adding support for new version of figma token (#1813)

* wip examples folder

* Rename examples.md to README.md

.github/workflows/detector-tests.yml

@@ -0,0 +1,46 @@
+name: detector test aggregation
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "0 8 * * *"
+
+jobs:
+  test-detectors:
+    runs-on: ubuntu-latest
+    permissions:
+      actions: "read"
+      contents: "read"
+      id-token: "write"
+    steps:
+      - name: Install Go
+        uses: actions/setup-go@v4
+        with:
+          go-version: "1.21"
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - id: "auth"
+        uses: "google-github-actions/auth@v1"
+        with:
+          workload_identity_provider: "projects/811013774421/locations/global/workloadIdentityPools/github-pool/providers/github-provider"
+          service_account: "github-ci-external@trufflehog-testing.iam.gserviceaccount.com"
+      - name: Test
+        run: make test-integration
+      - name: Set up gotestsum
+        run: |
+          go install gotest.tools/gotestsum@latest
+          mkdir -p tmp/test-results
+      - name: Test
+        run: |
+          CGO_ENABLED=1 gotestsum --junitfile tmp/test-results/test.xml --raw-command -- go test -json -tags=detectors -timeout=15m $(go list ./... | grep pkg/detectors)
+      - name: Upload test results to BuildPulse for flaky test detection
+        if: ${{ !cancelled() }} # Run this step even when the tests fail. Skip if the workflow is cancelled.
+        uses: buildpulse/buildpulse-action@main
+        with:
+          account: 79229934
+          repository: 694446374
+          path: |
+            tmp/test-results/*.xml
+          key: ${{ secrets.BUILDPULSE_DETECTORS_ACCESS_KEY_ID }}
+          secret: ${{ secrets.BUILDPULSE_DETECTORS_SECRET_ACCESS_KEY }}
+          tags: detectors

.github/workflows/test.yml

@@ -1,4 +1,3 @@
-
 name: Test
 
 on:
@@ -14,57 +13,57 @@ jobs:
     if: ${{ ! github.event.pull_request.head.repo.fork }}
     runs-on: ubuntu-latest
     permissions:
-      actions: 'read'
-      contents: 'read'
-      id-token: 'write'
+      actions: "read"
+      contents: "read"
+      id-token: "write"
     steps:
-    - name: Install Go
-      uses: actions/setup-go@v4
-      with:
-        go-version: '1.21'
-    - name: Checkout code
-      uses: actions/checkout@v3
-    - id: 'auth'
-      uses: 'google-github-actions/auth@v1'
-      with:
-        workload_identity_provider: 'projects/811013774421/locations/global/workloadIdentityPools/github-pool/providers/github-provider'
-        service_account: 'github-ci-external@trufflehog-testing.iam.gserviceaccount.com'
-    - name: Test
-      run: make test-integration
-  test-detectors:
-    if: ${{ ! github.event.pull_request.head.repo.fork }}
-    runs-on: ubuntu-latest
-    permissions:
-      actions: 'read'
-      contents: 'read'
-      id-token: 'write'
-    steps:
-    - name: Install Go
-      uses: actions/setup-go@v4
-      with:
-        go-version: '1.21'
-    - name: Checkout code
-      uses: actions/checkout@v3
-    - id: 'auth'
-      uses: 'google-github-actions/auth@v1'
-      with:
-        workload_identity_provider: 'projects/811013774421/locations/global/workloadIdentityPools/github-pool/providers/github-provider'
-        service_account: 'github-ci-external@trufflehog-testing.iam.gserviceaccount.com'
-    - name: Test
-      run: make test-detectors
-      continue-on-error: true
+      - name: Install Go
+        uses: actions/setup-go@v4
+        with:
+          go-version: "1.21"
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - id: "auth"
+        uses: "google-github-actions/auth@v1"
+        with:
+          workload_identity_provider: "projects/811013774421/locations/global/workloadIdentityPools/github-pool/providers/github-provider"
+          service_account: "github-ci-external@trufflehog-testing.iam.gserviceaccount.com"
+      - name: Set up gotestsum
+        run: |
+          go install gotest.tools/gotestsum@latest
+          mkdir -p tmp/test-results
+      - name: Test
+        run: |
+          CGO_ENABLED=1 gotestsum --junitfile tmp/test-results/test.xml --raw-command -- go test -json -tags=sources $(go list ./... | grep -v /vendor/ | grep -v pkg/detectors)
+        if: ${{ success() || failure() }} # always run this step, even if there were previous errors
+      - name: Upload test results to BuildPulse for flaky test detection
+        if: ${{ !cancelled() }} # Run this step even when the tests fail. Skip if the workflow is cancelled.
+        uses: buildpulse/buildpulse-action@main
+        with:
+          account: 79229934
+          repository: 77726177
+          path: |
+            tmp/test-results/*.xml
+          key: ${{ secrets.BUILDPULSE_ACCESS_KEY_ID }}
+          secret: ${{ secrets.BUILDPULSE_SECRET_ACCESS_KEY }}
+          tags: integration
+      - name: Annotate test results
+        uses: mikepenz/action-junit-report@v3
+        if: success() || failure() # always run even if the previous step fails
+        with:
+          report_paths: "tmp/test-results/*.xml"
   test-community:
     if: ${{ github.event.pull_request.head.repo.fork }}
     runs-on: ubuntu-latest
     permissions:
-      actions: 'read'
-      contents: 'read'
+      actions: "read"
+      contents: "read"
     steps:
-    - name: Install Go
-      uses: actions/setup-go@v4
-      with:
-        go-version: '1.21'
-    - name: Checkout code
-      uses: actions/checkout@v3
-    - name: Test
-      run: make test-community
+      - name: Install Go
+        uses: actions/setup-go@v4
+        with:
+          go-version: "1.21"
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Test
+        run: make test-community

Makefile

@@ -1,4 +1,4 @@
-PROTOS_IMAGE ?= trufflesecurity/protos:1.18-0
+PROTOS_IMAGE ?= trufflesecurity/protos:1.21-0
 
 .PHONY: check
 .PHONY: lint
@@ -59,7 +59,7 @@ protos-windows:
 
 release-protos-image:
 	docker buildx build --push --platform=linux/amd64,linux/arm64 \
-	-t trufflesecurity/protos:1.18-0 -f hack/Dockerfile.protos .
+	-t trufflesecurity/protos:1.21-0 -f hack/Dockerfile.protos .
 
 snifftest:
 	./hack/snifftest/snifftest.sh

examples/README.md

@@ -0,0 +1,14 @@
+# Examples
+This folder contains various examples like custom detectors, scripts, etc. Feel free to contribute!
+
+### Generic Detector
+An often requested feature for TruffleHog is a generic detector. By default, we do not support generic detection as it would result in lots of false positives. However, if you want to attempt detect generic secrets you can use a custom detector. 
+
+#### Try it out:
+```
+wget https://raw.githubusercontent.com/trufflesecurity/trufflehog/main/examples/generic.yml
+trufflehog filesystem --config=$PWD/generic.yml $PWD
+
+# to filter so that _only_ generic credentials are logged:
+trufflehog filesystem --config=$PWD/generic.yml --json --no-verification $PWD | awk '/generic-api-key/{print $0}'
+```

examples/generic.yml

@@ -0,0 +1,15 @@
+detectors:
+- name: generic-api-key
+  keywords:
+  - key
+  - api
+  - token
+  - secret
+  - client
+  - passwd
+  - password
+  - auth
+  - access
+  regex:
+    # borrowing the gitleaks generic-api-key regex
+    generic-api-key: "(?i)(?:key|api|token|secret|client|passwd|password|auth|access)(?:[0-9a-z\\-_\\t .]{0,20})(?:[\\s|']|[\\s|\"]){0,3}(?:=|>|:{1,3}=|\\|\\|:|<=|=>|:|\\?=)(?:'|\"|\\s|=|\\x60){0,5}([0-9a-z\\-_.=]{10,150})(?:['|\"|\\n|\\r|\\s|\\x60|;]|$)"

go.mod

@@ -9,8 +9,9 @@ replace github.com/couchbase/gocbcore/v10 => github.com/couchbase/gocbcore/v10 v
 
 require (
 	cloud.google.com/go/secretmanager v1.11.1
-	cloud.google.com/go/storage v1.31.0
+	cloud.google.com/go/storage v1.33.0
 	github.com/Azure/go-autorest/autorest/azure/auth v0.5.11
+	github.com/AzureAD/microsoft-authentication-library-for-go v0.5.1
 	github.com/BobuSumisu/aho-corasick v1.0.3
 	github.com/TheZeroSlave/zapsentry v1.17.0
 	github.com/aws/aws-sdk-go v1.44.83
@@ -20,7 +21,7 @@ require (
 	github.com/bitfinexcom/bitfinex-api-go v0.0.0-20210608095005-9e0b26f200fb
 	github.com/bradleyfalzon/ghinstallation/v2 v2.6.0
 	github.com/charmbracelet/bubbles v0.16.1
-	github.com/charmbracelet/bubbletea v0.24.1
+	github.com/charmbracelet/bubbletea v0.24.2
 	github.com/charmbracelet/glamour v0.6.0
 	github.com/charmbracelet/lipgloss v0.7.1
 	github.com/couchbase/gocb/v2 v2.6.3
@@ -29,10 +30,10 @@ require (
 	github.com/envoyproxy/protoc-gen-validate v1.0.2
 	github.com/fatih/color v1.15.0
 	github.com/felixge/fgprof v0.9.3
-	github.com/getsentry/sentry-go v0.22.0
+	github.com/getsentry/sentry-go v0.24.1
 	github.com/go-errors/errors v1.4.2
 	github.com/go-git/go-git/v5 v5.8.1
-	github.com/go-ldap/ldap/v3 v3.4.5
+	github.com/go-ldap/ldap/v3 v3.4.6
 	github.com/go-logr/logr v1.2.4
 	github.com/go-logr/zapr v1.2.4
 	github.com/go-redis/redis v6.15.9+incompatible
@@ -53,7 +54,8 @@ require (
 	github.com/launchdarkly/go-server-sdk/v6 v6.1.0
 	github.com/lib/pq v1.10.9
 	github.com/lrstanley/bubblezone v0.0.0-20221222153816-e95291e2243e
-	github.com/mattn/go-isatty v0.0.18
+	github.com/marusama/semaphore/v2 v2.5.0
+	github.com/mattn/go-isatty v0.0.19
 	github.com/mattn/go-sqlite3 v1.14.17
 	github.com/mholt/archiver/v4 v4.0.0-alpha.8
 	github.com/muesli/reflow v0.3.0
@@ -70,12 +72,12 @@ require (
 	go.mongodb.org/mongo-driver v1.12.0
 	go.uber.org/mock v0.2.0
 	go.uber.org/zap v1.24.0
-	golang.org/x/crypto v0.12.0
+	golang.org/x/crypto v0.13.0
 	golang.org/x/exp v0.0.0-20230206171751-46f607a40771
-	golang.org/x/oauth2 v0.10.0
+	golang.org/x/oauth2 v0.12.0
 	golang.org/x/sync v0.3.0
-	golang.org/x/text v0.12.0
-	google.golang.org/api v0.131.0
+	golang.org/x/text v0.13.0
+	google.golang.org/api v0.132.0
 	google.golang.org/protobuf v1.31.0
 	gopkg.in/alecthomas/kingpin.v2 v2.2.6
 	gopkg.in/h2non/gock.v1 v1.1.2
@@ -83,7 +85,7 @@ require (
 )
 
 require (
-	cloud.google.com/go v0.110.2 // indirect
+	cloud.google.com/go v0.110.4 // indirect
 	cloud.google.com/go/compute v1.20.1 // indirect
 	cloud.google.com/go/compute/metadata v0.2.3 // indirect
 	cloud.google.com/go/iam v1.1.0 // indirect
@@ -151,7 +153,7 @@ require (
 	github.com/emirpasic/gods v1.18.1 // indirect
 	github.com/form3tech-oss/jwt-go v3.2.5+incompatible // indirect
 	github.com/gabriel-vasile/mimetype v1.4.2 // indirect
-	github.com/go-asn1-ber/asn1-ber v1.5.4 // indirect
+	github.com/go-asn1-ber/asn1-ber v1.5.5 // indirect
 	github.com/go-git/gcfg v1.5.1-0.20230307220236-3a3c6141e376 // indirect
 	github.com/go-git/go-billy/v5 v5.4.1 // indirect
 	github.com/go-ole/go-ole v1.2.6 // indirect
@@ -168,7 +170,7 @@ require (
 	github.com/google/go-querystring v1.1.0 // indirect
 	github.com/google/pprof v0.0.0-20211214055906-6f57359322fd // indirect
 	github.com/google/s2a-go v0.1.4 // indirect
-	github.com/google/uuid v1.3.0 // indirect
+	github.com/google/uuid v1.3.1 // indirect
 	github.com/googleapis/enterprise-certificate-proxy v0.2.5 // indirect
 	github.com/gorilla/css v1.0.0 // indirect
 	github.com/gregjones/httpcache v0.0.0-20171119193500-2bcd89a1743f // indirect
@@ -203,7 +205,7 @@ require (
 	github.com/minio/asm2plan9s v0.0.0-20200509001527-cdd76441f9d8 // indirect
 	github.com/minio/c2goasm v0.0.0-20190812172519-36a3d3bbc4f3 // indirect
 	github.com/mitchellh/go-homedir v1.1.0 // indirect
-	github.com/montanaflynn/stats v0.0.0-20171201202039-1bf9dbcd8cbe // indirect
+	github.com/montanaflynn/stats v0.6.6 // indirect
 	github.com/mtibben/percent v0.2.1 // indirect
 	github.com/muesli/ansi v0.0.0-20211031195517-c9f0611b6c70 // indirect
 	github.com/muesli/cancelreader v0.2.2 // indirect
@@ -244,16 +246,16 @@ require (
 	go.uber.org/multierr v1.6.0 // indirect
 	go4.org v0.0.0-20200411211856-f5505b9728dd // indirect
 	golang.org/x/mod v0.12.0 // indirect
-	golang.org/x/net v0.14.0 // indirect
-	golang.org/x/sys v0.11.0 // indirect
-	golang.org/x/term v0.11.0 // indirect
+	golang.org/x/net v0.15.0 // indirect
+	golang.org/x/sys v0.12.0 // indirect
+	golang.org/x/term v0.12.0 // indirect
 	golang.org/x/time v0.3.0 // indirect
 	golang.org/x/tools v0.12.0 // indirect
 	golang.org/x/xerrors v0.0.0-20220907171357-04be3eba64a2 // indirect
 	google.golang.org/appengine v1.6.7 // indirect
-	google.golang.org/genproto v0.0.0-20230629202037-9506855d4529 // indirect
-	google.golang.org/genproto/googleapis/api v0.0.0-20230629202037-9506855d4529 // indirect
-	google.golang.org/genproto/googleapis/rpc v0.0.0-20230706204954-ccb25ca9f130 // indirect
+	google.golang.org/genproto v0.0.0-20230706204954-ccb25ca9f130 // indirect
+	google.golang.org/genproto/googleapis/api v0.0.0-20230706204954-ccb25ca9f130 // indirect
+	google.golang.org/genproto/googleapis/rpc v0.0.0-20230711160842-782d3b101e98 // indirect
 	google.golang.org/grpc v1.56.2 // indirect
 	gopkg.in/warnings.v0 v0.1.2 // indirect
 	gopkg.in/yaml.v2 v2.4.0 // indirect

go.sum

@@ -7,8 +7,8 @@ cloud.google.com/go v0.45.1/go.mod h1:RpBamKRgapWJb87xiFSdk4g1CME7QZg3uwTez+TSTj
 cloud.google.com/go v0.46.3/go.mod h1:a6bKKbmY7er1mI7TEI4lsAkts/mkhTSZK8w33B4RAg0=
 cloud.google.com/go v0.50.0/go.mod h1:r9sluTvynVuxRIOHXQEHMFffphuXHOMZMycpNR5e6To=
 cloud.google.com/go v0.53.0/go.mod h1:fp/UouUEsRkN6ryDKNW/Upv/JBKnv6WDthjR6+vze6M=
-cloud.google.com/go v0.110.2 h1:sdFPBr6xG9/wkBbfhmUz/JmZC7X6LavQgcrVINrKiVA=
-cloud.google.com/go v0.110.2/go.mod h1:k04UEeEtb6ZBRTv3dZz4CeJC3jKGxyhl0sAiVVquxiw=
+cloud.google.com/go v0.110.4 h1:1JYyxKMN9hd5dR2MYTPWkGUgcoxVVhg0LKNKEo0qvmk=
+cloud.google.com/go v0.110.4/go.mod h1:+EYjdK8e5RME/VY/qLCAtuyALQ9q67dvuum8i+H5xsI=
 cloud.google.com/go/bigquery v1.0.1/go.mod h1:i/xbL2UlR5RvWAURpBYZTtm/cXjCha9lbfbpx4poX+o=
 cloud.google.com/go/bigquery v1.3.0/go.mod h1:PjpwJnslEMmckchkHFfq+HTD2DmtT67aNFKH1/VBDHE=
 cloud.google.com/go/compute v1.20.1 h1:6aKEtlUiwEpJzM001l0yFkpXmUVXaN8W+fbkb2AZNbg=
@@ -25,8 +25,8 @@ cloud.google.com/go/secretmanager v1.11.1 h1:cLTCwAjFh9fKvU6F13Y4L9vPcx9yiWPyWXE
 cloud.google.com/go/secretmanager v1.11.1/go.mod h1:znq9JlXgTNdBeQk9TBW/FnR/W4uChEKGeqQWAJ8SXFw=
 cloud.google.com/go/storage v1.0.0/go.mod h1:IhtSnM/ZTZV8YYJWCY8RULGVqBDmpoyjwiyrjsg+URw=
 cloud.google.com/go/storage v1.5.0/go.mod h1:tpKbwo567HUNpVclU5sGELwQWBDZ8gh0ZeosJ0Rtdos=
-cloud.google.com/go/storage v1.31.0 h1:+S3LjjEN2zZ+L5hOwj4+1OkGCsLVe0NzpXKQ1pSdTCI=
-cloud.google.com/go/storage v1.31.0/go.mod h1:81ams1PrhW16L4kF7qg+4mTq7SRs5HsbDTM0bWvrwJ0=
+cloud.google.com/go/storage v1.33.0 h1:PVrDOkIC8qQVa1P3SXGpQvfuJhN2LHOoyZvWs8D2X5M=
+cloud.google.com/go/storage v1.33.0/go.mod h1:Hhh/dogNRGca7IWv1RC2YqEn0c0G77ctA/OxflYkiD8=
 dario.cat/mergo v1.0.0 h1:AGCNq9Evsj31mOgNPcLyXc+4PNABt905YmuqPYYpBWk=
 dario.cat/mergo v1.0.0/go.mod h1:uNxQE+84aUszobStD9th8a29P2fMDhsBdgRYvZOxGmk=
 dmitri.shuralyov.com/gpu/mtl v0.0.0-20190408044501-666a987793e9/go.mod h1:H6x//7gZCb22OMCxBHrMx7a5I7Hp++hsVxbQ4BYO7hU=
@@ -178,8 +178,8 @@ github.com/cespare/xxhash/v2 v2.2.0 h1:DC2CZ1Ep5Y4k3ZQ899DldepgrayRUGE6BBZ/cd9Cj
 github.com/cespare/xxhash/v2 v2.2.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
 github.com/charmbracelet/bubbles v0.16.1 h1:6uzpAAaT9ZqKssntbvZMlksWHruQLNxg49H5WdeuYSY=
 github.com/charmbracelet/bubbles v0.16.1/go.mod h1:2QCp9LFlEsBQMvIYERr7Ww2H2bA7xen1idUDIzm/+Xc=
-github.com/charmbracelet/bubbletea v0.24.1 h1:LpdYfnu+Qc6XtvMz6d/6rRY71yttHTP5HtrjMgWvixc=
-github.com/charmbracelet/bubbletea v0.24.1/go.mod h1:rK3g/2+T8vOSEkNHvtq40umJpeVYDn6bLaqbgzhL/hg=
+github.com/charmbracelet/bubbletea v0.24.2 h1:uaQIKx9Ai6Gdh5zpTbGiWpytMU+CfsPp06RaW2cx/SY=
+github.com/charmbracelet/bubbletea v0.24.2/go.mod h1:XdrNrV4J8GiyshTtx3DNuYkR1FDaJmO3l2nejekbsgg=
 github.com/charmbracelet/glamour v0.6.0 h1:wi8fse3Y7nfcabbbDuwolqTqMQPMnVPeZhDM273bISc=
 github.com/charmbracelet/glamour v0.6.0/go.mod h1:taqWV4swIMMbWALc0m7AfE9JkPSU8om2538k9ITBxOc=
 github.com/charmbracelet/lipgloss v0.7.1 h1:17WMwi7N1b1rVWOjMT+rCh7sQkvDU75B2hbZpc5Kc1E=
@@ -263,13 +263,13 @@ github.com/fsnotify/fsnotify v1.4.9 h1:hsms1Qyu0jgnwNXIxa+/V/PDsU6CfLf6CNO8H7IWo
 github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ=
 github.com/gabriel-vasile/mimetype v1.4.2 h1:w5qFW6JKBz9Y393Y4q372O9A7cUSequkh1Q7OhCmWKU=
 github.com/gabriel-vasile/mimetype v1.4.2/go.mod h1:zApsH/mKG4w07erKIaJPFiX0Tsq9BFQgN3qGY5GnNgA=
-github.com/getsentry/sentry-go v0.22.0 h1:XNX9zKbv7baSEI65l+H1GEJgSeIC1c7EN5kluWaP6dM=
-github.com/getsentry/sentry-go v0.22.0/go.mod h1:lc76E2QywIyW8WuBnwl8Lc4bkmQH4+w1gwTf25trprY=
+github.com/getsentry/sentry-go v0.24.1 h1:W6/0GyTy8J6ge6lVCc94WB6Gx2ZuLrgopnn9w8Hiwuk=
+github.com/getsentry/sentry-go v0.24.1/go.mod h1:lc76E2QywIyW8WuBnwl8Lc4bkmQH4+w1gwTf25trprY=
 github.com/ghodss/yaml v1.0.0/go.mod h1:4dBDuWmgqj2HViK6kFavaiC9ZROes6MMH2rRYeMEF04=
 github.com/gliderlabs/ssh v0.3.5 h1:OcaySEmAQJgyYcArR+gGGTHCyE7nvhEMTlYY+Dp8CpY=
 github.com/gliderlabs/ssh v0.3.5/go.mod h1:8XB4KraRrX39qHhT6yxPsHedjA08I/uBVwj4xC+/+z4=
-github.com/go-asn1-ber/asn1-ber v1.5.4 h1:vXT6d/FNDiELJnLb6hGNa309LMsrCoYFvpwHDF0+Y1A=
-github.com/go-asn1-ber/asn1-ber v1.5.4/go.mod h1:hEBeB/ic+5LoWskz+yKT7vGhhPYkProFKoKdwZRWMe0=
+github.com/go-asn1-ber/asn1-ber v1.5.5 h1:MNHlNMBDgEKD4TcKr36vQN68BA00aDfjIt3/bD50WnA=
+github.com/go-asn1-ber/asn1-ber v1.5.5/go.mod h1:hEBeB/ic+5LoWskz+yKT7vGhhPYkProFKoKdwZRWMe0=
 github.com/go-errors/errors v1.4.2 h1:J6MZopCL4uSllY1OfXM374weqZFFItUbrImctkmUxIA=
 github.com/go-errors/errors v1.4.2/go.mod h1:sIVyrIiJhuEF+Pj9Ebtd6P/rEYROXFi3BopGUQ5a5Og=
 github.com/go-git/gcfg v1.5.1-0.20230307220236-3a3c6141e376 h1:+zs/tPmkDkHx3U66DAb0lQFJrpS6731Oaa12ikc+DiI=
@@ -282,8 +282,8 @@ github.com/go-git/go-git/v5 v5.8.1 h1:Zo79E4p7TRk0xoRgMq0RShiTHGKcKI4+DI6BfJc/Q+
 github.com/go-git/go-git/v5 v5.8.1/go.mod h1:FHFuoD6yGz5OSKEBK+aWN9Oah0q54Jxl0abmj6GnqAo=
 github.com/go-gl/glfw v0.0.0-20190409004039-e6da0acd62b1/go.mod h1:vR7hzQXu2zJy9AVAgeJqvqgH9Q5CA+iKCZ2gyEVpxRU=
 github.com/go-gl/glfw/v3.3/glfw v0.0.0-20191125211704-12ad95a8df72/go.mod h1:tQ2UAYgL5IevRw8kRxooKSPJfGvJ9fJQFa0TUsXzTg8=
-github.com/go-ldap/ldap/v3 v3.4.5 h1:ekEKmaDrpvR2yf5Nc/DClsGG9lAmdDixe44mLzlW5r8=
-github.com/go-ldap/ldap/v3 v3.4.5/go.mod h1:bMGIq3AGbytbaMwf8wdv5Phdxz0FWHTIYMSzyrYgnQs=
+github.com/go-ldap/ldap/v3 v3.4.6 h1:ert95MdbiG7aWo/oPYp9btL3KJlMPKnP58r09rI8T+A=
+github.com/go-ldap/ldap/v3 v3.4.6/go.mod h1:IGMQANNtxpsOzj7uUAMjpGBaOVTC4DYyIy8VsTdxmtc=
 github.com/go-logr/logr v1.2.4 h1:g01GSCwiDw2xSZfjJ2/T9M+S6pFdcNtFYsp+Y43HYDQ=
 github.com/go-logr/logr v1.2.4/go.mod h1:jdQByPbusPIv2/zmleS9BjJVeZ6kBagPoEUsqbVz/1A=
 github.com/go-logr/zapr v1.2.4 h1:QHVo+6stLbfJmYGkQ7uGHUCu5hnAFAj6mDe6Ea0SeOo=
@@ -304,6 +304,7 @@ github.com/goccy/go-json v0.10.0 h1:mXKd9Qw4NuzShiRlOXKews24ufknHO7gx30lsDyokKA=
 github.com/goccy/go-json v0.10.0/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
 github.com/godbus/dbus v0.0.0-20190726142602-4481cbc300e2 h1:ZpnhV/YsD2/4cESfV5+Hoeu/iUR3ruzNvZ+yQfO03a0=
 github.com/godbus/dbus v0.0.0-20190726142602-4481cbc300e2/go.mod h1:bBOAhwG1umN6/6ZUMtDFBMQR8jRg9O75tm9K00oMsK4=
+github.com/golang-jwt/jwt v3.2.1+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
 github.com/golang-jwt/jwt v3.2.2+incompatible h1:IfV12K8xAKAnZqdXVzCZ+TOjboZ2keLg81eXfW3O+oY=
 github.com/golang-jwt/jwt v3.2.2+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
 github.com/golang-jwt/jwt/v4 v4.0.0/go.mod h1:/xlHOz8bRuivTWchD4jCa+NbatV+wEUSzwAxVc6locg=
@@ -382,9 +383,11 @@ github.com/google/pprof v0.0.0-20211214055906-6f57359322fd/go.mod h1:KgnwoLYCZ8I
 github.com/google/renameio v0.1.0/go.mod h1:KWCgfxg9yswjAJkECMjeO8J8rahYeXnNhOm40UhjYkI=
 github.com/google/s2a-go v0.1.4 h1:1kZ/sQM3srePvKs3tXAvQzo66XfcReoqFpIpIccE7Oc=
 github.com/google/s2a-go v0.1.4/go.mod h1:Ej+mSEMGRnqRzjc7VtF+jdBwYG5fuJfiZ8ELkjEwM0A=
+github.com/google/uuid v1.1.1/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/google/uuid v1.1.2/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
-github.com/google/uuid v1.3.0 h1:t6JiXgmwXMjEs8VusXIJk2BXHsn+wx8BZdTaoZ5fu7I=
 github.com/google/uuid v1.3.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/google/uuid v1.3.1 h1:KjJaJ9iWZ3jOFZIf1Lqf4laDRCasjl0BCmnEGxkdLb4=
+github.com/google/uuid v1.3.1/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/googleapis/enterprise-certificate-proxy v0.2.5 h1:UR4rDjcgpgEnqpIEvkiqTYKBCKLNmlge2eVjoZfySzM=
 github.com/googleapis/enterprise-certificate-proxy v0.2.5/go.mod h1:RxW0N9901Cko1VOCW3SXCpWP+mlIEkk2tP7jnHy9a3w=
 github.com/googleapis/gax-go/v2 v2.0.4/go.mod h1:0Wqv26UfaUD9n4G6kQubkQ+KchISgw+vpHVxEJEs9eg=
@@ -489,13 +492,15 @@ github.com/lucasb-eyer/go-colorful v1.2.0 h1:1nnpGOrhyZZuNyfu1QjKiUICQ74+3FNCN69
 github.com/lucasb-eyer/go-colorful v1.2.0/go.mod h1:R4dSotOR9KMtayYi1e77YzuveK+i7ruzyGqttikkLy0=
 github.com/mailru/easyjson v0.7.7 h1:UGYAvKxe3sBsEDzO8ZeWOSlIQfWFlxbzLZe7hwFURr0=
 github.com/mailru/easyjson v0.7.7/go.mod h1:xzfreul335JAWq5oZzymOObrkdz5UnU4kGfJJLY9Nlc=
+github.com/marusama/semaphore/v2 v2.5.0 h1:o/1QJD9DBYOWRnDhPwDVAXQn6mQYD0gZaS1Tpx6DJGM=
+github.com/marusama/semaphore/v2 v2.5.0/go.mod h1:z9nMiNUekt/LTpTUQdpp+4sJeYqUGpwMHfW0Z8V8fnQ=
 github.com/matryer/is v1.2.0 h1:92UTHpy8CDwaJ08GqLDzhhuixiBUUD1p3AU6PHddz4A=
 github.com/matryer/is v1.2.0/go.mod h1:2fLPjFQM9rhQ15aVEtbuwhJinnOqrmgXPNdZsdwlWXA=
 github.com/mattn/go-colorable v0.1.13 h1:fFA4WZxdEF4tXPZVKMLwD8oUnCTTo08duU7wxecdEvA=
 github.com/mattn/go-colorable v0.1.13/go.mod h1:7S9/ev0klgBDR4GtXTXX8a3vIGJpMovkB8vQcUbaXHg=
 github.com/mattn/go-isatty v0.0.16/go.mod h1:kYGgaQfpe5nmfYZH+SKPsOc2e4SrIfOl2e/yFXSvRLM=
-github.com/mattn/go-isatty v0.0.18 h1:DOKFKCQ7FNG2L1rbrmstDN4QVRdS89Nkh85u68Uwp98=
-github.com/mattn/go-isatty v0.0.18/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
+github.com/mattn/go-isatty v0.0.19 h1:JITubQf0MOLdlGRuRq+jtsDlekdYPia9ZFsB8h/APPA=
+github.com/mattn/go-isatty v0.0.19/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
 github.com/mattn/go-localereader v0.0.1 h1:ygSAOl7ZXTx4RdPYinUpg6W99U8jWvWi9Ye2JC/oIi4=
 github.com/mattn/go-localereader v0.0.1/go.mod h1:8fBrzywKY7BI3czFoHkuzRoWE9C+EiG4R1k4Cjx5p88=
 github.com/mattn/go-runewidth v0.0.9/go.mod h1:H031xJmbD/WCDINGzjvQ9THkh0rPKHF+m2gUSrubnMI=
@@ -518,8 +523,9 @@ github.com/minio/c2goasm v0.0.0-20190812172519-36a3d3bbc4f3/go.mod h1:RagcQ7I8Ie
 github.com/mitchellh/go-homedir v1.1.0 h1:lukF9ziXFxDFPkA1vsr5zpc1XuPDn/wFntq5mG+4E0Y=
 github.com/mitchellh/go-homedir v1.1.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrkLzIz1N1q0pr0=
 github.com/modocache/gover v0.0.0-20171022184752-b58185e213c5/go.mod h1:caMODM3PzxT8aQXRPkAt8xlV/e7d7w8GM5g0fa5F0D8=
-github.com/montanaflynn/stats v0.0.0-20171201202039-1bf9dbcd8cbe h1:iruDEfMl2E6fbMZ9s0scYfZQ84/6SPL6zC8ACM2oIL0=
 github.com/montanaflynn/stats v0.0.0-20171201202039-1bf9dbcd8cbe/go.mod h1:wL8QJuTMNUDYhXwkmfOly8iTdp5TEcJFWZD2D7SIkUc=
+github.com/montanaflynn/stats v0.6.6 h1:Duep6KMIDpY4Yo11iFsvyqJDyfzLF9+sndUKT+v64GQ=
+github.com/montanaflynn/stats v0.6.6/go.mod h1:etXPPgVO6n31NxCd9KQUMvCM+ve0ruNzt6R8Bnaayow=
 github.com/mtibben/percent v0.2.1 h1:5gssi8Nqo8QU/r2pynCm+hBQHpkB/uNK7BJCFogWdzs=
 github.com/mtibben/percent v0.2.1/go.mod h1:KG9uO+SZkUp+VkRHsCdYQV3XSZrrSpR3O9ibNBTZrns=
 github.com/muesli/ansi v0.0.0-20211031195517-c9f0611b6c70 h1:kMlmsLSbjkikxQJ1IPwaM+7LJ9ltFu/fi8CRzvSnQmA=
@@ -565,6 +571,7 @@ github.com/pingcap/errors v0.11.4/go.mod h1:Oi8TUi2kEtXXLMJk9l1cGmz20kV3TaQ0usTw
 github.com/pjbgf/sha1cd v0.3.0 h1:4D5XXmUUBUl/xQ6IjCkEAbqXskkq/4O7LmGn0AqMDs4=
 github.com/pjbgf/sha1cd v0.3.0/go.mod h1:nZ1rrWOcGJ5uZgEEVL1VUM9iRQiZvWdbZjkKyFzPPsI=
 github.com/pkg/browser v0.0.0-20180916011732-0a3d74bf9ce4/go.mod h1:4OwLy04Bl9Ef3GJJCoec+30X3LQs/0/m4HFRt/2LUSA=
+github.com/pkg/browser v0.0.0-20210115035449-ce105d075bb4/go.mod h1:N6UoU20jOqggOuDwUaBQpluzLNDqif3kq9z2wpdYEfQ=
 github.com/pkg/browser v0.0.0-20210911075715-681adbf594b8 h1:KoWmjvw+nsYOo29YJK9vDA65RGE3NrOnUtO7a+RF9HU=
 github.com/pkg/browser v0.0.0-20210911075715-681adbf594b8/go.mod h1:HKlIX3XHQyzLZPlr7++PzdhaXEj94dEiJgZDTsxEqUI=
 github.com/pkg/diff v0.0.0-20200914180035-5b29258ca4f7 h1:+/+DxvQaYifJ+grD4klzrS5y+KJXldn/2YTl5JG+vZ8=
@@ -701,8 +708,8 @@ golang.org/x/crypto v0.0.0-20220314234659-1baeb1ce4c0b/go.mod h1:IxCIyHEi3zRg3s0
 golang.org/x/crypto v0.0.0-20220622213112-05595931fe9d/go.mod h1:IxCIyHEi3zRg3s0A5j5BB6A9Jmi73HwBIUl50j+osU4=
 golang.org/x/crypto v0.3.1-0.20221117191849-2c476679df9a/go.mod h1:hebNnKkNXi2UzZN1eVRvBB7co0a+JxK6XbPiWVs/3J4=
 golang.org/x/crypto v0.7.0/go.mod h1:pYwdfH91IfpZVANVyUOhSIPZaFoJGxTFbZhFTx+dXZU=
-golang.org/x/crypto v0.12.0 h1:tFM/ta59kqch6LlvYnPa0yx5a83cL2nHflFhYKvv9Yk=
-golang.org/x/crypto v0.12.0/go.mod h1:NF0Gs7EO5K4qLn+Ylc+fih8BSTeIjAP05siRnAh98yw=
+golang.org/x/crypto v0.13.0 h1:mvySKfSWJ+UKUii46M40LOvyWfN0s2U+46/jDd0e6Ck=
+golang.org/x/crypto v0.13.0/go.mod h1:y6Z2r+Rw4iayiXXAIxJIDAJ1zMW4yaTpebo8fPOliYc=
 golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
 golang.org/x/exp v0.0.0-20190306152737-a1d7652674e8/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
 golang.org/x/exp v0.0.0-20190510132918-efd6b22b2522/go.mod h1:ZjyILWgesfNpC6sMxTJOJm9Kp84zZh5NQWvqDGG3Qr8=
@@ -767,16 +774,16 @@ golang.org/x/net v0.2.0/go.mod h1:KqCZLdyyvdV855qA2rE3GC2aiw5xGR5TEjj8smXukLY=
 golang.org/x/net v0.6.0/go.mod h1:2Tu9+aMcznHK/AK1HMvgo6xiTLG5rD5rZLDS+rp2Bjs=
 golang.org/x/net v0.8.0/go.mod h1:QVkue5JL9kW//ek3r6jTKnTFis1tRmNAW2P1shuFdJc=
 golang.org/x/net v0.10.0/go.mod h1:0qNGK6F8kojg2nk9dLZ2mShWaEBan6FAoqfSigmmuDg=
-golang.org/x/net v0.14.0 h1:BONx9s002vGdD9umnlX1Po8vOZmrgH34qlHcD1MfK14=
-golang.org/x/net v0.14.0/go.mod h1:PpSgVXXLK0OxS0F31C1/tv6XNguvCrnXIDrFMspZIUI=
+golang.org/x/net v0.15.0 h1:ugBLEUaxABaB5AJqW9enI0ACdci2RUd4eP51NTBvuJ8=
+golang.org/x/net v0.15.0/go.mod h1:idbUs1IY1+zTqbi8yxTbhexhEEk5ur9LInksu6HrEpk=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
 golang.org/x/oauth2 v0.0.0-20190226205417-e64efc72b421/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20190604053449-0f29369cfe45/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20191202225959-858c2ad4c8b6/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20200107190931-bf48bf16ab8d/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.8.0/go.mod h1:yr7u4HXZRm1R1kBWqr/xKNqewf0plRYoB7sla+BCIXE=
-golang.org/x/oauth2 v0.10.0 h1:zHCpF2Khkwy4mMB4bv0U37YtJdTGW8jI0glAApi0Kh8=
-golang.org/x/oauth2 v0.10.0/go.mod h1:kTpgurOux7LqtuxjuyZa4Gj2gdezIt/jQtGnNFfypQI=
+golang.org/x/oauth2 v0.12.0 h1:smVPGxink+n1ZI5pkQa8y6fZT0RW0MgCO5bFpepy4B4=
+golang.org/x/oauth2 v0.12.0/go.mod h1:A74bZ3aGXgCY0qaIC9Ahg6Lglin4AMAco8cIv9baba4=
 golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20181108010431-42b317875d0f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20181221193216-37e7f081c4d4/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
@@ -835,16 +842,16 @@ golang.org/x/sys v0.3.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.5.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.8.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.11.0 h1:eG7RXZHdqOJ1i+0lgLgCpSXAp6M3LYlAo6osgSi0xOM=
-golang.org/x/sys v0.11.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.12.0 h1:CM0HF96J0hcLAwsHPJZjfdNzs0gftsLfgKt57wWHJ0o=
+golang.org/x/sys v0.12.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/term v0.0.0-20201126162022-7de9c90e9dd1/go.mod h1:bj7SfCRtBDWHUb9snDiAeCFNEtKQo2Wmx5Cou7ajbmo=
 golang.org/x/term v0.0.0-20210927222741-03fcf44c2211/go.mod h1:jbD1KX2456YbFQfuXm/mYQcufACuNUgVhRMnK/tPxf8=
 golang.org/x/term v0.2.0/go.mod h1:TVmDHMZPmdnySmBfhjOoOdhjzdE1h4u1VwSiw2l1Nuc=
 golang.org/x/term v0.5.0/go.mod h1:jMB1sMXY+tzblOD4FWmEbocvup2/aLOaQEp7JmGp78k=
 golang.org/x/term v0.6.0/go.mod h1:m6U89DPEgQRMq3DNkDClhWw02AUbt2daBVO4cn4Hv9U=
 golang.org/x/term v0.8.0/go.mod h1:xPskH00ivmX89bAKVGSKKtLOWNx2+17Eiy94tnKShWo=
-golang.org/x/term v0.11.0 h1:F9tnn/DA/Im8nCwm+fX+1/eBwi4qFjRT++MhtVC4ZX0=
-golang.org/x/term v0.11.0/go.mod h1:zC9APTIj3jG3FdV/Ons+XE1riIZXG4aZ4GTHiPZJPIU=
+golang.org/x/term v0.12.0 h1:/ZfYdc3zq+q02Rv9vGqTeSItdzZTSNDmfTi0mBAuidU=
+golang.org/x/term v0.12.0/go.mod h1:owVbMEjm3cBLCHdkQu9b1opXd4ETQWc3BhuQGKgXgvU=
 golang.org/x/text v0.0.0-20170915032832-14c0d48ead0c/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.1-0.20180807135948-17ff2d5776d2/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
@@ -857,8 +864,8 @@ golang.org/x/text v0.4.0/go.mod h1:mrYo+phRRbMaCq/xk9113O4dZlRixOauAjOtrjsXDZ8=
 golang.org/x/text v0.7.0/go.mod h1:mrYo+phRRbMaCq/xk9113O4dZlRixOauAjOtrjsXDZ8=
 golang.org/x/text v0.8.0/go.mod h1:e1OnstbJyHTd6l/uOt8jFFHp6TRDWZR/bV3emEE/zU8=
 golang.org/x/text v0.9.0/go.mod h1:e1OnstbJyHTd6l/uOt8jFFHp6TRDWZR/bV3emEE/zU8=
-golang.org/x/text v0.12.0 h1:k+n5B8goJNdU7hSvEtMUz3d1Q6D/XW4COJSJR6fN0mc=
-golang.org/x/text v0.12.0/go.mod h1:TvPlkZtksWOMsz7fbANvkp4WM8x/WCo/om8BMLbz+aE=
+golang.org/x/text v0.13.0 h1:ablQoSUd0tRdKxZewP80B+BaqeKJuVhuRxj/dkrun3k=
+golang.org/x/text v0.13.0/go.mod h1:TvPlkZtksWOMsz7fbANvkp4WM8x/WCo/om8BMLbz+aE=
 golang.org/x/time v0.0.0-20181108054448-85acf8d2951c/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.0.0-20190308202827-9d24e82272b4/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.3.0 h1:rg5rLMjNzMS1RkNLzCG38eapWhnYLFYXDXj2gOlr8j4=
@@ -910,8 +917,8 @@ google.golang.org/api v0.13.0/go.mod h1:iLdEw5Ide6rF15KTC1Kkl0iskquN2gFfn9o9XIsb
 google.golang.org/api v0.14.0/go.mod h1:iLdEw5Ide6rF15KTC1Kkl0iskquN2gFfn9o9XIsbkAI=
 google.golang.org/api v0.15.0/go.mod h1:iLdEw5Ide6rF15KTC1Kkl0iskquN2gFfn9o9XIsbkAI=
 google.golang.org/api v0.17.0/go.mod h1:BwFmGc8tA3vsd7r/7kR8DY7iEEGSU04BFxCo5jP/sfE=
-google.golang.org/api v0.131.0 h1:AcgWS2edQ4chVEt/SxgDKubVu/9/idCJy00tBGuGB4M=
-google.golang.org/api v0.131.0/go.mod h1:7vtkbKv2REjJbxmHSkBTBQ5LUGvPdAqjjvt84XAfhpA=
+google.golang.org/api v0.132.0 h1:8t2/+qZ26kAOGSmOiHwVycqVaDg7q3JDILrNi/Z6rvc=
+google.golang.org/api v0.132.0/go.mod h1:AeTBC6GpJnJSRJjktDcPX0QwtS8pGYZOV6MSuSCusw0=
 google.golang.org/appengine v1.1.0/go.mod h1:EbEs0AVv82hx2wNQdGPgUI5lhzA/G0D9YwlJXL52JkM=
 google.golang.org/appengine v1.4.0/go.mod h1:xpcJRLb0r/rnEns0DIKYYv+WjYCduHsrkT7/EB5XEv4=
 google.golang.org/appengine v1.5.0/go.mod h1:xpcJRLb0r/rnEns0DIKYYv+WjYCduHsrkT7/EB5XEv4=
@@ -934,12 +941,12 @@ google.golang.org/genproto v0.0.0-20191230161307-f3c370f40bfb/go.mod h1:n3cpQtvx
 google.golang.org/genproto v0.0.0-20200212174721-66ed5ce911ce/go.mod h1:55QSHmfGQM9UVYDPBsyGGes0y52j32PQ3BqQfXhyH3c=
 google.golang.org/genproto v0.0.0-20200513103714-09dca8ec2884/go.mod h1:55QSHmfGQM9UVYDPBsyGGes0y52j32PQ3BqQfXhyH3c=
 google.golang.org/genproto v0.0.0-20200526211855-cb27e3aa2013/go.mod h1:NbSheEEYHJ7i3ixzK3sjbqSGDJWnxyFXZblF3eUsNvo=
-google.golang.org/genproto v0.0.0-20230629202037-9506855d4529 h1:9JucMWR7sPvCxUFd6UsOUNmA5kCcWOfORaT3tpAsKQs=
-google.golang.org/genproto v0.0.0-20230629202037-9506855d4529/go.mod h1:xZnkP7mREFX5MORlOPEzLMr+90PPZQ2QWzrVTWfAq64=
-google.golang.org/genproto/googleapis/api v0.0.0-20230629202037-9506855d4529 h1:s5YSX+ZH5b5vS9rnpGymvIyMpLRJizowqDlOuyjXnTk=
-google.golang.org/genproto/googleapis/api v0.0.0-20230629202037-9506855d4529/go.mod h1:vHYtlOoi6TsQ3Uk2yxR7NI5z8uoV+3pZtR4jmHIkRig=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20230706204954-ccb25ca9f130 h1:2FZP5XuJY9zQyGM5N0rtovnoXjiMUEIUMvw0m9wlpLc=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20230706204954-ccb25ca9f130/go.mod h1:8mL13HKkDa+IuJ8yruA3ci0q+0vsUz4m//+ottjwS5o=
+google.golang.org/genproto v0.0.0-20230706204954-ccb25ca9f130 h1:Au6te5hbKUV8pIYWHqOUZ1pva5qK/rwbIhoXEUB9Lu8=
+google.golang.org/genproto v0.0.0-20230706204954-ccb25ca9f130/go.mod h1:O9kGHb51iE/nOGvQaDUuadVYqovW56s5emA88lQnj6Y=
+google.golang.org/genproto/googleapis/api v0.0.0-20230706204954-ccb25ca9f130 h1:XVeBY8d/FaK4848myy41HBqnDwvxeV3zMZhwN1TvAMU=
+google.golang.org/genproto/googleapis/api v0.0.0-20230706204954-ccb25ca9f130/go.mod h1:mPBs5jNgx2GuQGvFwUvVKqtn6HsUw9nP64BedgvqEsQ=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20230711160842-782d3b101e98 h1:bVf09lpb+OJbByTj913DRJioFFAjf/ZGxEz7MajTp2U=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20230711160842-782d3b101e98/go.mod h1:TUfxEVdsvPg18p6AslUXFoLdpED4oBnGwyqk3dV1XzM=
 google.golang.org/grpc v1.19.0/go.mod h1:mqu4LbDTu4XGKhr4mRzUsmM4RtVoemTSY81AxZiDr8c=
 google.golang.org/grpc v1.20.1/go.mod h1:10oTOabMzJvdu6/UiuZezV6QK5dSlG84ov/aaiqXj38=
 google.golang.org/grpc v1.21.1/go.mod h1:oYelfM1adQP15Ek0mdvEgi9Df8B9CZIaU1084ijfRaM=

hack/Dockerfile.protos

@@ -1,6 +1,6 @@
 # trufflesecurity/protos:1.18-0
 
-FROM golang:1.18-buster
+FROM golang:1.21-bullseye
 
 ARG TARGETARCH
 ARG TARGETOS

main.go

@@ -105,11 +105,11 @@ var (
 	filesystemScanIncludePaths = filesystemScan.Flag("include-paths", "Path to file with newline separated regexes for files to include in scan.").Short('i').String()
 	filesystemScanExcludePaths = filesystemScan.Flag("exclude-paths", "Path to file with newline separated regexes for files to exclude in scan.").Short('x').String()
 
-	s3Scan    = cli.Command("s3", "Find credentials in S3 buckets.")
-	s3ScanKey = s3Scan.Flag("key", "S3 key used to authenticate. Can be provided with environment variable AWS_ACCESS_KEY_ID.").Envar("AWS_ACCESS_KEY_ID").String()
-	s3ScanRoleArns     = s3Scan.Flag("role-arn", "Specify the ARN of an IAM role to assume for scanning. You can repeat this flag.").Strings()
-	s3ScanSecret       = s3Scan.Flag("secret", "S3 secret used to authenticate. Can be provided with environment variable AWS_SECRET_ACCESS_KEY.").Envar("AWS_SECRET_ACCESS_KEY").String()
-	s3ScanSessionToken = s3Scan.Flag("session-token", "S3 session token used to authenticate temporary credentials. Can be provided with environment variable AWS_SESSION_TOKEN.").Envar("AWS_SESSION_TOKEN").String()
+	s3Scan              = cli.Command("s3", "Find credentials in S3 buckets.")
+	s3ScanKey           = s3Scan.Flag("key", "S3 key used to authenticate. Can be provided with environment variable AWS_ACCESS_KEY_ID.").Envar("AWS_ACCESS_KEY_ID").String()
+	s3ScanRoleArns      = s3Scan.Flag("role-arn", "Specify the ARN of an IAM role to assume for scanning. You can repeat this flag.").Strings()
+	s3ScanSecret        = s3Scan.Flag("secret", "S3 secret used to authenticate. Can be provided with environment variable AWS_SECRET_ACCESS_KEY.").Envar("AWS_SECRET_ACCESS_KEY").String()
+	s3ScanSessionToken  = s3Scan.Flag("session-token", "S3 session token used to authenticate temporary credentials. Can be provided with environment variable AWS_SESSION_TOKEN.").Envar("AWS_SESSION_TOKEN").String()
 	s3ScanCloudEnv      = s3Scan.Flag("cloud-environment", "Use IAM credentials in cloud environment.").Bool()
 	s3ScanBuckets       = s3Scan.Flag("bucket", "Name of S3 bucket to scan. You can repeat this flag.").Strings()
 	s3ScanMaxObjectSize = s3Scan.Flag("max-object-size", "Maximum size of objects to scan. Objects larger than this will be skipped. (Byte units eg. 512B, 2KB, 4MB)").Default("250MB").Bytes()
@@ -354,6 +354,10 @@ func run(state overseer.State) {
 		printer = new(output.PlainPrinter)
 	}
 
+	if !*jsonLegacy && !*jsonOut {
+		fmt.Fprintf(os.Stderr, "🐷🔑🐷  TruffleHog. Unearth your secrets. 🐷🔑🐷\n\n")
+	}
+
 	e, err := engine.Start(ctx,
 		engine.WithConcurrency(uint8(*concurrency)),
 		engine.WithDecoders(decoders.DefaultDecoders()...),
@@ -469,7 +473,7 @@ func run(state overseer.State) {
 			Secret:        *s3ScanSecret,
 			SessionToken:  *s3ScanSessionToken,
 			Buckets:       *s3ScanBuckets,
-			Roles:	       *s3ScanRoleArns,
+			Roles:         *s3ScanRoleArns,
 			CloudCred:     *s3ScanCloudEnv,
 			MaxObjectSize: int64(*s3ScanMaxObjectSize),
 		}
@@ -525,10 +529,6 @@ func run(state overseer.State) {
 		}
 	}
 
-	if !*jsonLegacy && !*jsonOut {
-		fmt.Fprintf(os.Stderr, "🐷🔑🐷  TruffleHog. Unearth your secrets. 🐷🔑🐷\n\n")
-	}
-
 	// Wait for all workers to finish.
 	if err = e.Finish(ctx); err != nil {
 		logFatal(err, "engine failed to finish execution")

pkg/detectors/aws/aws.go

@@ -208,11 +208,10 @@ func (s scanner) verifyMatch(ctx context.Context, resIDMatch, resSecretMatch str
 
 	res, err := client.Do(req)
 	if err == nil {
-
+		defer res.Body.Close()
 		if res.StatusCode >= 200 && res.StatusCode < 300 {
 			identityInfo := identityRes{}
 			err := json.NewDecoder(res.Body).Decode(&identityInfo)
-			res.Body.Close()
 			if err == nil {
 				extraData := map[string]string{
 					"account": identityInfo.GetCallerIdentityResponse.GetCallerIdentityResult.Account,
@@ -244,7 +243,6 @@ func (s scanner) verifyMatch(ctx context.Context, resIDMatch, resSecretMatch str
 			}
 			var body awsErrorResponseBody
 			err = json.NewDecoder(res.Body).Decode(&body)
-			res.Body.Close()
 			if err == nil {
 				// All instances of the code I've seen in the wild are PascalCased but this check is
 				// case-insensitive out of an abundance of caution

pkg/detectors/detectors.go

@@ -66,7 +66,7 @@ type ResultWithMetadata struct {
 	// SourceMetadata contains source-specific contextual information.
 	SourceMetadata *source_metadatapb.MetaData
 	// SourceID is the ID of the source that the API uses to map secrets to specific sources.
-	SourceID int64
+	SourceID sources.SourceID
 	// SourceType is the type of Source.
 	SourceType sourcespb.SourceType
 	// SourceName is the name of the Source.

pkg/detectors/figmapersonalaccesstoken/figmapersonalaccesstoken.go

@@ -15,6 +15,9 @@ type Scanner struct{}
 
 // Ensure the Scanner satisfies the interface at compile time.
 var _ detectors.Detector = (*Scanner)(nil)
+var _ detectors.Versioner = (*Scanner)(nil)
+
+func (Scanner) Version() int { return 1 }
 
 var (
 	client = common.SaneHttpClient()

pkg/detectors/figmapersonalaccesstoken_v2/figmapersonalaccesstoken_v2.go

@@ -0,0 +1,77 @@
+package figmapersonalaccesstoken
+
+import (
+	"context"
+	"net/http"
+	"regexp"
+	"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{}
+
+// Ensure the Scanner satisfies the interface at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+var _ detectors.Versioner = (*Scanner)(nil)
+
+func (Scanner) Version() int { return 2 }
+
+var (
+	client = common.SaneHttpClient()
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"figma"}) + `\b(fig[d|((u|o)(r|h)?)]_[a-z0-9A-Z_-]{40})\b`)
+)
+
+// 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{"figma"}
+}
+
+// FromData will find and optionally verify FigmaPersonalAccessToken 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)
+
+	matches := keyPat.FindAllStringSubmatch(dataStr, -1)
+
+	for _, match := range matches {
+		if len(match) != 2 {
+			continue
+		}
+		resMatch := strings.TrimSpace(match[1])
+
+		s1 := detectors.Result{
+			DetectorType: detectorspb.DetectorType_FigmaPersonalAccessToken,
+			Raw:          []byte(resMatch),
+		}
+
+		if verify {
+			req, err := http.NewRequestWithContext(ctx, "GET", "https://api.figma.com/v1/me", nil)
+			if err != nil {
+				continue
+			}
+			req.Header.Add("X-Figma-Token", resMatch)
+			res, err := client.Do(req)
+			if err == nil {
+				defer res.Body.Close()
+				if res.StatusCode >= 200 && res.StatusCode < 300 {
+					s1.Verified = true
+				} else {
+					if detectors.IsKnownFalsePositive(resMatch, detectors.DefaultFalsePositives, true) {
+						continue
+					}
+				}
+			}
+		}
+
+		results = append(results, s1)
+	}
+
+	return results, nil
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_FigmaPersonalAccessToken
+}

pkg/detectors/figmapersonalaccesstoken_v2/figmapersonalaccesstoken_v2_test.go

@@ -0,0 +1,120 @@
+//go:build detectors
+// +build detectors
+
+package figmapersonalaccesstoken
+
+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 TestFigmaPersonalAccessToken_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("FIGMAPERSONALACCESSTOKEN_V2_TOKEN")
+	inactiveSecret := testSecrets.MustGetField("FIGMAPERSONALACCESSTOKEN_V2_INACTIVE")
+
+	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 figmapersonalaccesstoken secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_FigmaPersonalAccessToken,
+					Verified:     true,
+				},
+			},
+			wantErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a figmapersonalaccesstoken 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_FigmaPersonalAccessToken,
+					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("FigmaPersonalAccessToken.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("FigmaPersonalAccessToken.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/gemini/gemini.go

@@ -96,7 +96,7 @@ func constructRequest(secret, keyID string) (*http.Request, error) {
 		return nil, err
 	}
 
-	params := map[string]interface{}{
+	params := map[string]any{
 		"request": accountDetail,
 		"nonce":   time.Now().UnixNano(),
 	}

pkg/detectors/mockaroo/mockaroo.go

@@ -83,8 +83,8 @@ func (s Scanner) Type() detectorspb.DetectorType {
 
 type typeRes struct {
 	Types []struct {
-		Name       string        `json:"name"`
-		Type       interface{}   `json:"type"`
-		Parameters []interface{} `json:"parameters"`
+		Name       string `json:"name"`
+		Type       any    `json:"type"`
+		Parameters []any  `json:"parameters"`
 	} `json:"types"`
 }

pkg/detectors/privatekey/cracker.go

@@ -9,7 +9,7 @@ import (
 	"golang.org/x/crypto/ssh"
 )
 
-//go:embed "rockyou-15.txt"
+//go:embed "list.txt"
 var rawCrackList []byte
 var passphrases [][]byte
 
@@ -21,17 +21,17 @@ var (
 	ErrUncrackable = errors.New("unable to crack encryption")
 )
 
-func crack(in []byte) (interface{}, error) {
+func crack(in []byte) (any, string, error) {
 	for _, passphrase := range passphrases {
 		parsed, err := ssh.ParseRawPrivateKeyWithPassphrase(in, passphrase)
 		if err != nil {
 			if errors.Is(err, x509.IncorrectPasswordError) {
 				continue
 			} else {
-				return nil, err
+				return nil, "", err
 			}
 		}
-		return parsed, nil
+		return parsed, string(passphrase), nil
 	}
-	return nil, ErrUncrackable
+	return nil, "", ErrUncrackable
 }

pkg/detectors/privatekey/cracker_test.go

@@ -56,23 +56,29 @@ Hx7UPVlTK8dyvk1Z+Yw0nrfNClI=
 
 func Test_crack(t *testing.T) {
 	tests := []struct {
-		name    string
-		in      []byte
-		wantErr bool
+		name             string
+		in               []byte
+		wantedPassphrase string
+		wantErr          bool
 	}{
 		{
-			name:    "crackable",
-			wantErr: false,
-			in:      testEncryptedKey,
+			name:             "crackable",
+			wantErr:          false,
+			in:               testEncryptedKey,
+			wantedPassphrase: string(testEncryptedKeyCorrectPassword),
 		},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			_, err := crack(tt.in)
+			_, passphrase, err := crack(tt.in)
 			if (err != nil) != tt.wantErr {
 				t.Errorf("crack() error = %v, wantErr %v", err, tt.wantErr)
 				return
 			}
+
+			if passphrase != string(tt.wantedPassphrase) {
+				t.Errorf("crack() passphrase = %v, want %v", passphrase, string(testEncryptedKeyCorrectPassword))
+			}
 		})
 	}
 }

pkg/detectors/privatekey/fingerprint.go

@@ -5,10 +5,12 @@ import (
 	"crypto/ed25519"
 	"crypto/rsa"
 	"crypto/sha1"
+	"crypto/sha256"
 	"crypto/x509"
+	"encoding/base64"
 	"encoding/hex"
 	"errors"
-	"strings"
+	"fmt"
 
 	"golang.org/x/crypto/ssh"
 )
@@ -18,18 +20,8 @@ var (
 	ErrEncryptedKey = errors.New("key is encrypted")
 )
 
-func FingerprintPEMKey(in []byte) (string, error) {
-	parsedKey, err := ssh.ParseRawPrivateKey(in)
-	if err != nil && strings.Contains(err.Error(), "private key is passphrase protected") {
-		parsedKey, err = crack(in)
-		if err != nil {
-			return "", err
-		}
-	} else if err != nil {
-		return "", err
-	}
-
-	var pubKey interface{}
+func FingerprintPEMKey(parsedKey any) (string, error) {
+	var pubKey any
 	switch privateKey := parsedKey.(type) {
 	case *rsa.PrivateKey:
 		pubKey = &privateKey.PublicKey
@@ -44,6 +36,10 @@ func FingerprintPEMKey(in []byte) (string, error) {
 		return "", ErrNotSupported
 	}
 
+	return fingerprintPublicKey(pubKey)
+}
+
+func fingerprintPublicKey(pubKey any) (string, error) {
 	publickeyBytes, err := x509.MarshalPKIXPublicKey(pubKey)
 	if err != nil {
 		return "", err
@@ -53,3 +49,8 @@ func FingerprintPEMKey(in []byte) (string, error) {
 	publicKeyFingerprintInHex := hex.EncodeToString(publicKeyFingerprint[:])
 	return publicKeyFingerprintInHex, nil
 }
+
+func fingerprintSSHPublicKey(pubKey ssh.PublicKey) string {
+	publicKeyFingerprint := sha256.Sum256(pubKey.Marshal())
+	return fmt.Sprintf("SHA256:%s", base64.RawStdEncoding.EncodeToString(publicKeyFingerprint[:]))
+}

pkg/detectors/privatekey/privatekey.go

@@ -6,11 +6,14 @@ import (
 	"fmt"
 	"net/http"
 	"regexp"
+	"strings"
 	"time"
 
+	"github.com/AzureAD/microsoft-authentication-library-for-go/apps/errors"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+	"golang.org/x/crypto/ssh"
 )
 
 type Scanner struct {
@@ -53,19 +56,67 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) ([]dete
 			Redacted:     token[0:64],
 		}
 
-		fingerprint, err := FingerprintPEMKey([]byte(token))
+		secret.ExtraData = make(map[string]string)
+
+		var passphrase string
+		parsedKey, err := ssh.ParseRawPrivateKey([]byte(token))
+		if err != nil && strings.Contains(err.Error(), "private key is passphrase protected") {
+			secret.ExtraData["encrypted"] = "true"
+			parsedKey, passphrase, err = crack([]byte(token))
+			if err != nil {
+				secret.VerificationError = err
+				continue
+			}
+			if passphrase != "" {
+				secret.ExtraData["cracked_encryption_passphrase"] = "true"
+			}
+		} else if err != nil {
+			// couldn't parse key, probably invalid
+			continue
+		}
+
+		fingerprint, err := FingerprintPEMKey(parsedKey)
 		if err != nil {
 			continue
 		}
 
 		if verify {
+			verificationErrors := []string{}
 			data, err := lookupFingerprint(fingerprint, s.IncludeExpired)
 			if err == nil {
-				secret.StructuredData = data
 				if data != nil {
 					secret.Verified = true
+					secret.ExtraData["certificate_urls"] = strings.Join(data.CertificateURLs, ", ")
 				}
+			} else {
+				verificationErrors = append(verificationErrors, err.Error())
 			}
+
+			user, err := verifyGitHubUser(parsedKey)
+			if err != nil && !errors.Is(err, errPermissionDenied) {
+				verificationErrors = append(verificationErrors, err.Error())
+			}
+			if user != nil {
+				secret.Verified = true
+				secret.ExtraData["github_user"] = *user
+			}
+
+			user, err = verifyGitLabUser(parsedKey)
+			if err != nil && !errors.Is(err, errPermissionDenied) {
+				verificationErrors = append(verificationErrors, err.Error())
+			}
+			if user != nil {
+				secret.Verified = true
+				secret.ExtraData["gitlab_user"] = *user
+			}
+
+			if !secret.Verified && len(verificationErrors) > 0 {
+				secret.VerificationError = fmt.Errorf("verification failures: %s", strings.Join(verificationErrors, ", "))
+			}
+		}
+
+		if len(secret.ExtraData) == 0 {
+			secret.ExtraData = nil
 		}
 
 		results = append(results, secret)
@@ -74,23 +125,30 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) ([]dete
 	return results, nil
 }
 
-func lookupFingerprint(publicKeyFingerprintInHex string, includeExpired bool) (data *detectorspb.StructuredData, err error) {
+type result struct {
+	CertificateURLs []string
+	GitHubUsername  string
+}
+
+func lookupFingerprint(publicKeyFingerprintInHex string, includeExpired bool) (*result, error) {
 	req, err := http.NewRequest("GET", fmt.Sprintf("https://keychecker.trufflesecurity.com/fingerprint/%s", publicKeyFingerprintInHex), nil)
 	if err != nil {
-		return
+		return nil, err
 	}
 	res, err := client.Do(req)
 	if err != nil {
-		return
+		return nil, err
 	}
 	defer res.Body.Close()
 
 	results := DriftwoodResult{}
 	err = json.NewDecoder(res.Body).Decode(&results)
 	if err != nil {
-		return
+		return nil, err
 	}
 
+	var data *result
+
 	seen := map[string]struct{}{}
 	for _, r := range results.CertificateResults {
 		if _, ok := seen[r.CertificateFingerprint]; ok {
@@ -100,36 +158,13 @@ func lookupFingerprint(publicKeyFingerprintInHex string, includeExpired bool) (d
 			continue
 		}
 		if data == nil {
-			data = &detectorspb.StructuredData{}
+			data = &result{}
 		}
-		if data.TlsPrivateKey == nil {
-			data.TlsPrivateKey = make([]*detectorspb.TlsPrivateKey, 0)
-		}
-		data.TlsPrivateKey = append(data.TlsPrivateKey, &detectorspb.TlsPrivateKey{
-			CertificateFingerprint: r.CertificateFingerprint,
-			ExpirationTimestamp:    r.ExpirationTimestamp.Unix(),
-			VerificationUrl:        fmt.Sprintf("https://crt.sh/?q=%s", r.CertificateFingerprint),
-		})
+		data.CertificateURLs = append(data.CertificateURLs, fmt.Sprintf("https://crt.sh/?q=%s", r.CertificateFingerprint))
 		seen[r.CertificateFingerprint] = struct{}{}
 	}
 
-	for _, r := range results.GitHubSSHResults {
-		if _, ok := seen[r.Username]; ok {
-			continue
-		}
-		if data == nil {
-			data = &detectorspb.StructuredData{}
-		}
-		if data.GithubSshKey == nil {
-			data.GithubSshKey = make([]*detectorspb.GitHubSSHKey, 0)
-		}
-		data.GithubSshKey = append(data.GithubSshKey, &detectorspb.GitHubSSHKey{
-			User: r.Username,
-		})
-		seen[r.Username] = struct{}{}
-	}
-
-	return
+	return data, nil
 }
 
 type DriftwoodResult struct {

pkg/detectors/privatekey/privatekey_test.go

@@ -20,13 +20,15 @@ import (
 func TestPrivatekey_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
-	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors2")
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")
 	if err != nil {
 		t.Fatalf("could not get test secrets from GCP: %s", err)
 	}
 	secretTLS := testSecrets.MustGetField("PRIVATEKEY_TLS")
 	secretGitHub := testSecrets.MustGetField("PRIVATEKEY_GITHUB")
+	secretGitHubEncrypted := testSecrets.MustGetField("PRIVATEKEY_GITHUB_ENCRYPTED")
 	secretInactive := testSecrets.MustGetField("PRIVATEKEY_UNVERIFIED")
+
 	type args struct {
 		ctx    context.Context
 		data   []byte
@@ -75,19 +77,8 @@ func TestPrivatekey_FromChunk(t *testing.T) {
 					DetectorType: detectorspb.DetectorType_PrivateKey,
 					Verified:     true,
 					Redacted:     "-----BEGIN PRIVATE KEY-----\nMIIEvgIBADANBgkqhkiG9w0BAQEFAASCBKgw",
-					StructuredData: &detectorspb.StructuredData{
-						TlsPrivateKey: []*detectorspb.TlsPrivateKey{
-							{
-								CertificateFingerprint: "1e20c40deb44a8539dd3ac3e8c53b72750cb19f9",
-								VerificationUrl:        "https://crt.sh/?q=1e20c40deb44a8539dd3ac3e8c53b72750cb19f9",
-								ExpirationTimestamp:    1497337731,
-							},
-							{
-								CertificateFingerprint: "0e9de31fb2ee16465a4d5d93b227d54f870326d1",
-								VerificationUrl:        "https://crt.sh/?q=0e9de31fb2ee16465a4d5d93b227d54f870326d1",
-								ExpirationTimestamp:    1528192832,
-							},
-						},
+					ExtraData: map[string]string{
+						"certificate_urls": "https://crt.sh/?q=1e20c40deb44a8539dd3ac3e8c53b72750cb19f9, https://crt.sh/?q=0e9de31fb2ee16465a4d5d93b227d54f870326d1",
 					},
 				},
 			},
@@ -106,12 +97,8 @@ func TestPrivatekey_FromChunk(t *testing.T) {
 					DetectorType: detectorspb.DetectorType_PrivateKey,
 					Verified:     true,
 					Redacted:     "-----BEGIN OPENSSH PRIVATE KEY-----\nb3BlbnNzaC1rZXktdjEAAAAABG5v",
-					StructuredData: &detectorspb.StructuredData{
-						GithubSshKey: []*detectorspb.GitHubSSHKey{
-							{
-								User: "thisisforgithub0",
-							},
-						},
+					ExtraData: map[string]string{
+						"github_user": "sirdetectsalot",
 					},
 				},
 			},
@@ -121,47 +108,8 @@ func TestPrivatekey_FromChunk(t *testing.T) {
 			name: "found encrypted GitHub SSH private key, verified",
 			s:    Scanner{},
 			args: args{
-				ctx: context.Background(),
-				data: []byte(`-----BEGIN OPENSSH PRIVATE KEY-----
-						b3BlbnNzaC1rZXktdjEAAAAACmFlczI1Ni1jdHIAAAAGYmNyeXB0AAAAGAAAABAjNIZuun
-						xgLkM8KuzfmQuRAAAAEAAAAAEAAAGXAAAAB3NzaC1yc2EAAAADAQABAAABgQDe3Al0EMPz
-						utVNk5DixaYrGMK56RqUoqGBinke6SWVWmqom1lBcJWzor6HlnMRPPr7YCEsJKL4IpuVwu
-						inRa5kdtNTyM7yyQTSR2xXCS0fUItNuq8pUktsH8VUggpMeew8hJv7rFA7tnIg3UXCl6iF
-						OLZKbDA5aa24idpcD8b1I9/RzTOB1fu0of5xd9vgODzGw5JvHQSJ0FaA42aNBMGwrDhDB3
-						sgnRNdWf6NNIh8KpXXMKJADf3klsyn6He8L2bPMp8a4wwys2YB35p5zQ0JURovsdewlOxH
-						NT7eP19eVf4dCreibxUmRUaob5DEoHEk8WrxjKWIYUuLeD6AfcW6oXyRU2Yy8Vrt6SqFl5
-						WAi47VMFTkDZYS/eCvG53q9UBHpCj7Qvb0vSkCZXBvBIhlw193F3PX4WvO1IXsMwvQ1D1X
-						lmomsItbqM0cJyKw6LU18QWiBHvE7BqcphaoL5E08W2ATTSRIMCp6rt4rptM7KyGK8rc6W
-						UYrCnWt6KlCA8AAAWQXk+lVx6bH5itIKKYmQr6cR/5xtZ2GHAxnYtvlW3xnGhU0MHv+lJ2
-						uoWlT2RXE5pdMUQj7rNWAMqkwifSKZs9wBfYeo1TaFDmC3nW7yHSN3XTuO78mPIW5JyvmE
-						Rj5qjsUn7fNmzECoAxnVERhwnF3KqUBEPzIAc6/7v/na9NTiiGaJPco9lvCoPWbVLN08WG
-						SuyU+0x5zc3ebzuPcYqu5/c5nmiGxhALrIhjIS0OV1mtAAFhvdMjMIHOijOzSKVCC7rRk5
-						kG9EMLNvOn/DUVSRHamw5gs2V3V+Zq2g5nYWfgq8aDSTB8XlIzOj1cz3HwfN6pfSNQ/3Qe
-						wOQfWfTWdO+JSL8aoBN5Wg8tDbgmvmbFrINsJfFfSm0wZgcHhC7Ul4U3v4c8PoNdK9HXwi
-						TKKzJ9nxLYb+vDh50cnkseu2gt0KwVpjIorxEqeK755mKPao3JmOMr6uFTQsb+g+ZNgPwl
-						nRHA4Igx+zADFj3twldnKIiRpBQ5J4acur3uQ+saanBTXgul1TiFiUGT2cnz+IiCsdPovg
-						TAMt868W5LmzpfH4Cy54JtaRC4/UuMnkTGbWgutVDnWj2stOAzsQ1YmhH5igUmc94mUL+W
-						8vQDCKpeI8n+quDS9zxTvy4L4H5Iz7OZlh0h6N13BDvCYXKcNF/ugkfxZbu8mZsZQQzXNR
-						wOrEtKoHc4AnXYNzsuHEoEyLyJxGfFRDSTLbyN9wFOS/c0k9Gjte+kQRZjBVGORE5sN6X3
-						akUnTF76RhbEc+LamrwM1h5340bwosRbR8I+UrsQdFfJBEj1ZSyMRJlMkFUNi6blt7bhyx
-						ea+Pm2A614nlYUBjw2KKzzn8N/0H2NpJjIptvDsbrx3BS/rKwOeJwavRrGnIlEzuAag4vx
-						Zb2TPVta45uz7fQP5IBl83b0BJKI5Zv/fniUeLI78W/UsZqb64YQbfRyBzFtI1T/SsCi0B
-						e0EyKMzbxtSceT1Mb8eJiVIq04Xpwez9fIUt5rSedZD8KPq8P6s0cGsR7Qmw6eXZ/dBR/a
-						s5vPhfIUmQawmnwAVuWNRdQQ79jUBSn5M+ZRVVTgEG+vFyvxr/bZqOo1JCoq5BmQhLWGRJ
-						Dk9TolbeFIVFrkuXkcu99a079ux7XSkON64oPzHrcsEzjPA1GPqs9CGBSO16wq/nI3zg+E
-						kcOCaurc9yHJJPwduem0+8WLX3WoGNfQRKurtQze2ppy8KarEtDhDd96sKkhYaqOg3GOX8
-						Yx827L4vuWSJSIqKuO2kH6kOCMUNO16piv0z/8u3CJxOGh9+4FZIop81fiFTKLhV3/gwLm
-						fzFY++KIZrLfZcUjzd80NNEja69F452Eb9HrI5BurN/PznDEi9bzM598Y7beyl4/kd4R2e
-						S7SW9/LOrGw5UgxtiU+kV8nPz1PdgxO4sRlnntSBEwkQBzMkLOpq2h2BuJ2TlMP/TWuwLQ
-						sDkv1Yk1pD0roGmtMzbujnURGxqRJ8gUmuIot4hpfyRSssvnRQQZ3lQCQCwHiE+HJxXWf5
-						c58zOMjW7o21tI8e13uUnbRoQVJM9XYqk1usPXIkYPYL9uOw3AW/Zn+cnDrsXvTK9ZxgGD
-						/90b1BNwVqMlUK+QggHNwl5qD8eoXK5cDvav66te+E+V7FYFQ06w3tytRVz8SjoaiChN02
-						muIjvl6G7Hoj1hObM2t/ZheN1EShS11z868hhS6Mx7GvIdtkXuvdiBYMiBLOshJQxB8Mzx
-						iug9W+Di3upLf0UMC1TqADGphsIHRU7RbmHQ8Rwp7dogswmDfpRSapPt9p0D+6Ad5VBzi3
-						f3BPXj76UBLMEJCrZR1P28vnAA7AyNHaLvMPlWDMG5v3V/UV+ugyFcoBAOyjiQgYST8F3e
-						Hx7UPVlTK8dyvk1Z+Yw0nrfNClI=
-						-----END OPENSSH PRIVATE KEY-----
-		`),
+				ctx:    context.Background(),
+				data:   []byte(secretGitHubEncrypted),
 				verify: true,
 			},
 			want: []detectors.Result{
@@ -169,12 +117,10 @@ func TestPrivatekey_FromChunk(t *testing.T) {
 					DetectorType: detectorspb.DetectorType_PrivateKey,
 					Verified:     true,
 					Redacted:     "-----BEGIN OPENSSH PRIVATE KEY-----\nb3BlbnNzaC1rZXktdjEAAAAACmFl",
-					StructuredData: &detectorspb.StructuredData{
-						GithubSshKey: []*detectorspb.GitHubSSHKey{
-							{
-								User: "thisisforgithub0",
-							},
-						},
+					ExtraData: map[string]string{
+						"github_user":                   "sirdetectsalot",
+						"encrypted":                     "true",
+						"cracked_encryption_passphrase": "true",
 					},
 				},
 			},
@@ -262,7 +208,7 @@ func Test_lookupFingerprint(t *testing.T) {
 				t.Errorf("lookupFingerprint() error = %v, wantErr %v", err, tt.wantErr)
 				return
 			}
-			if !reflect.DeepEqual(gotFingerprints != nil && len(gotFingerprints.TlsPrivateKey) > 0, tt.wantFingerprints) {
+			if !reflect.DeepEqual(gotFingerprints != nil && len(gotFingerprints.CertificateURLs) > 0, tt.wantFingerprints) {
 				t.Errorf("lookupFingerprint() = %v, want %v", gotFingerprints, tt.wantFingerprints)
 			}
 		})

pkg/detectors/privatekey/ssh.go

@@ -0,0 +1,125 @@
+package privatekey
+
+import (
+	"bytes"
+	"errors"
+	"fmt"
+	"net"
+	"strings"
+	"time"
+
+	"golang.org/x/crypto/ssh"
+)
+
+// https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/githubs-ssh-key-fingerprints
+var githubFingerprints = map[string]string{
+	"SHA256:uNiVztksCsDhcc0u9e8BujQXVUpKZIDTMczCvj3tD2s": "RSA",
+	"SHA256:br9IjFspm1vxR3iA35FWE+4VTyz1hYVLIE2t1/CeyWQ": "DSA - deprecated",
+	"SHA256:p2QAMXNIC1TJYWeIOttrVc98/R1BUFWu3/LiyKgUfQM": "ECDSA",
+	"SHA256:+DiY3wvvV6TuJJhbpZisF/zLDA0zPMSvHdkr4UvCOqU": "ED25519",
+}
+
+// https://docs.gitlab.com/ee/user/gitlab_com/index.html#ssh-host-keys-fingerprints
+var gitlabFingerprints = map[string]string{
+	"SHA256:HbW3g8zUjNSksFbqTiUWPWg2Bq1x8xdGUrliXFzSnUw": "ECDSA",
+	"SHA256:eUXGGm1YGsMAS7vkcx6JOJdOGHPem5gQp4taiCfCLB8": "ED25519",
+	"SHA256:ROQFvPThGrW4RuWLoL9tq9I9zJ42fK4XywyRtbOz/EQ": "RSA",
+}
+
+func firstResponseFromSSH(parsedKey any, username, hostport string) (string, error) {
+	signer, err := ssh.NewSignerFromKey(parsedKey)
+	if err != nil {
+		return "", err
+	}
+
+	// Verify the server fingerprint to ensure that there is no MITM replay attack
+	config := &ssh.ClientConfig{
+		Timeout: 3 * time.Second,
+		User:    username,
+		Auth: []ssh.AuthMethod{
+			ssh.PublicKeys(signer),
+		},
+		HostKeyCallback: func(hostname string, _ net.Addr, key ssh.PublicKey) error {
+			switch hostname {
+			case "github.com:22":
+				fingerprint := fingerprintSSHPublicKey(key)
+				if _, ok := githubFingerprints[fingerprint]; !ok {
+					return fmt.Errorf("unknown host fingerprint for github.com, got %s", fingerprint)
+				}
+			case "gitlab.com:22":
+				fingerprint := fingerprintSSHPublicKey(key)
+				if _, ok := gitlabFingerprints[fingerprint]; !ok {
+					return fmt.Errorf("unknown host fingerprint for gitlab.com, got %s", fingerprint)
+				}
+			default:
+				return errors.New("unknown host in fingerprint db")
+			}
+			return nil
+		},
+	}
+
+	client, err := ssh.Dial("tcp", hostport, config)
+	if err != nil {
+		if strings.Contains(err.Error(), "unable to authenticate") {
+			return "", errPermissionDenied
+		}
+		return "", err
+	}
+	defer client.Close()
+
+	session, err := client.NewSession()
+	if err != nil {
+		return "", err
+	}
+	defer session.Close()
+
+	var output bytes.Buffer
+	session.Stderr = &output
+
+	err = session.Shell()
+	if err != nil {
+		return "", err
+	}
+	_ = session.Wait()
+
+	return output.String(), err
+}
+
+var errPermissionDenied = errors.New("permission denied")
+
+func verifyGitHubUser(parsedKey any) (*string, error) {
+	output, err := firstResponseFromSSH(parsedKey, "git", "github.com:22")
+	if err != nil {
+		return nil, err
+	}
+
+	if strings.Contains(output, "Permission denied") {
+		return nil, errPermissionDenied
+	}
+
+	if strings.Contains(output, "successfully authenticated") {
+		username := strings.TrimSuffix(strings.Split(output, " ")[1], "!")
+		return &username, nil
+	}
+
+	return nil, nil
+}
+
+func verifyGitLabUser(parsedKey any) (*string, error) {
+	output, err := firstResponseFromSSH(parsedKey, "git", "gitlab.com:22")
+	if err != nil {
+		return nil, err
+	}
+
+	if strings.Contains(output, "Permission denied") {
+		return nil, errPermissionDenied
+	}
+
+	if strings.Contains(output, "Welcome to GitLab") {
+		split := strings.Split(output, " ")
+		username := strings.TrimPrefix(strings.TrimSuffix(split[len(split)-1], "!"), "@")
+		return &username, nil
+	}
+
+	return nil, nil
+}

pkg/detectors/privatekey/ssh_test.go

@@ -0,0 +1,34 @@
+package privatekey
+
+import (
+	"context"
+	"testing"
+	"time"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"golang.org/x/crypto/ssh"
+)
+
+func TestFirstResponseFromSSH(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors4")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+	secretGitHub := testSecrets.MustGetField("PRIVATEKEY_GITHUB")
+
+	parsedKey, err := ssh.ParseRawPrivateKey([]byte(normalize(secretGitHub)))
+	if err != nil {
+		t.Fatalf("could not parse test secret: %s", err)
+	}
+
+	output, err := firstResponseFromSSH(parsedKey, "git", "github.com:22")
+	if err != nil {
+		t.Fail()
+	}
+
+	if len(output) == 0 {
+		t.Fail()
+	}
+}

pkg/detectors/rubygems/rubygems.go

@@ -47,7 +47,7 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 		}
 
 		if verify {
-			req, err := http.NewRequestWithContext(ctx, "GET", "https://rubygems.org/api/v1/gems", nil)
+			req, err := http.NewRequestWithContext(ctx, "GET", "https://rubygems.org/api/v1/gems.json", nil)
 			if err != nil {
 				continue
 			}
@@ -56,7 +56,7 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 			res, err := client.Do(req)
 			if err == nil {
 				defer res.Body.Close()
-				if res.StatusCode >= 200 && res.StatusCode < 300 {
+				if res.StatusCode >= 200 && res.StatusCode < 300 || res.StatusCode == http.StatusForbidden {
 					s1.Verified = true
 				} else {
 					// 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.

pkg/detectors/sendbirdorganizationapi/sendbirdorganizationapi.go

@@ -2,6 +2,7 @@ package sendbirdorganizationapi
 
 import (
 	"context"
+	"fmt"
 	"net/http"
 	"regexp"
 	"strings"
@@ -11,13 +12,15 @@ import (
 	"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{"sendbird"}) + `\b([0-9a-f]{24})\b`)
@@ -47,25 +50,33 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 		}
 
 		if verify {
+			client := s.client
+			if client == nil {
+				client = defaultClient
+			}
+
 			req, err := http.NewRequestWithContext(ctx, "GET", "https://gate.sendbird.com/api/v2/applications", nil)
 			if err != nil {
-				continue
+				s1.VerificationError = err
 			}
 			req.Header.Add("SENDBIRDORGANIZATIONAPITOKEN", resMatch)
 			res, err := client.Do(req)
-			if err == nil {
+			if err != nil {
+				s1.VerificationError = err
+			} else {
 				defer res.Body.Close()
 				if res.StatusCode >= 200 && res.StatusCode < 300 {
 					s1.Verified = true
-				} else {
-					// 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 detectors.IsKnownFalsePositive(resMatch, detectors.DefaultFalsePositives, true) {
-						continue
-					}
+				} else if res.StatusCode != http.StatusForbidden {
+					s1.VerificationError = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
 				}
 			}
 		}
 
+		// 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(resMatch, detectors.DefaultFalsePositives, true) {
+			continue
+		}
 		results = append(results, s1)
 	}
 

pkg/detectors/sendbirdorganizationapi/sendbirdorganizationapi_test.go

@@ -9,7 +9,8 @@ import (
 	"testing"
 	"time"
 
-	"github.com/kylelemons/godebug/pretty"
+	"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"
@@ -19,7 +20,7 @@ import (
 func TestSendbirdOrganizationAPI_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
-	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors3")
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors5")
 	if err != nil {
 		t.Fatalf("could not get test secrets from GCP: %s", err)
 	}
@@ -32,11 +33,12 @@ func TestSendbirdOrganizationAPI_FromChunk(t *testing.T) {
 		verify bool
 	}
 	tests := []struct {
-		name    string
-		s       Scanner
-		args    args
-		want    []detectors.Result
-		wantErr bool
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
 	}{
 		{
 			name: "found, verified",
@@ -52,7 +54,8 @@ func TestSendbirdOrganizationAPI_FromChunk(t *testing.T) {
 					Verified:     true,
 				},
 			},
-			wantErr: false,
+			wantErr:             false,
+			wantVerificationErr: false,
 		},
 		{
 			name: "found, unverified",
@@ -68,7 +71,42 @@ func TestSendbirdOrganizationAPI_FromChunk(t *testing.T) {
 					Verified:     false,
 				},
 			},
-			wantErr: false,
+			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 sendbird secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SendbirdOrganizationAPI,
+					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 sendbird secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SendbirdOrganizationAPI,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
 		},
 		{
 			name: "not found",
@@ -84,20 +122,23 @@ func TestSendbirdOrganizationAPI_FromChunk(t *testing.T) {
 	}
 	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)
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
 			if (err != nil) != tt.wantErr {
 				t.Errorf("SendbirdOrganizationAPI.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 (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Fatalf("wantVerificationError = %v, verification error = %v", tt.wantVerificationErr, got[i].VerificationError)
+				}
 			}
-			if diff := pretty.Compare(got, tt.want); diff != "" {
-				t.Errorf("SendbirdOrganizationAPI.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
+			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "RawV2", "VerificationError")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Sendbird.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
 			}
 		})
 	}

pkg/detectors/sentrytoken/sentrytoken.go

@@ -122,6 +122,9 @@ func verifyToken(ctx context.Context, client *http.Client, token string) (bool,
 	if err = json.Unmarshal(bytes, &resp); err != nil {
 		return false, err
 	}
+	if len(resp) == 0 {
+		return false, fmt.Errorf("unexpected response body: %s", string(bytes))
+	}
 
 	return isVerified, err
 }

pkg/detectors/sentrytoken/sentrytoken_test.go

@@ -123,6 +123,40 @@ func TestSentryToken_FromChunk(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "found, account deactivated",
+			s:    Scanner{client: common.ConstantResponseHttpClient(200, reponseAccountDeactivated)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a sentry super secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SentryToken,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "found, account deactivated",
+			s:    Scanner{client: common.ConstantResponseHttpClient(200, responseEnmpty)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a sentry super secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SentryToken,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
 		{
 			name: "not found",
 			s:    Scanner{},
@@ -159,7 +193,8 @@ func TestSentryToken_FromChunk(t *testing.T) {
 	}
 }
 
-const responseBody403 = `
+const (
+	responseBody403 = `
 [
   {
     "organization": {
@@ -174,6 +209,9 @@ const responseBody403 = `
   }
 ]
 `
+	reponseAccountDeactivated = `{"detail": "Authentication credentials were not provided"}`
+	responseEnmpty            = `[]`
+)
 
 func BenchmarkFromData(benchmark *testing.B) {
 	ctx := context.Background()

pkg/detectors/slack/slack.go

@@ -13,13 +13,16 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
 )
 
-type Scanner struct{}
+type Scanner struct {
+	client *http.Client
+}
 
 // Check that the Slack scanner implements the SecretScanner interface at compile time.
 var _ detectors.Detector = Scanner{}
 
 var (
-	tokenPats = map[string]*regexp.Regexp{
+	defaultClient = common.SaneHttpClient()
+	tokenPats     = map[string]*regexp.Regexp{
 		"Slack Bot Token":               regexp.MustCompile(`xoxb\-[0-9]{10,13}\-[0-9]{10,13}[a-zA-Z0-9\-]*`),
 		"Slack User Token":              regexp.MustCompile(`xoxp\-[0-9]{10,13}\-[0-9]{10,13}[a-zA-Z0-9\-]*`),
 		"Slack Workspace Access Token":  regexp.MustCompile(`xoxa\-[0-9]{10,13}\-[0-9]{10,13}[a-zA-Z0-9\-]*`),
@@ -53,34 +56,54 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 		tokens := tokenPat.FindAllString(dataStr, -1)
 
 		for _, token := range tokens {
-			s := detectors.Result{
+			s1 := detectors.Result{
 				DetectorType: detectorspb.DetectorType_Slack,
 				Raw:          []byte(token),
 			}
 			if verify {
-				client := common.SaneHttpClient()
+				client := s.client
+				if s.client == nil {
+					client = defaultClient
+				}
+
 				req, err := http.NewRequestWithContext(ctx, "POST", verifyURL, nil)
 				if err != nil {
 					continue
 				}
+
 				req.Header.Add("Content-Type", "application/json; charset=utf-8")
 				req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", token))
+
 				res, err := client.Do(req)
 				if err == nil {
 					defer res.Body.Close()
 					var authResponse authRes
 					if err := json.NewDecoder(res.Body).Decode(&authResponse); err != nil {
-						continue
+						s1.VerificationError = fmt.Errorf("failed to decode auth response: %w", err)
 					}
-					s.Verified = authResponse.Ok
+
+					if authResponse.Ok {
+						s1.Verified = true
+						// Slack API returns 200 even if the token is invalid. We need to check the error field.
+					} else if authResponse.Error == "invalid_auth" {
+						// The secret is determinately not verified (nothing to do)
+					} else {
+						s1.VerificationError = fmt.Errorf("unexpected error auth response %+v", authResponse.Error)
+					}
+				} else {
+					s1.VerificationError = err
 				}
 			}
 
-			results = append(results, s)
+			if !s1.Verified && detectors.IsKnownFalsePositive(string(s1.Raw), detectors.DefaultFalsePositives, true) {
+				continue
+			}
+
+			results = append(results, s1)
 		}
 	}
 
-	return
+	return results, nil
 }
 
 func (s Scanner) Type() detectorspb.DetectorType {

pkg/detectors/slack/slack_test.go

@@ -9,13 +9,15 @@ import (
 	"testing"
 	"time"
 
-	"github.com/kylelemons/godebug/pretty"
+	"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/pb/detectorspb"
 )
 
-func TestScanner_FromChunk(t *testing.T) {
+func TestSlack_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
 	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors2")
@@ -24,17 +26,28 @@ func TestScanner_FromChunk(t *testing.T) {
 	}
 	secret := testSecrets.MustGetField("SLACK")
 	secretInactive := testSecrets.MustGetField("SLACK_INACTIVE")
+
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+
 	tests := []struct {
-		name        string
-		data        []byte
-		verify      bool
-		wantResults []detectors.Result
-		wantErr     bool
+		name                string
+		s                   Scanner
+		args                args
+		wantResults         []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
 	}{
 		{
-			name:   "found, verified",
-			data:   []byte(fmt.Sprintf("You can find a slack secret %s within", secret)),
-			verify: true,
+			name: "found, verified",
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slack secret %s within", secret)),
+				verify: true,
+			},
 			wantResults: []detectors.Result{
 				{
 					DetectorType: detectorspb.DetectorType_Slack,
@@ -44,9 +57,12 @@ func TestScanner_FromChunk(t *testing.T) {
 			wantErr: false,
 		},
 		{
-			name:   "found but unverified",
-			data:   []byte(fmt.Sprintf("You can find a slack secret %s within", secretInactive)),
-			verify: true,
+			name: "found but unverified",
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slack secret %s within", secretInactive)),
+				verify: true,
+			},
 			wantResults: []detectors.Result{
 				{
 					DetectorType: detectorspb.DetectorType_Slack,
@@ -55,23 +71,62 @@ func TestScanner_FromChunk(t *testing.T) {
 			},
 			wantErr: 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 slack secret %s within", secret)),
+				verify: true,
+			},
+			wantResults: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Slack,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
+		{
+			name: "unexpected auth response",
+			s:    Scanner{client: common.ConstantResponseHttpClient(200, `{"ok": false, "error": "unexpected_error"}`)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slack secret %s within", secret)),
+				verify: true,
+			},
+			wantResults: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_Slack,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			s := Scanner{}
-			got, err := s.FromData(context.Background(), tt.verify, tt.data)
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
 			if (err != nil) != tt.wantErr {
-				t.Errorf("Scanner.FromData) error = %v, wantErr %v", err, tt.wantErr)
+				t.Errorf("Slack.FromData() error = %v, wantErr %v", err, tt.wantErr)
 				return
 			}
+
 			for i := range got {
 				if len(got[i].Raw) == 0 {
 					t.Fatal("no raw secret present")
 				}
 				got[i].Raw = nil
+
+				if (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Fatalf("wantVerificationError = %v, verification error = %v", tt.wantVerificationErr, got[i].VerificationError)
+				}
 			}
-			if diff := pretty.Compare(got, tt.wantResults); diff != "" {
-				t.Errorf("Scanner.FromData) %s diff: (-got +want)\n%s", tt.name, diff)
+			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "VerificationError")
+			if diff := cmp.Diff(got, tt.wantResults, ignoreOpts); diff != "" {
+				t.Errorf("Slack.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
 			}
 		})
 	}

pkg/detectors/slackwebhook/slackwebhook.go

@@ -1,6 +1,7 @@
 package slackwebhook
 
 import (
+	"bytes"
 	"context"
 	"fmt"
 	"io"
@@ -69,11 +70,10 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 				if err != nil {
 					continue
 				}
-				body := string(bodyBytes)
 
 				defer res.Body.Close()
 
-				if res.StatusCode >= 200 && res.StatusCode < 300 || (res.StatusCode == 400 && (strings.Contains(body, "no_text") || strings.Contains(body, "missing_text"))) {
+				if res.StatusCode >= 200 && res.StatusCode < 300 || (res.StatusCode == 400 && (bytes.Equal(bodyBytes, []byte("no_text")) || bytes.Equal(bodyBytes, []byte("missing_text")))) {
 					s1.Verified = true
 				} else if res.StatusCode == 401 || res.StatusCode == 403 {
 					// The secret is determinately not verified (nothing to do)

pkg/detectors/slackwebhook/slackwebhook_test.go

@@ -6,14 +6,14 @@ package slackwebhook
 import (
 	"context"
 	"fmt"
-	"github.com/google/go-cmp/cmp"
-	"github.com/google/go-cmp/cmp/cmpopts"
 	"testing"
 	"time"
 
-	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"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/pb/detectorspb"
 )
 
@@ -57,6 +57,40 @@ func TestSlackWebhook_FromChunk(t *testing.T) {
 			wantErr:             false,
 			wantVerificationErr: false,
 		},
+		{
+			name: "found, verified, 400 no_text",
+			s:    Scanner{client: common.ConstantResponseHttpClient(400, "no_text")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slackwebhook secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SlackWebhook,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, verified, 400 missing_text",
+			s:    Scanner{client: common.ConstantResponseHttpClient(400, "missing_text")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slackwebhook secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SlackWebhook,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
 		{
 			name: "found, unverified",
 			s:    Scanner{},
@@ -120,6 +154,23 @@ func TestSlackWebhook_FromChunk(t *testing.T) {
 			wantErr:             false,
 			wantVerificationErr: true,
 		},
+		{
+			name: "found, account disabled",
+			s:    Scanner{client: common.ConstantResponseHttpClient(400, disabledAccountResponse)},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a slackwebhook secret %s within", secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_SlackWebhook,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
@@ -144,6 +195,8 @@ func TestSlackWebhook_FromChunk(t *testing.T) {
 	}
 }
 
+const disabledAccountResponse = `missing_text_or_fallback_or_attachments`
+
 func BenchmarkFromData(benchmark *testing.B) {
 	ctx := context.Background()
 	s := Scanner{}

pkg/detectors/sonarcloud/sonarcloud.go

@@ -21,13 +21,13 @@ 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{"sonarcloud"}) + `\b([0-9a-z]{40})\b`)
+	keyPat = regexp.MustCompile(detectors.PrefixRegex([]string{"sonar"}) + `\b([0-9a-z]{40})\b`)
 )
 
 // 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{"sonarcloud"}
+	return []string{"sonar"}
 }
 
 // FromData will find and optionally verify SonarCloud secrets in a given set of bytes.

pkg/detectors/uri/uri.go

@@ -15,6 +15,7 @@ import (
 
 type Scanner struct {
 	allowKnownTestSites bool
+	client              *http.Client
 }
 
 // Ensure the Scanner satisfies the interface at compile time.
@@ -23,7 +24,7 @@ var _ detectors.Detector = (*Scanner)(nil)
 var (
 	keyPat = regexp.MustCompile(`\b(?:https?:)?\/\/[\S]{3,50}:([\S]{3,50})@[-.%\w\/:]+\b`)
 
-	client = common.SaneHttpClient()
+	defaultClient = common.SaneHttpClient()
 )
 
 // Keywords are used for efficiently pre-filtering chunks.
@@ -71,7 +72,7 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 		// Removing the path causes possible deduplication issues if some paths have basic auth and some do not.
 		rawURL.Path = ""
 
-		s := detectors.Result{
+		s1 := detectors.Result{
 			DetectorType: detectorspb.DetectorType_URI,
 			Raw:          []byte(rawURL.String()),
 			RawV2:        []byte(rawURLStr),
@@ -79,27 +80,30 @@ func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 		}
 
 		if verify {
-			s.Verified = verifyURL(ctx, parsedURL)
+			if s.client == nil {
+				s.client = defaultClient
+			}
+			s1.Verified, s1.VerificationError = verifyURL(ctx, s.client, parsedURL)
 		}
 
-		if !s.Verified {
+		if !s1.Verified {
 			// Skip unverified findings where the password starts with a `$` - it's almost certainly a variable.
 			if strings.HasPrefix(password, "$") {
 				continue
 			}
 		}
 
-		if !s.Verified && detectors.IsKnownFalsePositive(string(s.Raw), detectors.DefaultFalsePositives, false) {
+		if !s1.Verified && !s.allowKnownTestSites && detectors.IsKnownFalsePositive(string(s1.Raw), detectors.DefaultFalsePositives, false) {
 			continue
 		}
 
-		results = append(results, s)
+		results = append(results, s1)
 	}
 
 	return results, nil
 }
 
-func verifyURL(ctx context.Context, u *url.URL) bool {
+func verifyURL(ctx context.Context, client *http.Client, u *url.URL) (bool, error) {
 	// defuse most SSRF payloads
 	u.Path = strings.TrimSuffix(u.Path, "/")
 	u.RawQuery = ""
@@ -112,41 +116,41 @@ func verifyURL(ctx context.Context, u *url.URL) bool {
 
 	req, err := http.NewRequest("GET", credentialedURL, nil)
 	if err != nil {
-		return false
+		return false, err
 	}
 	req = req.WithContext(ctx)
 	credentialedRes, err := client.Do(req)
 	if err != nil {
-		return false
+		return false, err
 	}
 	credentialedRes.Body.Close()
 
 	// If the credentialed URL returns a non 2XX code, we can assume it's a false positive.
 	if credentialedRes.StatusCode < 200 || credentialedRes.StatusCode > 299 {
-		return false
+		return false, nil
 	}
 
 	time.Sleep(time.Millisecond * 10)
 
 	req, err = http.NewRequest("GET", nonCredentialedURL, nil)
 	if err != nil {
-		return false
+		return false, err
 	}
 	req = req.WithContext(ctx)
 	nonCredentialedRes, err := client.Do(req)
 	if err != nil {
-		return false
+		return false, err
 	}
 	nonCredentialedRes.Body.Close()
 
 	// If the non-credentialed URL returns a non 400-428 code and basic auth header, we can assume it's verified now.
 	if nonCredentialedRes.StatusCode >= 400 && nonCredentialedRes.StatusCode < 429 {
 		if nonCredentialedRes.Header.Get("WWW-Authenticate") != "" {
-			return true
+			return true, nil
 		}
 	}
 
-	return false
+	return false, nil
 }
 
 func (s Scanner) Type() detectorspb.DetectorType {

pkg/detectors/uri/uri_test.go

@@ -7,8 +7,10 @@ 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"
 )
@@ -20,11 +22,12 @@ func TestURI_FromChunk(t *testing.T) {
 		verify bool
 	}
 	tests := []struct {
-		name    string
-		s       Scanner
-		args    args
-		want    []detectors.Result
-		wantErr bool
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
 	}{
 		{
 			name: "found, unverified, wrong username",
@@ -77,6 +80,24 @@ func TestURI_FromChunk(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "found, 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 uri secret %s within", "https://httpwatch:pass@www.httpwatch.com/httpgallery/authentication/authenticatedimage/default.aspx")),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_URI,
+					Verified:     false,
+					Redacted:     "https://httpwatch:********@www.httpwatch.com",
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: true,
+		},
 		{
 			name: "bad scheme",
 			s:    Scanner{},
@@ -101,8 +122,8 @@ func TestURI_FromChunk(t *testing.T) {
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			s := Scanner{allowKnownTestSites: true}
-			got, err := s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			tt.s.allowKnownTestSites = true
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
 			if (err != nil) != tt.wantErr {
 				t.Errorf("URI.FromData() error = %v, wantErr %v", err, tt.wantErr)
 				return
@@ -111,8 +132,13 @@ func TestURI_FromChunk(t *testing.T) {
 			// 	return
 			// }
 			for i := range got {
+				if (got[i].VerificationError != nil) != tt.wantVerificationErr {
+					t.Errorf("URI.FromData() error = %v, wantVerificationErr %v", got[i].VerificationError, tt.wantErr)
+					return
+				}
 				got[i].Raw = nil
 				got[i].RawV2 = nil
+				got[i].VerificationError = nil
 			}
 			if diff := pretty.Compare(got, tt.want); diff != "" {
 				t.Errorf("URI.FromData() %s diff: (-got +want)\n%s", tt.name, diff)

pkg/engine/circleci.go

@@ -8,7 +8,6 @@ import (
 
 	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
-	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources/circleci"
 )
 
@@ -27,17 +26,13 @@ func (e *Engine) ScanCircleCI(ctx context.Context, token string) error {
 		return err
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - Circle CI", new(circleci.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			circleSource := circleci.Source{}
-			if err := circleSource.Init(ctx, "trufflehog - Circle CI", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			return &circleSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - Circle CI"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, circleci.SourceType)
+
+	circleSource := &circleci.Source{}
+	if err := circleSource.Init(ctx, "trufflehog - Circle CI", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	_, err = e.sourceManager.Run(ctx, sourceName, circleSource)
 	return err
 }

pkg/engine/docker.go

@@ -6,23 +6,18 @@ import (
 	"google.golang.org/protobuf/types/known/anypb"
 
 	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
-	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources/docker"
 )
 
 // ScanDocker scans a given docker connection.
 func (e *Engine) ScanDocker(ctx context.Context, conn *anypb.Any) error {
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - docker", new(docker.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			dockerSource := docker.Source{}
-			if err := dockerSource.Init(ctx, "trufflehog - docker", jobID, sourceID, true, conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			return &dockerSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - docker"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, docker.SourceType)
+
+	dockerSource := &docker.Source{}
+	if err := dockerSource.Init(ctx, sourceName, jobID, sourceID, true, conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	_, err := e.sourceManager.Run(ctx, sourceName, dockerSource)
 	return err
 }

pkg/engine/engine.go

@@ -19,6 +19,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/decoders"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/giturl"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/output"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/source_metadatapb"
@@ -239,7 +240,7 @@ func (m *Metrics) getScanDuration() time.Duration {
 func (e *Engine) DetectorAvgTime() map[string][]time.Duration {
 	logger := context.Background().Logger()
 	avgTime := map[string][]time.Duration{}
-	e.metrics.detectorAvgTime.Range(func(k, v interface{}) bool {
+	e.metrics.detectorAvgTime.Range(func(k, v any) bool {
 		key, ok := k.(string)
 		if !ok {
 			logger.Info("expected detectorAvgTime key to be a string")
@@ -532,12 +533,12 @@ func (e *Engine) detectChunk(ctx context.Context, data detectableChunk) {
 	}
 
 	for _, res := range results {
-		e.processResult(data, res)
+		e.processResult(ctx, data, res)
 	}
 	data.wgDoneFn()
 }
 
-func (e *Engine) processResult(data detectableChunk, res detectors.Result) {
+func (e *Engine) processResult(ctx context.Context, data detectableChunk, res detectors.Result) {
 	ignoreLinePresent := false
 	if SupportsLineNumbers(data.chunk.SourceType) {
 		copyChunk := data.chunk
@@ -545,8 +546,12 @@ func (e *Engine) processResult(data detectableChunk, res detectors.Result) {
 		if copyMetaData, ok := copyMetaDataClone.(*source_metadatapb.MetaData); ok {
 			copyChunk.SourceMetadata = copyMetaData
 		}
-		fragStart, mdLine := FragmentFirstLine(&copyChunk)
+		fragStart, mdLine, link := FragmentFirstLineAndLink(&copyChunk)
 		ignoreLinePresent = SetResultLineNumber(&copyChunk, &res, fragStart, mdLine)
+		if err := UpdateLink(ctx, copyChunk.SourceMetadata, link, *mdLine); err != nil {
+			ctx.Logger().Error(err, "error setting link")
+			return
+		}
 		data.chunk = copyChunk
 	}
 	if ignoreLinePresent {
@@ -565,11 +570,18 @@ func (e *Engine) notifyResults(ctx context.Context) {
 		}
 		atomic.AddUint32(&e.numFoundResults, 1)
 
+		// Dedupe results by comparing the detector type, raw result, and source metadata.
+		// We want to avoid duplicate results with different decoder types, but we also
+		// want to include duplicate results with the same decoder type.
+		// Duplicate results with the same decoder type SHOULD have their own entry in the
+		// results list, this would happen if the same secret is found multiple times.
 		key := fmt.Sprintf("%s%s%s%+v", r.DetectorType.String(), r.Raw, r.RawV2, r.SourceMetadata)
-		if _, ok := e.dedupeCache.Get(key); ok {
-			continue
+		if val, ok := e.dedupeCache.Get(key); ok {
+			if res, ok := val.(detectorspb.DecoderType); ok && res != r.DecoderType {
+				continue
+			}
 		}
-		e.dedupeCache.Add(key, struct{}{})
+		e.dedupeCache.Add(key, r.DecoderType)
 
 		if r.Verified {
 			atomic.AddUint64(&e.metrics.VerifiedSecretsFound, 1)
@@ -593,7 +605,8 @@ func SupportsLineNumbers(sourceType sourcespb.SourceType) bool {
 		sourcespb.SourceType_SOURCE_TYPE_GERRIT,
 		sourcespb.SourceType_SOURCE_TYPE_GITHUB_UNAUTHENTICATED_ORG,
 		sourcespb.SourceType_SOURCE_TYPE_PUBLIC_GIT,
-		sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM:
+		sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM,
+		sourcespb.SourceType_SOURCE_TYPE_AZURE_REPOS:
 		return true
 	default:
 		return false
@@ -618,27 +631,45 @@ func FragmentLineOffset(chunk *sources.Chunk, result *detectors.Result) (int64,
 	return lineNumber, false
 }
 
-// FragmentFirstLine returns the first line number of a fragment along with a pointer to the value to update in the
-// chunk metadata.
-func FragmentFirstLine(chunk *sources.Chunk) (int64, *int64) {
-	var fragmentStart *int64
+// FragmentFirstLineAndLink extracts the first line number and the link from the chunk metadata.
+// It returns:
+//   - The first line number of the fragment.
+//   - A pointer to the line number, facilitating direct updates.
+//   - The link associated with the fragment. This link may be updated in the chunk metadata
+//     if there's a change in the line number.
+func FragmentFirstLineAndLink(chunk *sources.Chunk) (int64, *int64, string) {
+	if chunk.SourceMetadata == nil {
+		return 0, nil, ""
+	}
+
+	var (
+		fragmentStart *int64
+		link          string
+	)
 	switch metadata := chunk.SourceMetadata.GetData().(type) {
 	case *source_metadatapb.MetaData_Git:
 		fragmentStart = &metadata.Git.Line
 	case *source_metadatapb.MetaData_Github:
 		fragmentStart = &metadata.Github.Line
+		link = metadata.Github.Link
 	case *source_metadatapb.MetaData_Gitlab:
 		fragmentStart = &metadata.Gitlab.Line
+		link = metadata.Gitlab.Link
 	case *source_metadatapb.MetaData_Bitbucket:
 		fragmentStart = &metadata.Bitbucket.Line
+		link = metadata.Bitbucket.Link
 	case *source_metadatapb.MetaData_Gerrit:
 		fragmentStart = &metadata.Gerrit.Line
 	case *source_metadatapb.MetaData_Filesystem:
 		fragmentStart = &metadata.Filesystem.Line
+		link = metadata.Filesystem.Link
+	case *source_metadatapb.MetaData_AzureRepos:
+		fragmentStart = &metadata.AzureRepos.Line
+		link = metadata.AzureRepos.Link
 	default:
-		return 0, nil
+		return 0, nil, ""
 	}
-	return *fragmentStart, fragmentStart
+	return *fragmentStart, fragmentStart, link
 }
 
 // SetResultLineNumber sets the line number in the provided result.
@@ -647,3 +678,33 @@ func SetResultLineNumber(chunk *sources.Chunk, result *detectors.Result, fragSta
 	*mdLine = fragStart + offset
 	return skip
 }
+
+// UpdateLink updates the link of the provided source metadata.
+func UpdateLink(ctx context.Context, metadata *source_metadatapb.MetaData, link string, line int64) error {
+	if metadata == nil {
+		return fmt.Errorf("metadata is nil when setting the link")
+	}
+
+	if link == "" {
+		ctx.Logger().V(4).Info("link is empty, skipping update")
+		return nil
+	}
+
+	newLink := giturl.UpdateLinkLineNumber(ctx, link, line)
+
+	switch meta := metadata.GetData().(type) {
+	case *source_metadatapb.MetaData_Github:
+		meta.Github.Link = newLink
+	case *source_metadatapb.MetaData_Gitlab:
+		meta.Gitlab.Link = newLink
+	case *source_metadatapb.MetaData_Bitbucket:
+		meta.Bitbucket.Link = newLink
+	case *source_metadatapb.MetaData_Filesystem:
+		meta.Filesystem.Link = newLink
+	case *source_metadatapb.MetaData_AzureRepos:
+		meta.AzureRepos.Link = newLink
+	default:
+		return fmt.Errorf("unsupported metadata type")
+	}
+	return nil
+}

pkg/engine/engine_test.go

@@ -2,10 +2,16 @@ package engine
 
 import (
 	"fmt"
+	"path/filepath"
 	"testing"
+	"time"
 
+	"github.com/stretchr/testify/assert"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/decoders"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/source_metadatapb"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
@@ -146,21 +152,27 @@ func TestDefaultDecoders(t *testing.T) {
 }
 
 func TestSupportsLineNumbers(t *testing.T) {
-	testCases := []struct {
-		name     string
-		input    sourcespb.SourceType
-		expected bool
+	tests := []struct {
+		name          string
+		sourceType    sourcespb.SourceType
+		expectedValue bool
 	}{
-		{"Git source type", sourcespb.SourceType_SOURCE_TYPE_GIT, true},
-		{"Github source type", sourcespb.SourceType_SOURCE_TYPE_GITHUB, true},
-		{"Gitlab source type", sourcespb.SourceType_SOURCE_TYPE_GITLAB, true},
+		{"Git source", sourcespb.SourceType_SOURCE_TYPE_GIT, true},
+		{"Github source", sourcespb.SourceType_SOURCE_TYPE_GITHUB, true},
+		{"Gitlab source", sourcespb.SourceType_SOURCE_TYPE_GITLAB, true},
+		{"Bitbucket source", sourcespb.SourceType_SOURCE_TYPE_BITBUCKET, true},
+		{"Gerrit source", sourcespb.SourceType_SOURCE_TYPE_GERRIT, true},
+		{"Github unauthenticated org source", sourcespb.SourceType_SOURCE_TYPE_GITHUB_UNAUTHENTICATED_ORG, true},
+		{"Public Git source", sourcespb.SourceType_SOURCE_TYPE_PUBLIC_GIT, true},
+		{"Filesystem source", sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM, true},
+		{"Azure Repos source", sourcespb.SourceType_SOURCE_TYPE_AZURE_REPOS, true},
+		{"Unsupported type", sourcespb.SourceType_SOURCE_TYPE_BUILDKITE, false},
 	}
 
-	for _, tc := range testCases {
-		t.Run(tc.name, func(t *testing.T) {
-			if result := SupportsLineNumbers(tc.input); result != tc.expected {
-				t.Errorf("Expected %v for input %v, got %v", tc.expected, tc.input, result)
-			}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result := SupportsLineNumbers(tt.sourceType)
+			assert.Equal(t, tt.expectedValue, result)
 		})
 	}
 }
@@ -171,3 +183,217 @@ func BenchmarkSupportsLineNumbersLoop(b *testing.B) {
 		_ = SupportsLineNumbers(sourceType)
 	}
 }
+
+// TestEngine_DuplicatSecrets is a test that detects ALL duplicate secrets with the same decoder.
+func TestEngine_DuplicatSecrets(t *testing.T) {
+	ctx := context.Background()
+
+	absPath, err := filepath.Abs("./testdata")
+	assert.Nil(t, err)
+
+	ctx, cancel := context.WithTimeout(ctx, 10*time.Second)
+	defer cancel()
+
+	e, err := Start(ctx,
+		WithConcurrency(1),
+		WithDecoders(decoders.DefaultDecoders()...),
+		WithDetectors(true, DefaultDetectors()...),
+		WithPrinter(new(discardPrinter)),
+	)
+	assert.Nil(t, err)
+
+	cfg := sources.FilesystemConfig{Paths: []string{absPath}}
+	if err := e.ScanFileSystem(ctx, cfg); err != nil {
+		return
+	}
+
+	// Wait for all the chunks to be processed.
+	assert.Nil(t, e.Finish(ctx))
+	want := uint64(5)
+	assert.Equal(t, want, e.GetMetrics().UnverifiedSecretsFound)
+}
+
+func TestFragmentFirstLineAndLink(t *testing.T) {
+	tests := []struct {
+		name         string
+		chunk        *sources.Chunk
+		expectedLine int64
+		expectedLink string
+	}{
+		{
+			name: "Test Git Metadata",
+			chunk: &sources.Chunk{
+				SourceMetadata: &source_metadatapb.MetaData{
+					Data: &source_metadatapb.MetaData_Git{
+						Git: &source_metadatapb.Git{
+							Line: 10,
+						},
+					},
+				},
+			},
+			expectedLine: 10,
+			expectedLink: "", // Git doesn't support links
+		},
+		{
+			name: "Test Github Metadata",
+			chunk: &sources.Chunk{
+				SourceMetadata: &source_metadatapb.MetaData{
+					Data: &source_metadatapb.MetaData_Github{
+						Github: &source_metadatapb.Github{
+							Line: 5,
+							Link: "https://example.github.com",
+						},
+					},
+				},
+			},
+			expectedLine: 5,
+			expectedLink: "https://example.github.com",
+		},
+		{
+			name: "Test Azure Repos Metadata",
+			chunk: &sources.Chunk{
+				SourceMetadata: &source_metadatapb.MetaData{
+					Data: &source_metadatapb.MetaData_AzureRepos{
+						AzureRepos: &source_metadatapb.AzureRepos{
+							Line: 5,
+							Link: "https://example.azure.com",
+						},
+					},
+				},
+			},
+			expectedLine: 5,
+			expectedLink: "https://example.azure.com",
+		},
+		{
+			name:         "Unsupported Type",
+			chunk:        &sources.Chunk{},
+			expectedLine: 0,
+			expectedLink: "",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			line, linePtr, link := FragmentFirstLineAndLink(tt.chunk)
+			assert.Equal(t, tt.expectedLink, link, "Mismatch in link")
+			assert.Equal(t, tt.expectedLine, line, "Mismatch in line")
+
+			if linePtr != nil {
+				assert.Equal(t, tt.expectedLine, *linePtr, "Mismatch in linePtr value")
+			}
+		})
+	}
+}
+
+func TestSetLink(t *testing.T) {
+	tests := []struct {
+		name     string
+		input    *source_metadatapb.MetaData
+		link     string
+		line     int64
+		wantLink string
+		wantErr  bool
+	}{
+		{
+			name: "Github link set",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_Github{
+					Github: &source_metadatapb.Github{},
+				},
+			},
+			link:     "https://github.com/example",
+			line:     42,
+			wantLink: "https://github.com/example#L42",
+		},
+		{
+			name: "Gitlab link set",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_Gitlab{
+					Gitlab: &source_metadatapb.Gitlab{},
+				},
+			},
+			link:     "https://gitlab.com/example",
+			line:     10,
+			wantLink: "https://gitlab.com/example#L10",
+		},
+		{
+			name: "Bitbucket link set",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_Bitbucket{
+					Bitbucket: &source_metadatapb.Bitbucket{},
+				},
+			},
+			link:     "https://bitbucket.com/example",
+			line:     8,
+			wantLink: "https://bitbucket.com/example#L8",
+		},
+		{
+			name: "Filesystem link set",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_Filesystem{
+					Filesystem: &source_metadatapb.Filesystem{},
+				},
+			},
+			link:     "file:///path/to/example",
+			line:     3,
+			wantLink: "file:///path/to/example#L3",
+		},
+		{
+			name: "Azure Repos link set",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_AzureRepos{
+					AzureRepos: &source_metadatapb.AzureRepos{},
+				},
+			},
+			link:     "https://dev.azure.com/example",
+			line:     3,
+			wantLink: "https://dev.azure.com/example?line=3",
+		},
+		{
+			name: "Unsupported metadata type",
+			input: &source_metadatapb.MetaData{
+				Data: &source_metadatapb.MetaData_Git{
+					Git: &source_metadatapb.Git{},
+				},
+			},
+			link:    "https://git.example.com/link",
+			line:    5,
+			wantErr: true,
+		},
+		{
+			name:    "Metadata nil",
+			input:   nil,
+			link:    "https://some.link",
+			line:    1,
+			wantErr: true,
+		},
+	}
+
+	ctx := context.Background()
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			err := UpdateLink(ctx, tt.input, tt.link, tt.line)
+			if err != nil && !tt.wantErr {
+				t.Errorf("Unexpected error: %v", err)
+			}
+
+			if tt.wantErr {
+				return
+			}
+
+			switch data := tt.input.GetData().(type) {
+			case *source_metadatapb.MetaData_Github:
+				assert.Equal(t, tt.wantLink, data.Github.Link, "Github link mismatch")
+			case *source_metadatapb.MetaData_Gitlab:
+				assert.Equal(t, tt.wantLink, data.Gitlab.Link, "Gitlab link mismatch")
+			case *source_metadatapb.MetaData_Bitbucket:
+				assert.Equal(t, tt.wantLink, data.Bitbucket.Link, "Bitbucket link mismatch")
+			case *source_metadatapb.MetaData_Filesystem:
+				assert.Equal(t, tt.wantLink, data.Filesystem.Link, "Filesystem link mismatch")
+			case *source_metadatapb.MetaData_AzureRepos:
+				assert.Equal(t, tt.wantLink, data.AzureRepos.Link, "Azure Repos link mismatch")
+			}
+		})
+	}
+}

pkg/engine/filesystem.go

@@ -24,18 +24,14 @@ func (e *Engine) ScanFileSystem(ctx context.Context, c sources.FilesystemConfig)
 		return err
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - filesystem", new(filesystem.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			fileSystemSource := filesystem.Source{}
-			fileSystemSource.WithFilter(c.Filter)
-			if err := fileSystemSource.Init(ctx, "trufflehog - filesystem", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			return &fileSystemSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - filesystem"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, filesystem.SourceType)
+
+	fileSystemSource := &filesystem.Source{}
+	fileSystemSource.WithFilter(c.Filter)
+	if err := fileSystemSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	_, err = e.sourceManager.Run(ctx, sourceName, fileSystemSource)
 	return err
 }

pkg/engine/gcs.go

@@ -44,18 +44,14 @@ func (e *Engine) ScanGCS(ctx context.Context, c sources.GCSConfig) error {
 		return fmt.Errorf("failed to marshal GCS connection: %w", err)
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - gcs", new(gcs.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			gcsSource := gcs.Source{}
-			if err := gcsSource.Init(ctx, "trufflehog - gcs", jobID, sourceID, true, &conn, int(c.Concurrency)); err != nil {
-				return nil, err
-			}
-			return &gcsSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - gcs"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, gcs.SourceType)
+
+	gcsSource := &gcs.Source{}
+	if err := gcsSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, int(c.Concurrency)); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	_, err = e.sourceManager.Run(ctx, sourceName, gcsSource)
 	return err
 }
 

pkg/engine/git.go

@@ -51,21 +51,18 @@ func (e *Engine) ScanGit(ctx context.Context, c sources.GitConfig) error {
 		return err
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - git", new(git.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			gitSource := git.Source{}
-			if err := gitSource.Init(ctx, "trufflehog - git", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			gitSource.WithScanOptions(scanOptions)
-			// Don't try to clean up the provided directory. That's handled by the
-			// caller of ScanGit.
-			gitSource.WithPreserveTempDirs(true)
-			return &gitSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - git"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, git.SourceType)
+
+	gitSource := &git.Source{}
+	if err := gitSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	gitSource.WithScanOptions(scanOptions)
+	// Don't try to clean up the provided directory. That's handled by the
+	// caller of ScanGit.
+	gitSource.WithPreserveTempDirs(true)
+
+	_, err := e.sourceManager.Run(ctx, sourceName, gitSource)
 	return err
 }

pkg/engine/github.go

@@ -47,18 +47,14 @@ func (e *Engine) ScanGitHub(ctx context.Context, c sources.GithubConfig) error {
 	}
 	scanOptions := git.NewScanOptions(opts...)
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - github", new(github.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			githubSource := github.Source{}
-			if err := githubSource.Init(ctx, "trufflehog - github", jobID, sourceID, true, &conn, c.Concurrency); err != nil {
-				return nil, err
-			}
-			githubSource.WithScanOptions(scanOptions)
-			return &githubSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - github"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, github.SourceType)
+
+	githubSource := &github.Source{}
+	if err := githubSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, c.Concurrency); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	githubSource.WithScanOptions(scanOptions)
+	_, err = e.sourceManager.Run(ctx, sourceName, githubSource)
 	return err
 }

pkg/engine/gitlab.go

@@ -50,18 +50,14 @@ func (e *Engine) ScanGitLab(ctx context.Context, c sources.GitlabConfig) error {
 		return err
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - gitlab", new(gitlab.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			gitlabSource := gitlab.Source{}
-			if err := gitlabSource.Init(ctx, "trufflehog - gitlab", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			gitlabSource.WithScanOptions(scanOptions)
-			return &gitlabSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - gitlab"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, gitlab.SourceType)
+
+	gitlabSource := &gitlab.Source{}
+	if err := gitlabSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	gitlabSource.WithScanOptions(scanOptions)
+	_, err = e.sourceManager.Run(ctx, sourceName, gitlabSource)
 	return err
 }

pkg/engine/s3.go

@@ -58,17 +58,13 @@ func (e *Engine) ScanS3(ctx context.Context, c sources.S3Config) error {
 		return err
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - s3", new(s3.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			s3Source := s3.Source{}
-			if err := s3Source.Init(ctx, "trufflehog - s3", jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
-				return nil, err
-			}
-			return &s3Source, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - s3"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, s3.SourceType)
+
+	s3Source := &s3.Source{}
+	if err := s3Source.Init(ctx, sourceName, jobID, sourceID, true, &conn, runtime.NumCPU()); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	_, err = e.sourceManager.Run(ctx, sourceName, s3Source)
 	return err
 }

pkg/engine/syslog.go

@@ -41,18 +41,14 @@ func (e *Engine) ScanSyslog(ctx context.Context, c sources.SyslogConfig) error {
 		return errors.WrapPrefix(err, "error unmarshalling connection", 0)
 	}
 
-	handle, err := e.sourceManager.Enroll(ctx, "trufflehog - syslog", new(syslog.Source).Type(),
-		func(ctx context.Context, jobID, sourceID int64) (sources.Source, error) {
-			syslogSource := syslog.Source{}
-			if err := syslogSource.Init(ctx, "trufflehog - syslog", jobID, sourceID, true, &conn, c.Concurrency); err != nil {
-				return nil, err
-			}
-			syslogSource.InjectConnection(connection)
-			return &syslogSource, nil
-		})
-	if err != nil {
+	sourceName := "trufflehog - syslog"
+	sourceID, jobID, _ := e.sourceManager.GetIDs(ctx, sourceName, syslog.SourceType)
+	syslogSource := &syslog.Source{}
+	if err := syslogSource.Init(ctx, sourceName, jobID, sourceID, true, &conn, c.Concurrency); err != nil {
 		return err
 	}
-	_, err = e.sourceManager.ScheduleRun(ctx, handle)
+	syslogSource.InjectConnection(connection)
+
+	_, err = e.sourceManager.Run(ctx, sourceName, syslogSource)
 	return err
 }

pkg/engine/testdata/secrets.txt

@@ -0,0 +1,6 @@
+  1 aws AKIAWARWQKZNHMZBLY4I
+  2 secret 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
+ sentry 27ac84f4bcdb4fca9701f4d6f6f58cd7d96b69c9d9754d40800645a51d668f90
+ sentry 27ac84f4bcdb4fca9701f4d6f6f58cd7d96b69c9d9754d40800645a51d668f90
+ sentry 27ac84f4bcdb4fca9701f4d6f6f58cd7d96b69c9d9754d40800645a51d668f90
+ sentry 27ac84f4bcdb4fca9701f4d6f6f58cd7d96b69c9d9754d40800645a51d668f90

pkg/giturl/giturl.go

@@ -2,10 +2,13 @@ package giturl
 
 import (
 	"net/url"
+	"regexp"
 	"strconv"
 	"strings"
 
 	"github.com/pkg/errors"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 )
 
 type provider string
@@ -143,3 +146,43 @@ func GenerateLink(repo, commit, file string, line int64) string {
 		return baseLink
 	}
 }
+
+var linePattern = regexp.MustCompile(`L\d+`)
+
+// UpdateLinkLineNumber updates the line number in a repository link.
+// Used post-link generation to refine reported issue locations within large scanned blocks.
+func UpdateLinkLineNumber(ctx context.Context, link string, newLine int64) string {
+	parsedURL, err := url.Parse(link)
+	if err != nil {
+		ctx.Logger().Error(err, "unable to parse link to update line number", "link", link)
+		return link
+	}
+
+	switch determineProvider(link) {
+	case providerBitbucket:
+		// For Bitbucket, it doesn't support line links (based on the GenerateLink function).
+		// So we don't need to change anything.
+		return link
+
+	case providerAzure:
+		// For Azure, line numbers are appended as ?line=<number>.
+		query := parsedURL.Query()
+		query.Set("line", strconv.FormatInt(newLine, 10))
+		parsedURL.RawQuery = query.Encode()
+
+	case providerGithub, providerGitlab:
+		// If the provider name isn't one of the cloud defaults, it is probably an on-prem github or gitlab.
+		// So do the same thing.
+		fallthrough
+	default:
+		// Assumed format: .../blob/<commit>/file.go#L<number>
+		fragment := "L" + strconv.FormatInt(newLine, 10)
+		if linePattern.MatchString(parsedURL.Fragment) {
+			parsedURL.Fragment = linePattern.ReplaceAllString(parsedURL.Fragment, fragment)
+		} else {
+			parsedURL.Fragment += fragment
+		}
+	}
+
+	return parsedURL.String()
+}

pkg/giturl/giturl_test.go

@@ -4,6 +4,8 @@ import (
 	"testing"
 
 	"github.com/pkg/errors"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
 )
 
 func Test_NormalizeOrgRepoURL(t *testing.T) {
@@ -160,6 +162,24 @@ func TestGenerateLink(t *testing.T) {
 			},
 			want: "https://dev.azure.com/org/project/_git/repo/commit/abcdef/main.go?line=20",
 		},
+		{
+			name: "Unknown provider on-prem instance",
+			args: args{
+				repo:   "https://onprem.customdomain.com/org/repo.git",
+				commit: "xyz123",
+				file:   "main.go",
+				line:   int64(30),
+			},
+			want: "https://onprem.customdomain.com/org/repo/blob/xyz123/main.go#L30",
+		},
+		{
+			name: "Unknown provider on-prem instance - no file",
+			args: args{
+				repo:   "https://onprem.customdomain.com/org/repo.git",
+				commit: "xyz123",
+			},
+			want: "https://onprem.customdomain.com/org/repo/commit/xyz123",
+		},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
@@ -169,3 +189,84 @@ func TestGenerateLink(t *testing.T) {
 		})
 	}
 }
+
+func TestUpdateLinkLineNumber(t *testing.T) {
+	t.Parallel()
+
+	type args struct {
+		link    string
+		newLine int64
+	}
+	tests := []struct {
+		name    string
+		args    args
+		want    string
+		wantErr bool
+	}{
+		{
+			name: "Update bitbucket, no line number supported",
+			args: args{
+				link:    "https://bitbucket.org/org/repo/blob/xyz123/main.go",
+				newLine: int64(10),
+			},
+			want: "https://bitbucket.org/org/repo/blob/xyz123/main.go",
+		},
+		{
+			name: "Update github link with line",
+			args: args{
+				link:    "https://github.com/trufflesec-julian/confluence-go-api/blob/047b4a2ba42fc5b6c0bd535c5307434a666db5ec/.gitignore#L4",
+				newLine: int64(10),
+			},
+			want: "https://github.com/trufflesec-julian/confluence-go-api/blob/047b4a2ba42fc5b6c0bd535c5307434a666db5ec/.gitignore#L10",
+		},
+		{
+			name: "Update Azure link with line",
+			args: args{
+				link:    "https://dev.azure.com/org/project/_git/repo/commit/abcdef/main.go?line=20",
+				newLine: int64(40),
+			},
+			want: "https://dev.azure.com/org/project/_git/repo/commit/abcdef/main.go?line=40",
+		},
+		{
+			name: "Add line to github link without line",
+			args: args{
+				link:    "https://github.com/trufflesec-julian/confluence-go-api/blob/047b4a2ba42fc5b6c0bd535c5307434a666db5ec/.gitignore",
+				newLine: int64(7),
+			},
+			want: "https://github.com/trufflesec-julian/confluence-go-api/blob/047b4a2ba42fc5b6c0bd535c5307434a666db5ec/.gitignore#L7",
+		},
+		{
+			name: "Update Unknown provider on-prem instance with line",
+			args: args{
+				link:    "https://onprem.customdomain.com/org/repo/blob/xyz123/main.go#L30",
+				newLine: int64(50),
+			},
+			want: "https://onprem.customdomain.com/org/repo/blob/xyz123/main.go#L50",
+		},
+		{
+			name: "Update Unknown provider on-prem instance without line",
+			args: args{
+				link:    "https://onprem.customdomain.com/org/repo/commit/xyz123",
+				newLine: int64(50),
+			},
+			want: "https://onprem.customdomain.com/org/repo/commit/xyz123#L50",
+		},
+		{
+			name: "Invalid link",
+			args: args{
+				link:    "definitely not a link",
+				newLine: int64(50),
+			},
+			wantErr: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := UpdateLinkLineNumber(context.Background(), tt.args.link, tt.args.newLine)
+			if got != tt.want && !tt.wantErr {
+				t.Errorf("UpdateLinkLineNumber() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/handlers/archive.go

@@ -78,7 +78,7 @@ func (a *Archive) FromFile(originalCtx context.Context, data io.Reader) chan []b
 			if errors.Is(err, archiver.ErrNoMatch) {
 				return
 			}
-			logger.V(2).Info("Error unarchiving chunk.")
+			logger.Error(err, "error unarchiving chunk.")
 		}
 	}()
 	return archiveChan
@@ -102,6 +102,10 @@ func (a *Archive) openArchive(ctx context.Context, depth int, reader io.Reader,
 		return a.handleNonArchiveContent(ctx, arReader, archiveChan)
 	}
 
+	if err != nil {
+		return err
+	}
+
 	switch archive := format.(type) {
 	case archiver.Decompressor:
 		info := decompressorInfo{depth: depth, reader: arReader, archiveChan: archiveChan, archiver: archive}

pkg/handlers/archive_test.go

@@ -9,6 +9,7 @@ import (
 	"regexp"
 	"strings"
 	"testing"
+	"time"
 
 	diskbufferreader "github.com/bill-rich/disk-buffer-reader"
 	"github.com/stretchr/testify/assert"
@@ -236,3 +237,37 @@ func TestExtractRPMContent(t *testing.T) {
 	expectedLength := 1822720
 	assert.Equal(t, expectedLength, len(string(content)))
 }
+
+func TestOpenInvalidArchive(t *testing.T) {
+	reader := strings.NewReader("invalid archive")
+
+	ctx := logContext.AddLogger(context.Background())
+	a := &Archive{}
+
+	archiveChan := make(chan []byte)
+
+	err := a.openArchive(ctx, 0, reader, archiveChan)
+	assert.Error(t, err)
+}
+
+func TestNestedDirArchive(t *testing.T) {
+	file, err := os.Open("testdata/dir-archive.zip")
+	assert.Nil(t, err)
+	defer file.Close()
+
+	ctx, cancel := logContext.WithTimeout(logContext.Background(), 5*time.Second)
+	defer cancel()
+	sourceChan := make(chan *sources.Chunk, 1)
+
+	go func() {
+		defer close(sourceChan)
+		HandleFile(ctx, file, &sources.Chunk{}, sourceChan)
+	}()
+
+	count := 0
+	want := 4
+	for range sourceChan {
+		count++
+	}
+	assert.Equal(t, want, count)
+}

pkg/handlers/handlers.go

@@ -47,42 +47,44 @@ func HandleFile(ctx context.Context, file io.Reader, chunkSkel *sources.Chunk, c
 		// an io.MultiReader, which is used by the SpecializedHandler.
 		reReader, err := diskbufferreader.New(file)
 		if err != nil {
-			aCtx.Logger().Error(err, "error creating re-reader reader")
+			aCtx.Logger().Error(err, "error creating reusable reader")
 			return false
 		}
-		defer reReader.Close()
 
-		// Check if the handler implements SpecializedHandler and process accordingly.
-		if specialHandler, ok := h.(SpecializedHandler); ok {
-			file, isSpecial, err := specialHandler.HandleSpecialized(aCtx, reReader)
-			if isSpecial {
-				return handleChunks(aCtx, h.FromFile(ctx, file), chunkSkel, chunksChan)
-			}
-
-			if err != nil {
-				aCtx.Logger().Error(err, "error handling file")
-			}
+		if success := processHandler(aCtx, h, reReader, chunkSkel, chunksChan); success {
+			return true
 		}
-
-		if err := reReader.Reset(); err != nil {
-			aCtx.Logger().Error(err, "error resetting re-reader")
-			return false
-		}
-		if _, isType := h.IsFiletype(aCtx, reReader); !isType {
-			continue
-		}
-
-		if err := reReader.Reset(); err != nil {
-			aCtx.Logger().Error(err, "error resetting re-reader")
-			return false
-		}
-		reReader.Stop()
-		return handleChunks(aCtx, h.FromFile(ctx, reReader), chunkSkel, chunksChan)
 	}
 
 	return false
 }
 
+func processHandler(ctx logContext.Context, h Handler, reReader *diskbufferreader.DiskBufferReader, chunkSkel *sources.Chunk, chunksChan chan *sources.Chunk) bool {
+	defer reReader.Close()
+	defer reReader.Stop()
+
+	if specialHandler, ok := h.(SpecializedHandler); ok {
+		file, isSpecial, err := specialHandler.HandleSpecialized(ctx, reReader)
+		if isSpecial {
+			return handleChunks(ctx, h.FromFile(ctx, file), chunkSkel, chunksChan)
+		}
+		if err != nil {
+			ctx.Logger().Error(err, "error handling file")
+		}
+	}
+
+	if _, err := reReader.Seek(0, io.SeekStart); err != nil {
+		ctx.Logger().Error(err, "error seeking to start of file")
+		return false
+	}
+
+	if _, isType := h.IsFiletype(ctx, reReader); !isType {
+		return false
+	}
+
+	return handleChunks(ctx, h.FromFile(ctx, reReader), chunkSkel, chunksChan)
+}
+
 func handleChunks(ctx context.Context, handlerChan chan []byte, chunkSkel *sources.Chunk, chunksChan chan *sources.Chunk) bool {
 	for {
 		select {

pkg/output/json.go

@@ -10,6 +10,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/source_metadatapb"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
 // JSONPrinter is a printer that prints results in JSON format.
@@ -20,7 +21,7 @@ func (p *JSONPrinter) Print(_ context.Context, r *detectors.ResultWithMetadata)
 		// SourceMetadata contains source-specific contextual information.
 		SourceMetadata *source_metadatapb.MetaData
 		// SourceID is the ID of the source that the API uses to map secrets to specific sources.
-		SourceID int64
+		SourceID sources.SourceID
 		// SourceType is the type of Source.
 		SourceType sourcespb.SourceType
 		// SourceName is the name of the Source.

pkg/output/plain.go

@@ -76,7 +76,7 @@ func (p *PlainPrinter) Print(_ context.Context, r *detectors.ResultWithMetadata)
 		}
 	}
 
-	aggregateData := make(map[string]interface{})
+	aggregateData := make(map[string]any)
 	var aggregateDataKeys []string
 
 	for _, data := range meta {
@@ -93,7 +93,7 @@ func (p *PlainPrinter) Print(_ context.Context, r *detectors.ResultWithMetadata)
 	return nil
 }
 
-func structToMap(obj interface{}) (m map[string]map[string]interface{}, err error) {
+func structToMap(obj any) (m map[string]map[string]any, err error) {
 	data, err := json.Marshal(obj)
 	if err != nil {
 		return

pkg/sources/circleci/circleci.go

@@ -19,13 +19,17 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
-const baseURL = "https://circleci.com/api/v1.1/"
+const (
+	SourceType = sourcespb.SourceType_SOURCE_TYPE_CIRCLECI
+
+	baseURL = "https://circleci.com/api/v1.1/"
+)
 
 type Source struct {
 	name     string
 	token    string
-	sourceId int64
-	jobId    int64
+	sourceId sources.SourceID
+	jobId    sources.JobID
 	verify   bool
 	jobPool  *errgroup.Group
 	sources.Progress
@@ -40,19 +44,19 @@ var _ sources.SourceUnitUnmarshaller = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_CIRCLECI
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
 // Init returns an initialized CircleCI source.
-func (s *Source) Init(_ context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(_ context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.name = name
 	s.sourceId = sourceId
 	s.jobId = jobId

pkg/sources/docker/docker.go

@@ -24,10 +24,12 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
+const SourceType = sourcespb.SourceType_SOURCE_TYPE_DOCKER
+
 type Source struct {
 	name        string
-	sourceId    int64
-	jobId       int64
+	sourceId    sources.SourceID
+	jobId       sources.JobID
 	verify      bool
 	concurrency int
 	conn        sourcespb.Docker
@@ -42,19 +44,19 @@ var _ sources.SourceUnitUnmarshaller = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_DOCKER
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
 // Init initializes the source.
-func (s *Source) Init(_ context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(_ context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.name = name
 	s.sourceId = sourceId
 	s.jobId = jobId

pkg/sources/filesystem/filesystem.go

@@ -22,10 +22,12 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
+const SourceType = sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM
+
 type Source struct {
 	name     string
-	sourceId int64
-	jobId    int64
+	sourceId sources.SourceID
+	jobId    sources.JobID
 	verify   bool
 	paths    []string
 	log      logr.Logger
@@ -43,19 +45,19 @@ var _ sources.SourceUnitChunker = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_FILESYSTEM
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
 // Init returns an initialized Filesystem source.
-func (s *Source) Init(aCtx context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, _ int) error {
+func (s *Source) Init(aCtx context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, _ int) error {
 	s.log = aCtx.Logger()
 
 	s.name = name

pkg/sources/gcs/gcs.go

@@ -28,7 +28,11 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
-const defaultCachePersistIncrement = 2500
+const (
+	SourceType = sourcespb.SourceType_SOURCE_TYPE_GCS
+
+	defaultCachePersistIncrement = 2500
+)
 
 // Ensure the Source satisfies the interfaces at compile time.
 var _ sources.Source = (*Source)(nil)
@@ -37,16 +41,16 @@ var _ sources.SourceUnitUnmarshaller = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_GCS
+	return SourceType
 }
 
 // SourceID number for GCS Source.
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
 // JobID number for GCS Source.
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
@@ -58,8 +62,8 @@ type objectManager interface {
 // Source represents a GCS source.
 type Source struct {
 	name        string
-	jobId       int64
-	sourceId    int64
+	jobId       sources.JobID
+	sourceId    sources.SourceID
 	concurrency int
 	verify      bool
 
@@ -107,7 +111,7 @@ func (c *persistableCache) shouldPersist() (bool, string) {
 }
 
 // Init returns an initialized GCS source.
-func (s *Source) Init(aCtx context.Context, name string, id int64, sourceID int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(aCtx context.Context, name string, id sources.JobID, sourceID sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.log = aCtx.Logger()
 
 	s.name = name

pkg/sources/git/git.go

@@ -34,10 +34,12 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
+const SourceType = sourcespb.SourceType_SOURCE_TYPE_GIT
+
 type Source struct {
 	name     string
-	sourceId int64
-	jobId    int64
+	sourceId sources.SourceID
+	jobId    sources.JobID
 	verify   bool
 	git      *Git
 	sources.Progress
@@ -51,8 +53,8 @@ type Source struct {
 type Git struct {
 	sourceType         sourcespb.SourceType
 	sourceName         string
-	sourceID           int64
-	jobID              int64
+	sourceID           sources.SourceID
+	jobID              sources.JobID
 	sourceMetadataFunc func(file, email, commit, timestamp, repository string, line int64) *source_metadatapb.MetaData
 	verify             bool
 	metrics            metrics
@@ -63,7 +65,7 @@ type metrics struct {
 	commitsScanned uint64
 }
 
-func NewGit(sourceType sourcespb.SourceType, jobID, sourceID int64, sourceName string, verify bool, concurrency int,
+func NewGit(sourceType sourcespb.SourceType, jobID sources.JobID, sourceID sources.SourceID, sourceName string, verify bool, concurrency int,
 	sourceMetadataFunc func(file, email, commit, timestamp, repository string, line int64) *source_metadatapb.MetaData,
 ) *Git {
 	return &Git{
@@ -84,14 +86,14 @@ var _ sources.SourceUnitUnmarshaller = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_GIT
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
@@ -109,7 +111,7 @@ func (s *Source) WithPreserveTempDirs(preserve bool) {
 }
 
 // Init returns an initialized GitHub source.
-func (s *Source) Init(aCtx context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(aCtx context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.name = name
 	s.sourceId = sourceId
 	s.jobId = jobId
@@ -370,6 +372,32 @@ func CloneRepo(ctx context.Context, userInfo *url.Userinfo, gitUrl string, args
 	return clonePath, repo, nil
 }
 
+// PingRepoUsingToken executes git ls-remote on a repo and returns any error that occurs. It can be used to validate
+// that a repo actually exists and is reachable.
+//
+// Pinging using other authentication methods is only unimplemented because there's been no pressing need for it yet.
+func PingRepoUsingToken(ctx context.Context, token, gitUrl, user string) error {
+	if err := GitCmdCheck(); err != nil {
+		return err
+	}
+	lsUrl, err := GitURLParse(gitUrl)
+	if err != nil {
+		return err
+	}
+	if lsUrl.User == nil {
+		lsUrl.User = url.UserPassword(user, token)
+	}
+
+	// We don't actually care about any refs on the remote, we just care whether can can list them at all. So we query
+	// only for a ref that we know won't exist to minimize the search time on the remote. (By default, ls-remote exits
+	// with 0 even if it doesn't find any matching refs.)
+	fakeRef := "TRUFFLEHOG_CHECK_GIT_REMOTE_URL_REACHABILITY"
+	gitArgs := []string{"ls-remote", lsUrl.String(), "--quiet", fakeRef}
+	cmd := exec.Command("git", gitArgs...)
+	_, err = cmd.CombinedOutput()
+	return err
+}
+
 // CloneRepoUsingToken clones a repo using a provided token.
 func CloneRepoUsingToken(ctx context.Context, token, gitUrl, user string, args ...string) (string, *git.Repository, error) {
 	userInfo := url.UserPassword(user, token)

pkg/sources/github/github.go

@@ -39,6 +39,8 @@ import (
 )
 
 const (
+	SourceType = sourcespb.SourceType_SOURCE_TYPE_GITHUB
+
 	unauthGithubOrgRateLimt = 30
 	defaultPagination       = 100
 	membersAppPagination    = 500
@@ -51,8 +53,8 @@ type Source struct {
 	githubUser  string
 	githubToken string
 
-	sourceID          int64
-	jobID             int64
+	sourceID          sources.SourceID
+	jobID             sources.JobID
 	verify            bool
 	repos             []string
 	members           []string
@@ -104,14 +106,14 @@ var endsWithGithub = regexp.MustCompile(`github\.com/?$`)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_GITHUB
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceID
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobID
 }
 
@@ -200,7 +202,7 @@ func (c *filteredRepoCache) includeRepo(s string) bool {
 }
 
 // Init returns an initialized GitHub source.
-func (s *Source) Init(aCtx context.Context, name string, jobID, sourceID int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(aCtx context.Context, name string, jobID sources.JobID, sourceID sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.log = aCtx.Logger()
 
 	s.name = name

pkg/sources/github/github_test.go

@@ -190,13 +190,13 @@ func TestAddMembersByApp(t *testing.T) {
 	gock.New("https://api.github.com").
 		Get("/app/installations").
 		Reply(200).
-		JSON([]map[string]interface{}{
+		JSON([]map[string]any{
 			{"account": map[string]string{"login": "super-secret-org", "type": "Organization"}},
 		})
 	gock.New("https://api.github.com").
 		Get("/orgs/super-secret-org/members").
 		Reply(200).
-		JSON([]map[string]interface{}{
+		JSON([]map[string]any{
 			{"login": "ssm1"},
 			{"login": "ssm2"},
 			{"login": "ssm3"},
@@ -221,7 +221,7 @@ func TestAddReposByApp(t *testing.T) {
 	gock.New("https://api.github.com").
 		Get("/installation/repositories").
 		Reply(200).
-		JSON(map[string]interface{}{
+		JSON(map[string]any{
 			"repositories": []map[string]string{
 				{"clone_url": "https://github/ssr1.git", "full_name": "ssr1"},
 				{"clone_url": "https://github/ssr2.git", "full_name": "ssr2"},
@@ -247,7 +247,7 @@ func TestAddOrgsByUser(t *testing.T) {
 	gock.New("https://api.github.com").
 		Get("/user/orgs").
 		Reply(200).
-		JSON([]map[string]interface{}{
+		JSON([]map[string]any{
 			{"login": "sso2"},
 		})
 

pkg/sources/gitlab/gitlab.go

@@ -29,10 +29,12 @@ import (
 	"google.golang.org/protobuf/types/known/anypb"
 )
 
+const SourceType = sourcespb.SourceType_SOURCE_TYPE_GITLAB
+
 type Source struct {
 	name            string
-	sourceId        int64
-	jobId           int64
+	sourceId        sources.SourceID
+	jobId           sources.JobID
 	verify          bool
 	authMethod      string
 	user            string
@@ -53,23 +55,24 @@ type Source struct {
 // Ensure the Source satisfies the interfaces at compile time.
 var _ sources.Source = (*Source)(nil)
 var _ sources.SourceUnitUnmarshaller = (*Source)(nil)
+var _ sources.Validator = (*Source)(nil)
 
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_GITLAB
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
 // Init returns an initialized Gitlab source.
-func (s *Source) Init(_ context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(_ context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.name = name
 	s.sourceId = sourceId
 	s.jobId = jobId
@@ -146,9 +149,9 @@ func (s *Source) Chunks(ctx context.Context, chunksChan chan *sources.Chunk, _ .
 		return errors.New(err)
 	}
 	// Get repo within target.
-	repos, errs := s.getRepos()
+	repos, errs := normalizeRepos(s.repos)
 	for _, repoErr := range errs {
-		ctx.Logger().Info("error getting repo", "error", repoErr)
+		ctx.Logger().Info("error normalizing repo", "error", repoErr)
 	}
 
 	// End early if we had errors getting specified repos but none were validated.
@@ -158,26 +161,14 @@ func (s *Source) Chunks(ctx context.Context, chunksChan chan *sources.Chunk, _ .
 
 	// Get all repos if not specified.
 	if len(repos) == 0 {
-		projects, err := s.getAllProjects(ctx, apiClient)
-		if err != nil {
-			return fmt.Errorf("error getting all projects: %v", err)
-		}
-		// Turn projects into URLs for Git cloner.
-		for _, prj := range projects {
-			if s.ignoreRepo(ctx, prj.PathWithNamespace) {
-				continue
-			}
-
-			// Ensure the urls are valid before adding them to the repo list.
-			_, err := url.Parse(prj.HTTPURLToRepo)
-			if err != nil {
-				fmt.Printf("could not parse url given by project: %s", prj.HTTPURLToRepo)
-			}
-			repos = append(repos, prj.HTTPURLToRepo)
-		}
-		if len(repos) == 0 {
-			return errors.Errorf("unable to discover any repos")
+		ignoreRepo := buildIgnorer(s.ignoreRepos, func(err error, pattern string) {
+			ctx.Logger().Error(err, "could not compile ignore repo glob", "glob", pattern)
+		})
+		gitlabRepos, err2, done := s.getReposFromGitlab(ctx, apiClient, ignoreRepo)
+		if done {
+			return err2
 		}
+		repos = gitlabRepos
 	}
 
 	s.repos = repos
@@ -187,6 +178,69 @@ func (s *Source) Chunks(ctx context.Context, chunksChan chan *sources.Chunk, _ .
 	return s.scanRepos(ctx, chunksChan)
 }
 
+func (s *Source) Validate(ctx context.Context) []error {
+	// The client is only used to query Gitlab for a repo list - it's not used to actually clone anything. Thus, we
+	// don't use it if there is a list of explicitly configured repos. However, constructing it validates that the
+	// configured authentication method is sensible, so we'll do it here.
+	apiClient, err := s.newClient()
+	if err != nil {
+		return []error{err}
+	}
+
+	_, _, err = apiClient.Users.CurrentUser()
+	if err != nil {
+		msg := fmt.Sprintf("gitlab authentication failed using method %v", s.authMethod)
+		return []error{errors.WrapPrefix(err, msg, 0)}
+	}
+
+	explicitlyConfiguredRepos, errs := normalizeRepos(s.repos)
+
+	if len(explicitlyConfiguredRepos) > 0 {
+		user := s.user
+		if user == "" {
+			user = "placeholder"
+		}
+
+		// We only check reachability for explicitly configured repositories. The purpose of source validation is to
+		// help users validate their configuration files, and Gitlab telling us about repositories that it won't let us
+		// access isn't a local configuration issue.
+		for _, r := range explicitlyConfiguredRepos {
+			if err := git.PingRepoUsingToken(ctx, s.token, r, user); err != nil {
+				msg := fmt.Sprintf("could not reach git repository %q", r)
+				err = errors.WrapPrefix(err, msg, 0)
+				errs = append(errs, err)
+			}
+		}
+
+		if len(s.ignoreRepos) > 0 {
+			errs = append(errs, fmt.Errorf("both repositories and ignore patterns were explicitly configured; ignore patterns will not be used"))
+		}
+	}
+
+	if len(explicitlyConfiguredRepos) > 0 || len(errs) > 0 {
+		return errs
+	}
+
+	ignoreProject := buildIgnorer(s.ignoreRepos, func(err error, pattern string) {
+		msg := fmt.Sprintf("could not compile ignore repo pattern %q", pattern)
+		errs = append(errs, errors.WrapPrefix(err, msg, 0))
+	})
+
+	projects, err := s.getAllProjects(ctx, apiClient)
+	if err != nil {
+		errs = append(errs, err)
+		return errs
+	}
+
+	for _, p := range projects {
+		if !ignoreProject(p.PathWithNamespace) {
+			return errs
+		}
+	}
+
+	return append(errs, fmt.Errorf("ignore patterns excluded all projects"))
+}
+
 func (s *Source) newClient() (*gitlab.Client, error) {
 	// Initialize a new api instance.
 	switch s.authMethod {
@@ -322,23 +376,30 @@ func (s *Source) getAllProjects(ctx context.Context, apiClient *gitlab.Client) (
 	return projectList, nil
 }
 
-func (s *Source) getRepos() ([]string, []error) {
-	if len(s.repos) == 0 {
-		return nil, nil
+func (s *Source) getReposFromGitlab(ctx context.Context, apiClient *gitlab.Client, ignoreRepo func(repo string) bool) ([]string, error, bool) {
+	projects, err := s.getAllProjects(ctx, apiClient)
+	if err != nil {
+		return nil, fmt.Errorf("error getting all projects: %v", err), true
 	}
 
-	var validRepos []string
-	var errs []error
-	for _, prj := range s.repos {
-		repo, err := giturl.NormalizeGitlabRepo(prj)
-		if err != nil {
-			errs = append(errs, errors.WrapPrefix(err, fmt.Sprintf("unable to normalize gitlab repo url %s", prj), 0))
+	// Turn projects into URLs for Git cloner.
+	var repos []string
+	for _, prj := range projects {
+		if ignoreRepo(prj.PathWithNamespace) {
 			continue
 		}
 
-		validRepos = append(validRepos, repo)
+		// Ensure the urls are valid before adding them to the repo list.
+		_, err := url.Parse(prj.HTTPURLToRepo)
+		if err != nil {
+			fmt.Printf("could not parse url given by project: %s", prj.HTTPURLToRepo)
+		}
+		repos = append(repos, prj.HTTPURLToRepo)
 	}
-	return validRepos, errs
+	if len(repos) == 0 {
+		return nil, errors.Errorf("unable to discover any repos"), true
+	}
+	return repos, nil, false
 }
 
 func (s *Source) scanRepos(ctx context.Context, chunksChan chan *sources.Chunk) error {
@@ -411,21 +472,6 @@ func (s *Source) scanRepos(ctx context.Context, chunksChan chan *sources.Chunk)
 	return nil
 }
 
-func (s *Source) ignoreRepo(ctx context.Context, r string) bool {
-	for _, ignore := range s.ignoreRepos {
-		g, err := glob.Compile(ignore)
-		if err != nil {
-			ctx.Logger().Error(err, "could not compile ignore repo glob", "glob", ignore)
-			continue
-		}
-		if g.Match(r) {
-			ctx.Logger().V(2).Info("Ignoring repo", "repo", r)
-			return true
-		}
-	}
-	return false
-}
-
 // setProgressCompleteWithRepo calls the s.SetProgressComplete after safely setting up the encoded resume info string.
 func (s *Source) setProgressCompleteWithRepo(index int, offset int, repoURL string) {
 	s.resumeInfoMutex.Lock()
@@ -445,3 +491,42 @@ func (s *Source) setProgressCompleteWithRepo(index int, offset int, repoURL stri
 func (s *Source) WithScanOptions(scanOptions *git.ScanOptions) {
 	s.scanOptions = scanOptions
 }
+
+func buildIgnorer(patterns []string, onCompileErr func(err error, pattern string)) func(repo string) bool {
+	var globs []glob.Glob
+
+	for _, pattern := range patterns {
+		g, err := glob.Compile(pattern)
+		if err != nil {
+			onCompileErr(err, pattern)
+			continue
+		}
+		globs = append(globs, g)
+	}
+
+	f := func(repo string) bool {
+		for _, g := range globs {
+			if g.Match(repo) {
+				return true
+			}
+		}
+		return false
+	}
+
+	return f
+}
+
+func normalizeRepos(repos []string) ([]string, []error) {
+	var validRepos []string
+	var errs []error
+	for _, prj := range repos {
+		repo, err := giturl.NormalizeGitlabRepo(prj)
+		if err != nil {
+			errs = append(errs, errors.WrapPrefix(err, fmt.Sprintf("unable to normalize gitlab repo url %s", prj), 0))
+			continue
+		}
+
+		validRepos = append(validRepos, repo)
+	}
+	return validRepos, errs
+}

pkg/sources/gitlab/gitlab_test.go

@@ -177,6 +177,150 @@ func TestSource_Scan(t *testing.T) {
 	}
 }
 
+func TestSource_Validate(t *testing.T) {
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	secret, err := common.GetTestSecret(ctx)
+	if err != nil {
+		t.Fatal(fmt.Errorf("failed to access secret: %v", err))
+	}
+	token := secret.MustGetField("GITLAB_TOKEN")
+	tokenWrongScope := secret.MustGetField("GITLAB_TOKEN_WRONG_SCOPE")
+
+	tests := []struct {
+		name         string
+		connection   *sourcespb.GitLab
+		wantErrCount int
+		wantErrs     []string
+	}{
+		{
+			name: "basic auth did not authenticate",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_BasicAuth{
+					BasicAuth: &credentialspb.BasicAuth{
+						Username: "bad-user",
+						Password: "bad-password",
+					},
+				},
+			},
+			wantErrCount: 1,
+		},
+		{
+			name: "token did not authenticate",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: "bad-token",
+				},
+			},
+			wantErrCount: 1,
+		},
+		{
+			name: "bad repo urls",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: token,
+				},
+				Repositories: []string{
+					"https://gitlab.com/testermctestface/testy",  // valid
+					"https://gitlab.com/testermctestface/testy/", // trailing slash
+					"ssh:git@gitlab.com/testermctestface/testy",  // bad protocol
+					"https://gitlab.com",                         // no path
+					"https://gitlab.com/",                        // no org name
+					"https://gitlab.com//testy",                  // no org name
+					"https://gitlab.com/testermctestface/",       // no repo name
+				},
+			},
+			wantErrCount: 6,
+		},
+		{
+			name: "token does not have permission to list projects",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: tokenWrongScope,
+				},
+			},
+			wantErrCount: 1,
+		},
+		{
+			name: "repositories and ignore globs both configured",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: token,
+				},
+				Repositories: []string{
+					"https://gitlab.com/testermctestface/testy", // valid
+				},
+				IgnoreRepos: []string{
+					"tes1188/*-gitlab",
+					"[", // glob doesn't compile, but this won't be checked
+				},
+			},
+			wantErrCount: 1,
+		},
+		{
+			name: "could not compile ignore glob(s)",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: token,
+				},
+				IgnoreRepos: []string{
+					"tes1188/*-gitlab",
+					"[",    // glob doesn't compile
+					"[a-]", // glob doesn't compile
+				},
+			},
+			wantErrCount: 2,
+		},
+		{
+			name: "repositories do not exist or are not accessible",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: token,
+				},
+				Repositories: []string{
+					"https://gitlab.com/testermctestface/testy",
+					"https://gitlab.com/testermctestface/doesn't-exist",
+					"https://gitlab.com/testermctestface/also-doesn't-exist",
+				},
+			},
+			wantErrCount: 2,
+		},
+		{
+			name: "ignore globs exclude all repos",
+			connection: &sourcespb.GitLab{
+				Credential: &sourcespb.GitLab_Token{
+					Token: token,
+				},
+				IgnoreRepos: []string{
+					"*",
+				},
+			},
+			wantErrCount: 1,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			s := Source{}
+
+			conn, err := anypb.New(tt.connection)
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			err = s.Init(ctx, tt.name, 0, 0, false, conn, 1)
+			if err != nil {
+				t.Fatalf("Source.Init() error: %v", err)
+			}
+
+			errs := s.Validate(ctx)
+
+			assert.Equal(t, tt.wantErrCount, len(errs))
+		})
+	}
+}
+
 func Test_setProgressCompleteWithRepo_resumeInfo(t *testing.T) {
 	tests := []struct {
 		startingResumeInfoSlice []string

pkg/sources/job_progress.go

@@ -43,8 +43,8 @@ type JobProgressHook interface {
 // If the job supports it, the reference can also be used to cancel running via
 // CancelRun.
 type JobProgressRef struct {
-	JobID       int64
-	SourceID    int64
+	JobID       JobID
+	SourceID    SourceID
 	SourceName  string
 	jobProgress *JobProgress
 }
@@ -101,8 +101,8 @@ func (f ChunkError) Unwrap() error { return f.err }
 // JobProgress aggregates information about a run of a Source.
 type JobProgress struct {
 	// Unique identifiers for this job.
-	JobID      int64
-	SourceID   int64
+	JobID      JobID
+	SourceID   SourceID
 	SourceName string
 	// Tracks whether the job is finished or not.
 	ctx    context.Context
@@ -155,7 +155,7 @@ func WithCancel(cancel context.CancelCauseFunc) func(*JobProgress) {
 }
 
 // NewJobProgress creates a new job report for the given source and job ID.
-func NewJobProgress(jobID, sourceID int64, sourceName string, opts ...func(*JobProgress)) *JobProgress {
+func NewJobProgress(jobID JobID, sourceID SourceID, sourceName string, opts ...func(*JobProgress)) *JobProgress {
 	ctx, cancel := context.WithCancel(context.Background())
 	jp := &JobProgress{
 		JobID:      jobID,

pkg/sources/job_progress_test.go

@@ -38,8 +38,8 @@ func TestJobProgressFatalErrors(t *testing.T) {
 func TestJobProgressRef(t *testing.T) {
 	jp := NewJobProgress(123, 456, "source name")
 	ref := jp.Ref()
-	assert.Equal(t, int64(123), ref.JobID)
-	assert.Equal(t, int64(456), ref.SourceID)
+	assert.Equal(t, JobID(123), ref.JobID)
+	assert.Equal(t, SourceID(456), ref.SourceID)
 
 	// Test Done() blocks until Finish() is called.
 	select {

pkg/sources/mock_job_progress_test.go

@@ -41,7 +41,7 @@ func (m *MockJobProgressHook) End(arg0 JobProgressRef, arg1 time.Time) {
 }
 
 // End indicates an expected call of End.
-func (mr *MockJobProgressHookMockRecorder) End(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) End(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "End", reflect.TypeOf((*MockJobProgressHook)(nil).End), arg0, arg1)
 }
@@ -53,7 +53,7 @@ func (m *MockJobProgressHook) EndEnumerating(arg0 JobProgressRef, arg1 time.Time
 }
 
 // EndEnumerating indicates an expected call of EndEnumerating.
-func (mr *MockJobProgressHookMockRecorder) EndEnumerating(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) EndEnumerating(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "EndEnumerating", reflect.TypeOf((*MockJobProgressHook)(nil).EndEnumerating), arg0, arg1)
 }
@@ -65,7 +65,7 @@ func (m *MockJobProgressHook) EndUnitChunking(arg0 JobProgressRef, arg1 SourceUn
 }
 
 // EndUnitChunking indicates an expected call of EndUnitChunking.
-func (mr *MockJobProgressHookMockRecorder) EndUnitChunking(arg0, arg1, arg2 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) EndUnitChunking(arg0, arg1, arg2 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "EndUnitChunking", reflect.TypeOf((*MockJobProgressHook)(nil).EndUnitChunking), arg0, arg1, arg2)
 }
@@ -77,7 +77,7 @@ func (m *MockJobProgressHook) Finish(arg0 JobProgressRef) {
 }
 
 // Finish indicates an expected call of Finish.
-func (mr *MockJobProgressHookMockRecorder) Finish(arg0 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) Finish(arg0 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "Finish", reflect.TypeOf((*MockJobProgressHook)(nil).Finish), arg0)
 }
@@ -89,7 +89,7 @@ func (m *MockJobProgressHook) ReportChunk(arg0 JobProgressRef, arg1 SourceUnit,
 }
 
 // ReportChunk indicates an expected call of ReportChunk.
-func (mr *MockJobProgressHookMockRecorder) ReportChunk(arg0, arg1, arg2 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) ReportChunk(arg0, arg1, arg2 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "ReportChunk", reflect.TypeOf((*MockJobProgressHook)(nil).ReportChunk), arg0, arg1, arg2)
 }
@@ -101,7 +101,7 @@ func (m *MockJobProgressHook) ReportError(arg0 JobProgressRef, arg1 error) {
 }
 
 // ReportError indicates an expected call of ReportError.
-func (mr *MockJobProgressHookMockRecorder) ReportError(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) ReportError(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "ReportError", reflect.TypeOf((*MockJobProgressHook)(nil).ReportError), arg0, arg1)
 }
@@ -113,7 +113,7 @@ func (m *MockJobProgressHook) ReportUnit(arg0 JobProgressRef, arg1 SourceUnit) {
 }
 
 // ReportUnit indicates an expected call of ReportUnit.
-func (mr *MockJobProgressHookMockRecorder) ReportUnit(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) ReportUnit(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "ReportUnit", reflect.TypeOf((*MockJobProgressHook)(nil).ReportUnit), arg0, arg1)
 }
@@ -125,7 +125,7 @@ func (m *MockJobProgressHook) Start(arg0 JobProgressRef, arg1 time.Time) {
 }
 
 // Start indicates an expected call of Start.
-func (mr *MockJobProgressHookMockRecorder) Start(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) Start(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "Start", reflect.TypeOf((*MockJobProgressHook)(nil).Start), arg0, arg1)
 }
@@ -137,7 +137,7 @@ func (m *MockJobProgressHook) StartEnumerating(arg0 JobProgressRef, arg1 time.Ti
 }
 
 // StartEnumerating indicates an expected call of StartEnumerating.
-func (mr *MockJobProgressHookMockRecorder) StartEnumerating(arg0, arg1 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) StartEnumerating(arg0, arg1 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "StartEnumerating", reflect.TypeOf((*MockJobProgressHook)(nil).StartEnumerating), arg0, arg1)
 }
@@ -149,7 +149,7 @@ func (m *MockJobProgressHook) StartUnitChunking(arg0 JobProgressRef, arg1 Source
 }
 
 // StartUnitChunking indicates an expected call of StartUnitChunking.
-func (mr *MockJobProgressHookMockRecorder) StartUnitChunking(arg0, arg1, arg2 interface{}) *gomock.Call {
+func (mr *MockJobProgressHookMockRecorder) StartUnitChunking(arg0, arg1, arg2 any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "StartUnitChunking", reflect.TypeOf((*MockJobProgressHook)(nil).StartUnitChunking), arg0, arg1, arg2)
 }

pkg/sources/s3/s3.go

@@ -31,6 +31,8 @@ import (
 )
 
 const (
+	SourceType = sourcespb.SourceType_SOURCE_TYPE_S3
+
 	defaultAWSRegion     = "us-east-1"
 	defaultMaxObjectSize = 250 * 1024 * 1024 // 250 MiB
 	maxObjectSizeLimit   = 250 * 1024 * 1024 // 250 MiB
@@ -38,8 +40,8 @@ const (
 
 type Source struct {
 	name        string
-	sourceId    int64
-	jobId       int64
+	sourceId    sources.SourceID
+	jobId       sources.JobID
 	verify      bool
 	concurrency int
 	log         logr.Logger
@@ -58,19 +60,19 @@ var _ sources.Validator = (*Source)(nil)
 
 // Type returns the type of source
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_S3
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
 // Init returns an initialized AWS source
-func (s *Source) Init(aCtx context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(aCtx context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 	s.log = context.WithValues(aCtx, "source", s.Type(), "name", name).Logger()
 
 	s.name = name
@@ -198,7 +200,7 @@ func (s *Source) scanBuckets(ctx context.Context, client *s3.S3, role string, bu
 		}
 
 		s.SetProgressComplete(i, len(bucketsToScan), fmt.Sprintf("Bucket: %s", bucket), "")
-		logger.Info("Scanning bucket")
+		logger.V(3).Info("Scanning bucket")
 
 		regionalClient, err := s.getRegionalClientForBucket(ctx, client, role, bucket)
 		if err != nil {
@@ -278,7 +280,7 @@ func (s *Source) pageChunker(ctx context.Context, client *s3.S3, chunksChan chan
 
 		// ignore large files
 		if *obj.Size > s.maxObjectSize {
-			s.log.V(3).Info("Skipping %d byte file (over maxObjectSize limit)", "object", *obj.Key)
+			s.log.V(5).Info("Skipping %d byte file (over maxObjectSize limit)", "object", *obj.Key)
 			continue
 		}
 

pkg/sources/source_manager.go

@@ -2,10 +2,12 @@ package sources
 
 import (
 	"fmt"
+	"runtime"
 	"sync"
 	"sync/atomic"
 	"time"
 
+	"github.com/marusama/semaphore/v2"
 	"golang.org/x/sync/errgroup"
 
 	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
@@ -13,32 +15,12 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
 )
 
-// handle uniquely identifies a Source given to the manager to manage. If the
-// SourceManager is connected to the API, it will be equivalent to the unique
-// source ID, otherwise it behaves as a counter.
-type handle int64
-
-// SourceInitFunc is a function that takes a source and job ID and returns an
-// initialized Source.
-type SourceInitFunc func(ctx context.Context, jobID, sourceID int64) (Source, error)
-
-// sourceInfo is an aggregate struct to store source information provided on
-// initialization.
-type sourceInfo struct {
-	initFunc SourceInitFunc
-	name     string
-}
-
 type SourceManager struct {
 	api   apiClient
 	hooks []JobProgressHook
-	// Map of handle to source initializer.
-	handles     map[handle]sourceInfo
-	handlesLock sync.Mutex
 	// Pool limiting the amount of concurrent sources running.
-	pool                errgroup.Group
-	poolLimit           int
-	currentRunningCount int32
+	sem semaphore.Semaphore
+	wg  sync.WaitGroup
 	// Max number of units to scan concurrently per source.
 	concurrentUnits int
 	// Run the sources using source unit enumeration / chunking if available.
@@ -46,15 +28,16 @@ type SourceManager struct {
 	// Downstream chunks channel to be scanned.
 	outputChunks chan *Chunk
 	// Set when Wait() returns.
-	done bool
+	firstErr chan error
+	waitErr  error
+	done     bool
 }
 
 // apiClient is an interface for optionally communicating with an external API.
 type apiClient interface {
-	// RegisterSource lets the API know we have a source and it returns a unique source ID for it.
-	RegisterSource(ctx context.Context, name string, kind sourcespb.SourceType) (int64, error)
-	// GetJobID queries the API for an existing job or new job ID.
-	GetJobID(ctx context.Context, id int64) (int64, error)
+	// GetIDs informs the API of the source that's about to run and returns
+	// two identifiers used during source initialization.
+	GetIDs(ctx context.Context, name string, kind sourcespb.SourceType) (SourceID, JobID, error)
 }
 
 // WithAPI adds an API client to the manager for tracking jobs and progress. If
@@ -72,8 +55,7 @@ func WithReportHook(hook JobProgressHook) func(*SourceManager) {
 // WithConcurrentSources limits the concurrent number of sources a manager can run.
 func WithConcurrentSources(concurrency int) func(*SourceManager) {
 	return func(mgr *SourceManager) {
-		mgr.pool.SetLimit(concurrency)
-		mgr.poolLimit = concurrency
+		mgr.sem.SetLimit(concurrency)
 	}
 }
 
@@ -99,8 +81,9 @@ func NewManager(opts ...func(*SourceManager)) *SourceManager {
 	mgr := SourceManager{
 		// Default to the headless API. Can be overwritten by the WithAPI option.
 		api:          &headlessAPI{},
-		handles:      make(map[handle]sourceInfo),
+		sem:          semaphore.New(runtime.NumCPU()),
 		outputChunks: make(chan *Chunk),
+		firstErr:     make(chan error, 1),
 	}
 	for _, opt := range opts {
 		opt(&mgr)
@@ -108,81 +91,53 @@ func NewManager(opts ...func(*SourceManager)) *SourceManager {
 	return &mgr
 }
 
-// Enroll informs the SourceManager to track and manage a Source.
-func (s *SourceManager) Enroll(ctx context.Context, name string, kind sourcespb.SourceType, f SourceInitFunc) (handle, error) {
-	if s.done {
-		return 0, fmt.Errorf("manager is done")
-	}
-	id, err := s.api.RegisterSource(ctx, name, kind)
-	if err != nil {
-		return 0, err
-	}
-	handleID := handle(id)
-	s.handlesLock.Lock()
-	defer s.handlesLock.Unlock()
-	if _, ok := s.handles[handleID]; ok {
-		// TODO: smartly handle this?
-		return 0, fmt.Errorf("handle ID '%d' already in use", handleID)
-	}
-	s.handles[handleID] = sourceInfo{
-		initFunc: f,
-		name:     name,
-	}
-	return handleID, nil
+func (s *SourceManager) GetIDs(ctx context.Context, sourceName string, kind sourcespb.SourceType) (SourceID, JobID, error) {
+	return s.api.GetIDs(ctx, sourceName, kind)
 }
 
 // Run blocks until a resource is available to run the source, then
-// synchronously runs it. The first fatal error, if any, will be returned.
-func (s *SourceManager) Run(ctx context.Context, handle handle) (JobProgressRef, error) {
-	progress, err := s.asyncRun(ctx, handle)
-	if err != nil {
-		return progress, err
-	}
-	<-progress.Done()
-	return progress, progress.Snapshot().FatalError()
-}
-
-// ScheduleRun blocks until a resource is available to run the source, then
 // asynchronously runs it. Error information is stored and accessible via the
 // JobProgressRef as it becomes available.
-func (s *SourceManager) ScheduleRun(ctx context.Context, handle handle) (JobProgressRef, error) {
-	return s.asyncRun(ctx, handle)
+func (s *SourceManager) Run(ctx context.Context, sourceName string, source Source) (JobProgressRef, error) {
+	return s.asyncRun(ctx, sourceName, source)
 }
 
-// asyncRun is a helper method to asynchronously run the Source. It calls out
-// to the API to get a job ID for this run, creates a JobProgress object, then
-// waits for an available goroutine to asynchronously run it.
-func (s *SourceManager) asyncRun(ctx context.Context, handle handle) (JobProgressRef, error) {
+// asyncRun is a helper method to asynchronously run the Source.
+func (s *SourceManager) asyncRun(ctx context.Context, sourceName string, source Source) (JobProgressRef, error) {
+	sourceID, jobID := source.SourceID(), source.JobID()
 	// Do preflight checks before waiting on the pool.
-	if err := s.preflightChecks(ctx, handle); err != nil {
-		return JobProgressRef{}, err
-	}
-	// Get the name. Should never fail due to preflight checks.
-	sourceInfo, ok := s.getSourceInfo(handle)
-	if !ok {
-		return JobProgressRef{SourceID: int64(handle)}, fmt.Errorf("unrecognized handle")
-	}
-	sourceName := sourceInfo.name
-	// Get a Job ID.
-	ctx = context.WithValue(ctx, "source_id", int64(handle))
-	jobID, err := s.api.GetJobID(ctx, int64(handle))
-	if err != nil {
-		return JobProgressRef{SourceID: int64(handle), SourceName: sourceName}, err
+	if err := s.preflightChecks(ctx); err != nil {
+		return JobProgressRef{
+			SourceName: sourceName,
+			SourceID:   sourceID,
+			JobID:      jobID,
+		}, err
 	}
 	// Create a JobProgress object for tracking progress.
 	ctx, cancel := context.WithCancelCause(ctx)
-	progress := NewJobProgress(jobID, int64(handle), sourceName, WithHooks(s.hooks...), WithCancel(cancel))
-	s.pool.Go(func() error {
-		atomic.AddInt32(&s.currentRunningCount, 1)
-		defer atomic.AddInt32(&s.currentRunningCount, -1)
+	progress := NewJobProgress(jobID, sourceID, sourceName, WithHooks(s.hooks...), WithCancel(cancel))
+	if err := s.sem.Acquire(ctx, 1); err != nil {
+		// Context cancelled.
+		progress.ReportError(Fatal{err})
+		return progress.Ref(), Fatal{err}
+	}
+	s.wg.Add(1)
+	go func() {
+		defer s.sem.Release(1)
+		defer s.wg.Done()
 		ctx := context.WithValues(ctx,
 			"job_id", jobID,
 			"source_manager_worker_id", common.RandomID(5),
 		)
 		defer common.Recover(ctx)
 		defer cancel(nil)
-		return s.run(ctx, handle, jobID, progress)
-	})
+		if err := s.run(ctx, source, progress); err != nil {
+			select {
+			case s.firstErr <- err:
+			default:
+			}
+		}
+	}()
 	return progress.Ref(), nil
 }
 
@@ -199,13 +154,19 @@ func (s *SourceManager) Chunks() <-chan *Chunk {
 func (s *SourceManager) Wait() error {
 	// Check if the manager has been Waited.
 	if s.done {
-		return s.pool.Wait()
+		return s.waitErr
 	}
-	defer close(s.outputChunks)
-	defer func() { s.done = true }()
+	s.done = true
 
 	// Return the first error returned by run.
-	return s.pool.Wait()
+	s.wg.Wait()
+	select {
+	case s.waitErr = <-s.firstErr:
+	default:
+	}
+	close(s.outputChunks)
+	close(s.firstErr)
+	return s.waitErr
 }
 
 // ScanChunk injects a chunk into the output stream of chunks to be scanned.
@@ -216,33 +177,45 @@ func (s *SourceManager) ScanChunk(chunk *Chunk) {
 }
 
 // AvailableCapacity returns the number of concurrent jobs the manager can
-// accommodate at this time. If there is no limit, -1 is returned.
+// accommodate at this time.
 func (s *SourceManager) AvailableCapacity() int {
-	if s.poolLimit == 0 {
-		return -1
-	}
-	runCount := atomic.LoadInt32(&s.currentRunningCount)
-	return s.poolLimit - int(runCount)
+	return s.sem.GetLimit() - s.sem.GetCount()
+}
+
+// MaxConcurrentSources returns the maximum configured limit of concurrent
+// sources the manager will run.
+func (s *SourceManager) MaxConcurrentSources() int {
+	return s.sem.GetLimit()
+}
+
+// ConcurrentSources returns the current number of concurrently running
+// sources.
+func (s *SourceManager) ConcurrentSources() int {
+	return s.sem.GetCount()
+}
+
+// SetMaxConcurrentSources sets the maximum number of concurrently running
+// sources. If the count is lower than the already existing number of
+// concurrently running sources, no sources will be scheduled to run until the
+// existing sources complete.
+func (s *SourceManager) SetMaxConcurrentSources(maxRunCount int) {
+	s.sem.SetLimit(maxRunCount)
 }
 
 // preflightChecks is a helper method to check the Manager or the context isn't
-// done and that the handle is valid.
-func (s *SourceManager) preflightChecks(ctx context.Context, handle handle) error {
+// done.
+func (s *SourceManager) preflightChecks(ctx context.Context) error {
 	// Check if the manager has been Waited.
 	if s.done {
 		return fmt.Errorf("manager is done")
 	}
-	// Check the handle is valid.
-	if _, ok := s.getSourceInfo(handle); !ok {
-		return fmt.Errorf("unrecognized handle")
-	}
 	return ctx.Err()
 }
 
 // run is a helper method to sychronously run the source. It does not check for
 // acquired resources. An error is returned if there was a fatal error during
 // the run. This information is also recorded in the JobProgress.
-func (s *SourceManager) run(ctx context.Context, handle handle, jobID int64, report *JobProgress) error {
+func (s *SourceManager) run(ctx context.Context, source Source, report *JobProgress) error {
 	defer report.Finish()
 	report.Start(time.Now())
 	defer func() { report.End(time.Now()) }()
@@ -253,34 +226,21 @@ func (s *SourceManager) run(ctx context.Context, handle handle, jobID int64, rep
 		}
 	}()
 
-	// Initialize the source.
-	sourceInfo, ok := s.getSourceInfo(handle)
-	if !ok {
-		// Shouldn't happen due to preflight checks.
-		err := fmt.Errorf("unrecognized handle")
-		report.ReportError(Fatal{err})
-		return Fatal{err}
-	}
-	source, err := sourceInfo.initFunc(ctx, jobID, int64(handle))
-	if err != nil {
-		report.ReportError(Fatal{err})
-		return Fatal{err}
-	}
 	report.TrackProgress(source.GetProgress())
 	ctx = context.WithValues(ctx,
 		"source_type", source.Type().String(),
-		"source_name", sourceInfo.name,
+		"source_name", report.SourceName,
 	)
 	// Check for the preferred method of tracking source units.
 	if enumChunker, ok := source.(SourceUnitEnumChunker); ok && s.useSourceUnits {
-		return s.runWithUnits(ctx, handle, enumChunker, report)
+		return s.runWithUnits(ctx, enumChunker, report)
 	}
-	return s.runWithoutUnits(ctx, handle, source, report)
+	return s.runWithoutUnits(ctx, source, report)
 }
 
 // runWithoutUnits is a helper method to run a Source. It has coarse-grained
 // job reporting.
-func (s *SourceManager) runWithoutUnits(ctx context.Context, handle handle, source Source, report *JobProgress) error {
+func (s *SourceManager) runWithoutUnits(ctx context.Context, source Source, report *JobProgress) error {
 	// Introspect on the chunks we get from the Chunks method.
 	ch := make(chan *Chunk, 1)
 	var wg sync.WaitGroup
@@ -310,7 +270,7 @@ func (s *SourceManager) runWithoutUnits(ctx context.Context, handle handle, sour
 // runWithUnits is a helper method to run a Source that is also a
 // SourceUnitEnumChunker. This allows better introspection of what is getting
 // scanned and any errors encountered.
-func (s *SourceManager) runWithUnits(ctx context.Context, handle handle, source SourceUnitEnumChunker, report *JobProgress) error {
+func (s *SourceManager) runWithUnits(ctx context.Context, source SourceUnitEnumChunker, report *JobProgress) error {
 	unitReporter := &mgrUnitReporter{
 		unitCh: make(chan SourceUnit, 1),
 		report: report,
@@ -385,28 +345,15 @@ func (s *SourceManager) runWithUnits(ctx context.Context, handle handle, source
 	}
 }
 
-// getSourceInfo is a helper method for safe concurrent access to the
-// map[handle]SourceInitFunc map.
-func (s *SourceManager) getSourceInfo(handle handle) (sourceInfo, bool) {
-	s.handlesLock.Lock()
-	defer s.handlesLock.Unlock()
-	f, ok := s.handles[handle]
-	return f, ok
-}
-
 // headlessAPI implements the apiClient interface locally.
 type headlessAPI struct {
-	// Counters for assigning handle and job IDs.
+	// Counters for assigning source and job IDs.
 	sourceIDCounter int64
 	jobIDCounter    int64
 }
 
-func (api *headlessAPI) RegisterSource(ctx context.Context, name string, kind sourcespb.SourceType) (int64, error) {
-	return atomic.AddInt64(&api.sourceIDCounter, 1), nil
-}
-
-func (api *headlessAPI) GetJobID(ctx context.Context, id int64) (int64, error) {
-	return atomic.AddInt64(&api.jobIDCounter, 1), nil
+func (api *headlessAPI) GetIDs(context.Context, string, sourcespb.SourceType) (SourceID, JobID, error) {
+	return SourceID(atomic.AddInt64(&api.sourceIDCounter, 1)), JobID(atomic.AddInt64(&api.jobIDCounter, 1)), nil
 }
 
 // mgrUnitReporter implements the UnitReporter interface.

pkg/sources/source_manager_test.go

@@ -15,15 +15,15 @@ import (
 
 // DummySource implements Source and is used for testing a SourceManager.
 type DummySource struct {
-	sourceID int64
-	jobID    int64
+	sourceID SourceID
+	jobID    JobID
 	chunker
 }
 
 func (d *DummySource) Type() sourcespb.SourceType { return 1337 }
-func (d *DummySource) SourceID() int64            { return d.sourceID }
-func (d *DummySource) JobID() int64               { return d.jobID }
-func (d *DummySource) Init(_ context.Context, _ string, jobID, sourceID int64, _ bool, _ *anypb.Any, _ int) error {
+func (d *DummySource) SourceID() SourceID         { return d.sourceID }
+func (d *DummySource) JobID() JobID               { return d.jobID }
+func (d *DummySource) Init(_ context.Context, _ string, jobID JobID, sourceID SourceID, _ bool, _ *anypb.Any, _ int) error {
 	d.sourceID = sourceID
 	d.jobID = jobID
 	return nil
@@ -80,16 +80,13 @@ func (c errorChunker) Chunks(context.Context, chan *Chunk, ...ChunkingTarget) er
 func (c errorChunker) Enumerate(context.Context, UnitReporter) error                { return c }
 func (c errorChunker) ChunkUnit(context.Context, SourceUnit, ChunkReporter) error   { return c }
 
-// enrollDummy is a helper function to enroll a DummySource with a SourceManager.
-func enrollDummy(mgr *SourceManager, chunkMethod chunker) (handle, error) {
-	return mgr.Enroll(context.Background(), "dummy", 1337,
-		func(ctx context.Context, jobID, sourceID int64) (Source, error) {
-			source := &DummySource{chunker: chunkMethod}
-			if err := source.Init(ctx, "dummy", jobID, sourceID, true, nil, 42); err != nil {
-				return nil, err
-			}
-			return source, nil
-		})
+// buildDummy is a helper function to enroll a DummySource with a SourceManager.
+func buildDummy(chunkMethod chunker) (Source, error) {
+	source := &DummySource{chunker: chunkMethod}
+	if err := source.Init(context.Background(), "dummy", 123, 456, true, nil, 42); err != nil {
+		return nil, err
+	}
+	return source, nil
 }
 
 // tryRead is a helper function that will try to read from a channel and return
@@ -105,11 +102,13 @@ func tryRead(ch <-chan *Chunk) (*Chunk, error) {
 
 func TestSourceManagerRun(t *testing.T) {
 	mgr := NewManager(WithBufferedOutput(8))
-	handle, err := enrollDummy(mgr, &counterChunker{count: 1})
+	source, err := buildDummy(&counterChunker{count: 1})
 	assert.NoError(t, err)
 	for i := 0; i < 3; i++ {
-		_, err = mgr.Run(context.Background(), handle)
+		ref, err := mgr.Run(context.Background(), "dummy", source)
+		<-ref.Done()
 		assert.NoError(t, err)
+		assert.NoError(t, ref.Snapshot().FatalError())
 		chunk, err := tryRead(mgr.Chunks())
 		assert.NoError(t, err)
 		assert.Equal(t, []byte{byte(i)}, chunk.Data)
@@ -121,35 +120,31 @@ func TestSourceManagerRun(t *testing.T) {
 
 func TestSourceManagerWait(t *testing.T) {
 	mgr := NewManager()
-	handle, err := enrollDummy(mgr, &counterChunker{count: 1})
+	source, err := buildDummy(&counterChunker{count: 1})
 	assert.NoError(t, err)
 	// Asynchronously run the source.
-	_, err = mgr.ScheduleRun(context.Background(), handle)
+	_, err = mgr.Run(context.Background(), "dummy", source)
 	assert.NoError(t, err)
 	// Read the 1 chunk we're expecting so Waiting completes.
 	<-mgr.Chunks()
 	// Wait for all resources to complete.
 	assert.NoError(t, mgr.Wait())
-	// Enroll and run should return an error now.
-	_, err = enrollDummy(mgr, &counterChunker{count: 1})
-	assert.Error(t, err)
-	_, err = mgr.ScheduleRun(context.Background(), handle)
+	// Run should return an error now.
+	_, err = buildDummy(&counterChunker{count: 1})
+	assert.NoError(t, err)
+	_, err = mgr.Run(context.Background(), "dummy", source)
 	assert.Error(t, err)
 }
 
 func TestSourceManagerError(t *testing.T) {
 	mgr := NewManager()
-	handle, err := enrollDummy(mgr, errorChunker{fmt.Errorf("oops")})
+	source, err := buildDummy(errorChunker{fmt.Errorf("oops")})
 	assert.NoError(t, err)
-	// A synchronous run should fail.
-	_, err = mgr.Run(context.Background(), handle)
-	assert.Error(t, err)
-	// Scheduling a run should not fail, but the error should surface in
-	// Wait().
-	ref, err := mgr.ScheduleRun(context.Background(), handle)
+	ref, err := mgr.Run(context.Background(), "dummy", source)
 	assert.NoError(t, err)
-	assert.Error(t, mgr.Wait())
+	<-ref.Done()
 	assert.Error(t, ref.Snapshot().FatalError())
+	assert.Error(t, mgr.Wait())
 }
 
 func TestSourceManagerReport(t *testing.T) {
@@ -159,11 +154,11 @@ func TestSourceManagerReport(t *testing.T) {
 		{WithBufferedOutput(8), WithSourceUnits(), WithConcurrentUnits(1)},
 	} {
 		mgr := NewManager(opts...)
-		handle, err := enrollDummy(mgr, &counterChunker{count: 4})
+		source, err := buildDummy(&counterChunker{count: 4})
 		assert.NoError(t, err)
-		// Synchronously run the source.
-		ref, err := mgr.Run(context.Background(), handle)
+		ref, err := mgr.Run(context.Background(), "dummy", source)
 		assert.NoError(t, err)
+		<-ref.Done()
 		assert.Equal(t, 0, len(ref.Snapshot().Errors))
 		assert.Equal(t, uint64(4), ref.Snapshot().TotalChunks)
 	}
@@ -221,10 +216,11 @@ func TestSourceManagerNonFatalError(t *testing.T) {
 		{unit: "three", err: "not again"},
 	}
 	mgr := NewManager(WithBufferedOutput(8), WithSourceUnits())
-	handle, err := enrollDummy(mgr, &unitChunker{input})
+	source, err := buildDummy(&unitChunker{input})
 	assert.NoError(t, err)
-	ref, err := mgr.Run(context.Background(), handle)
+	ref, err := mgr.Run(context.Background(), "dummy", source)
 	assert.NoError(t, err)
+	<-ref.Done()
 	report := ref.Snapshot()
 	assert.Equal(t, len(input), int(report.TotalUnits))
 	assert.Equal(t, len(input), int(report.FinishedUnits))
@@ -235,11 +231,11 @@ func TestSourceManagerNonFatalError(t *testing.T) {
 
 func TestSourceManagerContextCancelled(t *testing.T) {
 	mgr := NewManager(WithBufferedOutput(8))
-	handle, err := enrollDummy(mgr, &counterChunker{count: 100})
+	source, err := buildDummy(&counterChunker{count: 100})
 	assert.NoError(t, err)
 
 	ctx, cancel := context.WithCancel(context.Background())
-	ref, err := mgr.ScheduleRun(ctx, handle)
+	ref, err := mgr.Run(ctx, "dummy", source)
 	assert.NoError(t, err)
 
 	cancel()
@@ -249,47 +245,18 @@ func TestSourceManagerContextCancelled(t *testing.T) {
 }
 
 type DummyAPI struct {
-	registerSource func(context.Context, string, sourcespb.SourceType) (int64, error)
-	getJobID       func(context.Context, int64) (int64, error)
+	registerSource func(context.Context, string, sourcespb.SourceType) (SourceID, error)
+	getJobID       func(context.Context, SourceID) (JobID, error)
 }
 
-func (api DummyAPI) RegisterSource(ctx context.Context, name string, kind sourcespb.SourceType) (int64, error) {
+func (api DummyAPI) RegisterSource(ctx context.Context, name string, kind sourcespb.SourceType) (SourceID, error) {
 	return api.registerSource(ctx, name, kind)
 }
 
-func (api DummyAPI) GetJobID(ctx context.Context, id int64) (int64, error) {
+func (api DummyAPI) GetJobID(ctx context.Context, id SourceID) (JobID, error) {
 	return api.getJobID(ctx, id)
 }
 
-func TestSourceManagerJobAndSourceIDs(t *testing.T) {
-	mgr := NewManager(WithAPI(DummyAPI{
-		registerSource: func(context.Context, string, sourcespb.SourceType) (int64, error) {
-			return 1337, nil
-		},
-		getJobID: func(context.Context, int64) (int64, error) {
-			return 9001, nil
-		},
-	}))
-	var (
-		initializedJobID    int64
-		initializedSourceID int64
-	)
-	handle, err := mgr.Enroll(context.Background(), "dummy", 1337,
-		func(ctx context.Context, jobID, sourceID int64) (Source, error) {
-			initializedJobID = jobID
-			initializedSourceID = sourceID
-			return nil, fmt.Errorf("ignore")
-		})
-	assert.NoError(t, err)
-
-	ref, _ := mgr.Run(context.Background(), handle)
-	assert.Equal(t, int64(1337), initializedSourceID)
-	assert.Equal(t, int64(1337), ref.SourceID)
-	assert.Equal(t, int64(9001), initializedJobID)
-	assert.Equal(t, int64(9001), ref.JobID)
-	assert.Equal(t, "dummy", ref.SourceName)
-}
-
 // Chunk method that has a custom callback for the Chunks method.
 type callbackChunker struct {
 	cb func(context.Context, chan *Chunk) error
@@ -304,7 +271,7 @@ func (c callbackChunker) ChunkUnit(context.Context, SourceUnit, ChunkReporter) e
 func TestSourceManagerCancelRun(t *testing.T) {
 	mgr := NewManager(WithBufferedOutput(8))
 	var returnedErr error
-	handle, err := enrollDummy(mgr, callbackChunker{func(ctx context.Context, _ chan *Chunk) error {
+	source, err := buildDummy(callbackChunker{func(ctx context.Context, _ chan *Chunk) error {
 		// The context passed to Chunks should get cancelled when ref.CancelRun() is called.
 		<-ctx.Done()
 		returnedErr = fmt.Errorf("oh no: %w", ctx.Err())
@@ -312,7 +279,7 @@ func TestSourceManagerCancelRun(t *testing.T) {
 	}})
 	assert.NoError(t, err)
 
-	ref, err := mgr.ScheduleRun(context.Background(), handle)
+	ref, err := mgr.Run(context.Background(), "dummy", source)
 	assert.NoError(t, err)
 
 	cancelErr := fmt.Errorf("abort! abort!")
@@ -326,7 +293,7 @@ func TestSourceManagerCancelRun(t *testing.T) {
 func TestSourceManagerAvailableCapacity(t *testing.T) {
 	mgr := NewManager(WithConcurrentSources(1337))
 	start, end := make(chan struct{}), make(chan struct{})
-	handle, err := enrollDummy(mgr, callbackChunker{func(context.Context, chan *Chunk) error {
+	source, err := buildDummy(callbackChunker{func(context.Context, chan *Chunk) error {
 		start <- struct{}{} // Send start signal.
 		<-end               // Wait for end signal.
 		return nil
@@ -334,7 +301,7 @@ func TestSourceManagerAvailableCapacity(t *testing.T) {
 	assert.NoError(t, err)
 
 	assert.Equal(t, 1337, mgr.AvailableCapacity())
-	ref, err := mgr.ScheduleRun(context.Background(), handle)
+	ref, err := mgr.Run(context.Background(), "dummy", source)
 	assert.NoError(t, err)
 
 	<-start // Wait for start signal.

pkg/sources/sources.go

@@ -11,14 +11,19 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
 )
 
+type (
+	SourceID int64
+	JobID    int64
+)
+
 // Chunk contains data to be decoded and scanned along with context on where it came from.
 type Chunk struct {
 	// SourceName is the name of the Source that produced the chunk.
 	SourceName string
 	// SourceID is the ID of the source that the Chunk originated from.
-	SourceID int64
+	SourceID SourceID
 	// JobID is the ID of the job that the Chunk originated from.
-	JobID int64
+	JobID JobID
 	// SourceType is the type of Source that produced the chunk.
 	SourceType sourcespb.SourceType
 	// SourceMetadata holds the context of where the Chunk was found.
@@ -45,11 +50,11 @@ type Source interface {
 	// Type returns the source type, used for matching against configuration and jobs.
 	Type() sourcespb.SourceType
 	// SourceID returns the initialized source ID used for tracking relationships in the DB.
-	SourceID() int64
+	SourceID() SourceID
 	// JobID returns the initialized job ID used for tracking relationships in the DB.
-	JobID() int64
+	JobID() JobID
 	// Init initializes the source.
-	Init(aCtx context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error
+	Init(aCtx context.Context, name string, jobId JobID, sourceId SourceID, verify bool, connection *anypb.Any, concurrency int) error
 	// Chunks emits data over a channel which is then decoded and scanned for secrets.
 	// By default, data is obtained indiscriminately. However, by providing one or more
 	// ChunkingTarget parameters, the caller can direct the function to retrieve
@@ -272,6 +277,9 @@ type Validator interface {
 // scope should be the len(scopedItems)
 // message is the public facing user information about the current progress
 // encodedResumeInfo is an optional string representing any information necessary to resume the job if interrupted
+//
+//	NOTE: SetProgressOngoing should be used when source does not yet know how many items it is scanning (scope)
+//	and does not want to display a percentage complete
 func (p *Progress) SetProgressComplete(i, scope int, message, encodedResumeInfo string) {
 	p.mut.Lock()
 	defer p.mut.Unlock()
@@ -290,6 +298,23 @@ func (p *Progress) SetProgressComplete(i, scope int, message, encodedResumeInfo
 	p.PercentComplete = int64((float64(i) / float64(scope)) * 100)
 }
 
+// SetProgressOngoing sets information about the current running job based on
+// the highest level objects in the source.
+// message is the public facing user information about the current progress
+// encodedResumeInfo is an optional string representing any information necessary to resume the job if interrupted
+//
+//	NOTE: This method should be used over SetProgressComplete when the source does
+//	not yet know how many items it is scanning and does not want to display a percentage complete.
+func (p *Progress) SetProgressOngoing(message string, encodedResumeInfo string) {
+	p.mut.Lock()
+	defer p.mut.Unlock()
+
+	p.Message = message
+	p.EncodedResumeInfo = encodedResumeInfo
+	// Explicitly set SectionsRemaining to 0 so the frontend does not display a percent.
+	p.SectionsRemaining = 0
+}
+
 // GetProgress gets job completion percentage for metrics reporting.
 func (p *Progress) GetProgress() *Progress {
 	p.mut.Lock()

pkg/sources/syslog/syslog.go

@@ -23,12 +23,16 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
 )
 
-const nilString = ""
+const (
+	SourceType = sourcespb.SourceType_SOURCE_TYPE_SYSLOG
+
+	nilString = ""
+)
 
 type Source struct {
 	name     string
-	sourceId int64
-	jobId    int64
+	sourceId sources.SourceID
+	jobId    sources.JobID
 	verify   bool
 	syslog   *Syslog
 	sources.Progress
@@ -38,14 +42,14 @@ type Source struct {
 type Syslog struct {
 	sourceType         sourcespb.SourceType
 	sourceName         string
-	sourceID           int64
-	jobID              int64
+	sourceID           sources.SourceID
+	jobID              sources.JobID
 	sourceMetadataFunc func(hostname, appname, procid, timestamp, facility, client string) *source_metadatapb.MetaData
 	verify             bool
 	concurrency        *semaphore.Weighted
 }
 
-func NewSyslog(sourceType sourcespb.SourceType, jobID, sourceID int64, sourceName string, verify bool, concurrency int,
+func NewSyslog(sourceType sourcespb.SourceType, jobID sources.JobID, sourceID sources.SourceID, sourceName string, verify bool, concurrency int,
 	sourceMetadataFunc func(hostname, appname, procid, timestamp, facility, client string) *source_metadatapb.MetaData,
 ) *Syslog {
 	return &Syslog{
@@ -103,14 +107,14 @@ var _ sources.Source = (*Source)(nil)
 // Type returns the type of source.
 // It is used for matching source types in configuration and job input.
 func (s *Source) Type() sourcespb.SourceType {
-	return sourcespb.SourceType_SOURCE_TYPE_SYSLOG
+	return SourceType
 }
 
-func (s *Source) SourceID() int64 {
+func (s *Source) SourceID() sources.SourceID {
 	return s.sourceId
 }
 
-func (s *Source) JobID() int64 {
+func (s *Source) JobID() sources.JobID {
 	return s.jobId
 }
 
@@ -119,7 +123,7 @@ func (s *Source) InjectConnection(conn *sourcespb.Syslog) {
 }
 
 // Init returns an initialized Syslog source.
-func (s *Source) Init(_ context.Context, name string, jobId, sourceId int64, verify bool, connection *anypb.Any, concurrency int) error {
+func (s *Source) Init(_ context.Context, name string, jobId sources.JobID, sourceId sources.SourceID, verify bool, connection *anypb.Any, concurrency int) error {
 
 	s.name = name
 	s.sourceId = sourceId