")
+
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("AzureBatch.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

@@ -62,6 +62,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/aylien"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/ayrshare"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azure"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azurebatch"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/bannerbear"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/baremetrics"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/beamer"
@@ -780,6 +781,7 @@ func DefaultDetectors() []detectors.Detector {
 		&alibaba.Scanner{},
 		aws.New(),
 		&azure.Scanner{},
+		&azurebatch.Scanner{},
 		&slack.Scanner{}, // has 4 secret types
 		&gitlab.Scanner{},
 		&gitlabv2.Scanner{},

proto/detectors.proto

@@ -965,6 +965,7 @@ enum DetectorType {
   DenoDeploy = 954;
   Stripo = 955;
   ReplyIO = 956;
+  AzureBatch = 957;
 }
 
 message Result {