feat(desktop): custom in-app window title bar on Windows (#7781 )

<img width="1750" height="1223" alt="image" src="https://github.com/user-attachments/assets/c115a30c-be62-4eaa-8de3-26a93a605919" /> ## Custom windows top bar * Doesn't work on mac * doesn't effect web * little effort been put into mobile view
Combine Policies and Pipelines pages (#7681 )
2026-09-03 05:10:16 +03:00 · 2026-09-02 21:31:45 +00:00 · 2026-09-02 16:08:15 +00:00 · 2026-09-02 14:18:57 +00:00 · 2026-09-02 13:57:56 +00:00 · 2026-09-02 12:35:22 +00:00
896 changed files with 86092 additions and 14603 deletions
@@ -0,0 +1,19 @@
+{
+  "$schema": "https://json.schemastore.org/claude-code-settings.json",
+  "hooks": {
+    "Stop": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node",
+            "args": [
+              "${CLAUDE_PROJECT_DIR}/scripts/lint/comment-lint-hook.mjs"
+            ],
+            "timeout": 60
+          }
+        ]
+      }
+    ]
+  }
+}
@@ -20,6 +20,7 @@ Closes #(issue_number)
 - [ ] I have read the [Stirling-PDF Developer Guide](https://github.com/Stirling-Tools/Stirling-PDF/blob/main/DeveloperGuide.md) (if applicable)
 - [ ] I have read the [How to add new languages to Stirling-PDF](https://github.com/Stirling-Tools/Stirling-PDF/blob/main/devGuide/HowToAddNewLanguage.md) (if applicable)
 - [ ] I have performed a self-review of my own code
+- [ ] Every comment I added says something the code does not ([guide](https://github.com/Stirling-Tools/Stirling-PDF/blob/main/devGuide/CODE_COMMENTS.md))
 - [ ] My changes generate no new warnings

 ### Documentation
@@ -182,7 +182,7 @@ jobs:
          fetch-depth: 0 # Fetch full history for commit hash detection

      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Get version number
        id: versionNumber
@@ -220,6 +220,42 @@ jobs:
            echo "app_short=${APP_HASH:0:8}" >> $GITHUB_OUTPUT
          fi

+      # The Stirling account previews connect to. Derived from the ref rather than stored as a URL
+      # so it cannot drift from the key: a mismatched pair is accepted by the browser and rejected
+      # by Supabase, surfacing much later as "session expired" on Usage rather than at sign-in.
+      # Secret only to match Saas-Dev-Deploy.yml, which owns the same value; a project ref is not
+      # itself sensitive, which is why SAAS_API_BASE_URL next to it is a plain variable.
+      - name: Resolve Stirling account config
+        id: saas
+        env:
+          PROJECT_REF: ${{ secrets.SAAS_DB_PROJECT_REF }}
+          API_BASE_OVERRIDE: ${{ vars.SAAS_API_BASE_URL }}
+        run: |
+          # Set, this is the one value both halves use: the browser's portal reads and the backend's
+          # register/entitlement calls have to land on the same SaaS, and nothing checks that they
+          # do. Unset, only the backend gets a base, from its own compiled-in default.
+          API_BASE="${API_BASE_OVERRIDE:-https://stirling.com/app}"
+          echo "backend_base=${API_BASE}" >> "$GITHUB_OUTPUT"
+
+          if [ -z "${PROJECT_REF}" ]; then
+            echo "Not configured for this environment: the preview will build without a Stirling"
+            echo "account, and the connect dialog will say so. To wire one up, set on the"
+            echo "pr-preview environment the secrets SAAS_DB_PROJECT_REF and"
+            echo "SAAS_SUPABASE_PUBLISHABLE_KEY, both from the same Supabase project."
+            echo "supabase_url=" >> "$GITHUB_OUTPUT"
+            echo "frontend_base=" >> "$GITHUB_OUTPUT"
+          else
+            # Only whether, not which: the ref is a secret here, so Actions masks it out of any
+            # line it appears in, derived URL included.
+            echo "Stirling account configured, at ${API_BASE}."
+            echo "supabase_url=https://${PROJECT_REF}.supabase.co" >> "$GITHUB_OUTPUT"
+            # Deliberately the override and not API_BASE: the backend's default is a subpath URL
+            # nobody has confirmed answers /api/v1, and prod CORS does not list preview hostnames,
+            # so portal reads stay off until someone sets a base they have checked. Empty leaves the
+            # committed .env default alone, which is the clean "not configured" state.
+            echo "frontend_base=${API_BASE_OVERRIDE}" >> "$GITHUB_OUTPUT"
+          fi
+
      - name: Check if image exists
        id: check-image
        run: |
@@ -246,6 +282,9 @@ jobs:
          build-args: |
            VERSION_TAG=v2-alpha
            BUILD_PORTAL=${{ env.BUILD_PORTAL }}
+            VITE_SUPABASE_URL=${{ steps.saas.outputs.supabase_url }}
+            VITE_SUPABASE_PUBLISHABLE_DEFAULT_KEY=${{ secrets.SAAS_SUPABASE_PUBLISHABLE_KEY }}
+            VITE_SAAS_API_URL=${{ steps.saas.outputs.frontend_base }}
          platforms: linux/amd64

      - name: Set up SSH
@@ -279,6 +318,13 @@ jobs:
              environment:
                DISABLE_ADDITIONAL_FEATURES: "false"
                STIRLING_BILLING_ACCOUNT_LINK_ENABLED: "true"
+                STIRLING_BILLING_ACCOUNT_LINK_SAAS_BASE_URL: "${{ steps.saas.outputs.backend_base }}"
+                # Off so preview traffic never accrues against a real wallet or trips its cap. The
+                # 402 gate is separate and stays on, so gating is still testable here.
+                STIRLING_BILLING_ACCOUNT_LINK_METERING_ENABLED: "false"
+                # Stated rather than inferred from the request: the callback has to come back to the
+                # preview hostname, not to the container's own :8080 behind this proxy.
+                SYSTEM_FRONTENDURL: "https://${V2_PORT}.ssl.stirlingpdf.cloud"
                SECURITY_ENABLELOGIN: "true"
                SECURITY_INITIALLOGIN_USERNAME: "${TEST_LOGIN_USERNAME}"
                SECURITY_INITIALLOGIN_PASSWORD: "${TEST_LOGIN_PASSWORD}"
@@ -353,7 +399,7 @@ jobs:

      - name: Install Task for Storybook
        if: steps.sb-changes.outputs.storybook == 'true'
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Build and deploy Storybook
        id: storybook
@@ -206,7 +206,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Run Gradle Command
        run: |
          if [ "${{ needs.check-comment.outputs.disable_security }}" == "true" ]; then
@@ -222,7 +222,7 @@ jobs:
          STIRLING_PDF_DESKTOP_UI: false

      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Login to GitHub Container Registry
        uses: docker/login-action@dbcb813823bdd20940b903addbd779551569679f # v4.6.0
@@ -0,0 +1,246 @@
+name: Auto SaaS Dev Deployment
+
+on:
+  push:
+    branches:
+      - saas-prod
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+env:
+  FRONTEND_PORT: "901"
+  BACKEND_PORT: "902"
+  DEPLOY_DIR: /stirling/SAAS-DEV
+
+jobs:
+  deploy-saas-dev:
+    runs-on: ubuntu-latest
+    environment: saas-dev
+    concurrency:
+      group: saas-dev-deploy
+      cancel-in-progress: true
+    permissions:
+      contents: read
+      packages: write
+
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@05e31511f85b41b11d1cf0ef85d0992719546e2c # v2.21.0
+        with:
+          egress-policy: audit
+
+      - name: Check SaaS configuration
+        id: config
+        env:
+          PROJECT_REF: ${{ secrets.SAAS_DB_PROJECT_REF }}
+        run: |
+          echo "supabase_url=https://${PROJECT_REF}.supabase.co" >> "$GITHUB_OUTPUT"
+          echo "meter_endpoint=https://${PROJECT_REF}.supabase.co/functions/v1/meter-payg-units" >> "$GITHUB_OUTPUT"
+
+      - name: Checkout repository
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        with:
+          fetch-depth: 0
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0
+
+      - name: Login to GitHub Container Registry
+        uses: docker/login-action@dbcb813823bdd20940b903addbd779551569679f # v4.6.0
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ github.token }}
+
+      - name: Convert repository owner to lowercase
+        id: repoowner
+        run: echo "lowercase=$(echo ${{ github.repository_owner }} | awk '{print tolower($0)}')" >> $GITHUB_OUTPUT
+
+      - name: Get commit hash
+        id: commit-hash
+        run: echo "app_short=$(git rev-parse --short=8 HEAD)" >> $GITHUB_OUTPUT
+
+      - name: Build and push backend image
+        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7.3.0
+        with:
+          context: .
+          file: ./docker/backend/Dockerfile
+          push: true
+          cache-from: type=gha,scope=stirling-saas-backend
+          cache-to: type=gha,mode=max,scope=stirling-saas-backend
+          tags: |
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-backend-${{ steps.commit-hash.outputs.app_short }}
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-backend-latest
+          build-args: |
+            VERSION_TAG=v2-alpha
+            STIRLING_FLAVOR=saas
+          platforms: linux/amd64
+
+      - name: Build and push frontend image
+        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7.3.0
+        with:
+          context: .
+          file: ./docker/frontend/Dockerfile
+          push: true
+          cache-from: type=gha,scope=stirling-saas-frontend
+          cache-to: type=gha,mode=max,scope=stirling-saas-frontend
+          tags: |
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-frontend-${{ steps.commit-hash.outputs.app_short }}
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-frontend-latest
+          build-args: |
+            VERSION_TAG=v2-alpha
+            STIRLING_FLAVOR=saas
+            VITE_BUILD_MODE=development
+            VITE_SUPABASE_URL=${{ steps.config.outputs.supabase_url }}
+            VITE_SUPABASE_PUBLISHABLE_DEFAULT_KEY=${{ secrets.SAAS_SUPABASE_PUBLISHABLE_KEY }}
+          platforms: linux/amd64
+
+      - name: Build and push AI engine image
+        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7.3.0
+        with:
+          context: .
+          file: ./engine/Dockerfile
+          push: true
+          cache-from: type=gha,scope=stirling-saas-engine
+          cache-to: type=gha,mode=max,scope=stirling-saas-engine
+          tags: |
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-engine-${{ steps.commit-hash.outputs.app_short }}
+            ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test:saas-engine-latest
+          platforms: linux/amd64
+
+      - name: Set up SSH
+        env:
+          SSH_KEY: ${{ secrets.NEW_VPS_SSH_KEY }}
+        run: |
+          mkdir -p ~/.ssh/
+          echo "$SSH_KEY" > ../private.key
+          sudo chmod 600 ../private.key
+
+      - name: Deploy to VPS
+        env:
+          IMAGE_BASE: ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-pdf-test
+          IMAGE_TAG: ${{ steps.commit-hash.outputs.app_short }}
+          GHCR_USER: ${{ github.actor }}
+          GHCR_TOKEN: ${{ github.token }}
+          VPS_USERNAME: ${{ secrets.NEW_VPS_USERNAME }}
+          VPS_HOST: ${{ secrets.NEW_VPS_HOST }}
+          SAAS_DB_URL: ${{ secrets.SAAS_DB_URL }}
+          SAAS_DB_USERNAME: ${{ secrets.SAAS_DB_USERNAME || 'postgres' }}
+          SAAS_DB_PASSWORD: ${{ secrets.SAAS_DB_PASSWORD }}
+          SAAS_DB_PROJECT_REF: ${{ secrets.SAAS_DB_PROJECT_REF }}
+          SUPABASE_EDGE_FUNCTION_SECRET: ${{ secrets.SUPABASE_EDGE_FUNCTION_SECRET }}
+          PAYG_METER_ENDPOINT: ${{ steps.config.outputs.meter_endpoint }}
+          STIRLING_KEYGEN_ENABLED: ${{ secrets.KEYGEN_ACCOUNT_ID != '' && secrets.KEYGEN_API_TOKEN != '' && secrets.KEYGEN_POLICY_ID != '' }}
+          KEYGEN_ACCOUNT_ID: ${{ secrets.KEYGEN_ACCOUNT_ID }}
+          KEYGEN_API_TOKEN: ${{ secrets.KEYGEN_API_TOKEN }}
+          KEYGEN_POLICY_ID: ${{ secrets.KEYGEN_POLICY_ID }}
+          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+          VOYAGE_API_KEY: ${{ secrets.VOYAGE_API_KEY }}
+        run: |
+          set -euo pipefail
+
+          BASE_URL="http://${VPS_HOST}:${FRONTEND_PORT}"
+
+          yaml() {
+            printf "'%s'" "$(printf '%s' "$1" | sed -e "s/'/''/g" -e 's/\$/$$/g')"
+          }
+
+            ENGINE_SECRET="$(openssl rand -hex 32)"
+            AI_BACKEND_VARS="
+                SYSTEM_AIENGINE_ENABLED: \"true\"
+                SYSTEM_AIENGINE_URL: \"http://saas-engine:5001\"
+                APP_AI_SERVICEBASEURL: \"http://saas-engine:5001\"
+                STIRLING_ENGINE_SHARED_SECRET: $(yaml "$ENGINE_SECRET")"
+            AI_SERVICE="
+
+            saas-engine:
+              container_name: stirling-saas-dev-engine
+              image: ${IMAGE_BASE}:saas-engine-${IMAGE_TAG}
+              environment:
+                ANTHROPIC_API_KEY: $(yaml "$ANTHROPIC_API_KEY")
+                VOYAGE_API_KEY: $(yaml "$VOYAGE_API_KEY")
+                STIRLING_ENGINE_SHARED_SECRET: $(yaml "$ENGINE_SECRET")
+              restart: on-failure:5"
+
+          cat > docker-compose.yml << EOF
+          version: '3.3'
+          services:
+            saas-backend:
+              container_name: stirling-saas-dev-backend
+              image: ${IMAGE_BASE}:saas-backend-${IMAGE_TAG}
+              ports:
+                - "${BACKEND_PORT}:8080"
+              volumes:
+                - ${DEPLOY_DIR}/config:/configs:rw
+                - ${DEPLOY_DIR}/logs:/logs:rw
+                - ${DEPLOY_DIR}/storage:/storage:rw
+              environment:
+                SPRING_PROFILES_ACTIVE: "saas"
+                DISABLE_ADDITIONAL_FEATURES: "false"
+                SAAS_DB_URL: $(yaml "$SAAS_DB_URL")
+                SAAS_DB_USERNAME: $(yaml "$SAAS_DB_USERNAME")
+                SAAS_DB_PASSWORD: $(yaml "$SAAS_DB_PASSWORD")
+                SAAS_DB_PROJECT_REF: $(yaml "$SAAS_DB_PROJECT_REF")
+                SUPABASE_EDGE_FUNCTION_SECRET: $(yaml "$SUPABASE_EDGE_FUNCTION_SECRET")
+                PAYG_METER_ENDPOINT: $(yaml "$PAYG_METER_ENDPOINT")
+                STIRLING_KEYGEN_ENABLED: $(yaml "$STIRLING_KEYGEN_ENABLED")
+                KEYGEN_ACCOUNT_ID: $(yaml "$KEYGEN_ACCOUNT_ID")
+                KEYGEN_API_TOKEN: $(yaml "$KEYGEN_API_TOKEN")
+                KEYGEN_POLICY_ID: $(yaml "$KEYGEN_POLICY_ID")
+                SYSTEM_DEFAULTLOCALE: en-US
+                SYSTEM_MAXFILESIZE: "100"
+                METRICS_ENABLED: "true"
+                SYSTEM_GOOGLEVISIBILITY: "false"
+                SWAGGER_SERVER_URL: "${BASE_URL}"
+                baseUrl: "${BASE_URL}"${AI_BACKEND_VARS}
+              restart: on-failure:5
+
+            saas-frontend:
+              container_name: stirling-saas-dev-frontend
+              image: ${IMAGE_BASE}:saas-frontend-${IMAGE_TAG}
+              ports:
+                - "${FRONTEND_PORT}:80"
+              environment:
+                VITE_API_BASE_URL: "http://saas-backend:8080"
+              depends_on:
+                - saas-backend
+              restart: on-failure:5${AI_SERVICE}
+          EOF
+
+          SSH_OPTS=(-i ../private.key -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null)
+
+          scp "${SSH_OPTS[@]}" docker-compose.yml "${VPS_USERNAME}@${VPS_HOST}:/tmp/saas-dev-docker-compose.yml"
+
+          ssh "${SSH_OPTS[@]}" -T "${VPS_USERNAME}@${VPS_HOST}" << ENDSSH
+            set -e
+            mkdir -p ${DEPLOY_DIR}/{config,logs,storage}
+            mv /tmp/saas-dev-docker-compose.yml ${DEPLOY_DIR}/docker-compose.yml
+            chmod 600 ${DEPLOY_DIR}/docker-compose.yml
+            cd ${DEPLOY_DIR}
+            printf '%s' "${GHCR_TOKEN}" | docker login ghcr.io -u "${GHCR_USER}" --password-stdin
+            docker-compose down --remove-orphans 2>/dev/null || true
+            docker-compose pull
+            docker-compose up -d
+            docker logout ghcr.io >/dev/null 2>&1 || true
+            docker image prune -af --filter "until=336h" --filter "label!=keep=true" || true
+          ENDSSH
+
+      - name: Wait for the backend to answer
+        env:
+          VPS_HOST: ${{ secrets.NEW_VPS_HOST }}
+        run: |
+          URL="http://${VPS_HOST}:${BACKEND_PORT}/api/v1/info/status"
+          for i in $(seq 1 60); do
+            code=$(curl -s -o /dev/null -w '%{http_code}' --max-time 5 "$URL" || true)
+            if [ "$code" = "200" ]; then echo "Healthy after $((i * 10))s"; exit 0; fi
+            sleep 10
+          done
+          echo "::error::SaaS dev backend did not become healthy within 10 minutes"
+          exit 1
+
+      - name: Cleanup temporary files
+        if: always()
+        run: rm -f ../private.key docker-compose.yml
+        continue-on-error: true
@@ -34,10 +34,9 @@ jobs:
          cache-dependency-glob: |
            engine/pyproject.toml
            engine/uv.lock
-          cache-suffix: ai-engine

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Quality-check engine
        id: engine-check
@@ -52,7 +52,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Check Java formatting (Spotless)
        # Runs once per matrix combination - pick the cheapest leg
        # (core - no proprietary, no saas) so we don't wait for the
@@ -95,7 +95,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Install Playwright (chromium only)
        run: task e2e:install -- chromium
      - name: Build frontend (needed for playwright's vite preview webServer)
@@ -42,7 +42,6 @@ jobs:
          cache-dependency-glob: |
            engine/pyproject.toml
            engine/uv.lock
-          cache-suffix: generated-models

      - name: Restore cache Gradle User Home
        if: inputs.use_shared_cache
@@ -76,7 +75,7 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Verify generated models are up to date
        id: models-check
@@ -38,7 +38,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Check licenses for compatibility
        run: task backend:licenses:check
        env:
@@ -39,7 +39,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Generate OpenAPI documentation
        run: task backend:swagger
        env:
@@ -57,7 +57,7 @@ jobs:
      # runtime token isn't exposed) since the docker driver can't use it.
      - name: Set up Docker Buildx
        if: inputs.docker-base-changed != 'true'
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      # Expose ACTIONS_RUNTIME_TOKEN / ACTIONS_RESULTS_URL for docker buildx type=gha cache backend.
      - name: Expose GitHub runtime for Buildx cache
@@ -45,7 +45,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Install Playwright (chromium only)
        run: task e2e:install -- chromium
      - name: Build frontend (production bundle for vite preview)
@@ -44,7 +44,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Build frontend (production bundle for vite preview)
        env:
          VITE_BUILD_FOR_PREVIEW: "1"
@@ -36,7 +36,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: a11y gate (changed stories)
        run: task frontend:storybook:a11y:changed -- origin/${{ github.base_ref || 'main' }}
      - name: Upload scan reports
@@ -97,7 +97,7 @@ jobs:
        run: npm ci --ignore-scripts --audit=false --fund=false

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Generate frontend license report (Push only)
        if: github.event_name == 'push'
@@ -367,7 +367,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Check licenses and generate report
        id: license-check
@@ -27,7 +27,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Quality-check frontend
        id: frontend-check
        run: task frontend:check:all
@@ -69,7 +69,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Get version number
        id: versionNumber
        run: |
@@ -169,7 +169,7 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Build JAR
        run: ./gradlew build ${{ matrix.variant.build_frontend && '-PbuildWithFrontend=true' || '' }} -x spotlessApply -x spotlessCheck -x test -x sonarqube
@@ -268,7 +268,7 @@ jobs:
          distribution: ${{ matrix.platform == 'windows-11-arm' && 'microsoft' || 'temurin' }}

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      # Build the universal JRE before desktop:prepare so the jlink:runtime
      # task short-circuits on its `test -d runtime/jre` status check.
@@ -38,7 +38,7 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Install all Playwright browsers
        run: task e2e:install

@@ -89,7 +89,7 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: a11y gate (every story, ${{ matrix.theme }})
        run: task frontend:storybook:a11y:${{ matrix.theme }}
@@ -162,7 +162,7 @@ jobs:
            engine/uv.lock

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Start the fat image with login and storage enabled
        run: docker compose -f docker/embedded/compose/test_cicd.yml up -d --build
@@ -31,10 +31,14 @@ jobs:
          cache-dependency-glob: |
            engine/pyproject.toml
            engine/uv.lock
-          cache-suffix: pre-commit

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Run pre-commit checks
        run: task pre-commit
+
+      # The fixture corpus checks the comment rules themselves, so it runs here
+      # rather than on every local commit.
+      - name: Check the comment-lint fixture corpus
+        run: task pre-commit:comment-lint:selftest
@@ -69,7 +69,7 @@ jobs:

      - name: Set up Docker Buildx
        id: buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Set up QEMU
        uses: docker/setup-qemu-action@96fe6ef7f33517b61c61be40b68a1882f3264fb8 # v4.2.0
@@ -85,10 +85,10 @@ jobs:

      - name: Set up Docker Buildx
        id: buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Get version number
        id: versionNumber
        run: echo "versionNumber=$(./gradlew printVersion --quiet | tail -1)" >> $GITHUB_OUTPUT
@@ -75,6 +75,6 @@ jobs:

      # Upload the results to GitHub's code scanning dashboard.
      - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@ff2f1c621b7f889edc0d3c761ac2e6a3f8cdb0dd # v4.37.7
+        uses: github/codeql-action/upload-sarif@db488ddef3bf6cb639b32c2e9a7c0a7ea8271d28 # v4.37.8
        with:
          sarif_file: results.sarif
@@ -63,7 +63,7 @@ jobs:
          SWAGGERHUB_USER: "Frooodle"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Get version number
        id: versionNumber
        run: echo "versionNumber=$(./gradlew printVersion --quiet | tail -1)" >> $GITHUB_OUTPUT
@@ -59,14 +59,13 @@ jobs:
          cache-dependency-glob: |
            engine/pyproject.toml
            engine/uv.lock
-          cache-suffix: sync-files

      - name: Install Python dependencies
        run: |
          uv sync --project engine --locked --group tools

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Sync translation TOML files
        run: |
@@ -212,7 +212,7 @@ jobs:
          distribution: ${{ matrix.platform == 'windows-11-arm' && 'microsoft' || 'temurin' }}

      - name: Setup Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0

      - name: Build universal macOS JRE
        if: matrix.platform == 'macos-15'
@@ -127,7 +127,7 @@ jobs:
          distribution: "temurin"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@a00fbb05ce67b35648be3c78cbc9fd85354c757e # v2.2.0
      - name: Build application
        run: task backend:build
        env:
@@ -142,7 +142,7 @@ jobs:

      - name: Set up Docker Buildx
        id: buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Set base image and platform for this build
        id: build-params
@@ -229,7 +229,7 @@ jobs:

      - name: Set up Docker Buildx
        id: buildx
-        uses: docker/setup-buildx-action@bb05f3f5519dd87d3ba754cc423b652a5edd6d2c # v4.2.0
+        uses: docker/setup-buildx-action@37fe631027851001ddb9b187196cc803df7f5f0e # v4.3.0

      - name: Build docker/unoserver/Dockerfile
        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7.3.0
@@ -298,8 +298,13 @@ docs/type3/signatures/

 **/application-dev-local.properties

-# Claude
-.claude/
+# Claude. Contents are ignored so personal config stays local, with the two
+# shared pieces re-included: settings.json (the comment-lint hook) and skills/.
+# The directory itself cannot be ignored or git will not look inside it.
+.claude/*
+!.claude/settings.json
+!.claude/skills/
+.claude/settings.local.json

 # Playwright MCP screenshots / traces
 .playwright-mcp/
@@ -307,3 +312,4 @@ docs/type3/signatures/

 # Local screenshot artifacts from *-screenshots.spec.ts
 frontend/editor/screenshots/
+frontend/editor/src-tauri/libs/.variant
@@ -11,6 +11,7 @@ vars:
    '.github/scripts/*.py'
    'app/core/src/main/resources/static/python/*.py'
    ':(exclude)*split_photos.py'
+    ':(exclude)scripts/lint/fixtures/*'
  SPELL_FILES: >-
    '*.html'
    '*.css'
@@ -59,6 +60,7 @@ tasks:
      - task: gitleaks
      - task: whitespace
      - task: toml-sort
+      - task: comment-lint

  fix:
    desc: "Auto-fix formatting, spelling, and secrets issues across the repo"
@@ -75,6 +77,7 @@ tasks:
        vars: { FIX: '1' }
      - task: codespell
      - task: gitleaks
+      - task: comment-lint

  install:
    desc: "Install the pinned pre-commit Python tools"
@@ -130,6 +133,85 @@ tasks:
    cmds:
      - "{{.GITLEAKS_BIN}} git --pre-commit --redact --staged --verbose"

+  comment-lint:
+    desc: "Check comment quality on the lines this branch adds"
+    summary: |
+      Blocks a comment that restates the code below it, a section banner, or a
+      block of commented-out code. Everything else it reports is advisory.
+
+      Scoped to added lines, so touching an old file never surfaces the standing
+      backlog. The standard is devGuide/CODE_COMMENTS.md.
+
+      With no arguments it diffs the working tree against HEAD, which is what a
+      pre-commit run wants: the lines you are about to commit. On a CI pull request
+      it diffs against the target branch instead, via GITHUB_BASE_REF.
+
+      To ask what a whole branch adds instead, use the branch variant, which
+      needs no argument passing:
+        task comment-lint:branch
+
+      Full tree (report only):     task pre-commit:comment-lint:all
+      Fixture corpus:             task pre-commit:comment-lint:selftest
+    # Depends on the frontend install because the .ts/.tsx half of the rule set
+    # runs as an oxlint plugin. Without it the TS engine warns and skips, which
+    # would leave the frontend silently unchecked on CI.
+    deps: [":frontend:install"]
+    cmds:
+      - node scripts/lint/comment-lint.mjs {{.CLI_ARGS}}
+
+  comment-lint:branch:
+    desc: "Check comment quality on everything this branch adds over its base"
+    summary: |
+      Like `task comment-lint`, but scoped to the whole branch rather than to
+      uncommitted work, so it still reports after you commit.
+
+      Exists as its own task because passing `-- --since origin/main` through Task
+      is not portable: with the npm build of Task the launcher is a PowerShell
+      script, and PowerShell strips the `--` before Task sees it, leaving Task to
+      print its own usage.
+
+      Override the base with BASE=<ref>.
+    vars:
+      BASE: '{{.BASE | default "origin/main"}}'
+    deps: [":frontend:install"]
+    cmds:
+      - node scripts/lint/comment-lint.mjs --since {{.BASE}}
+
+  comment-lint:ci:
+    desc: "Comment gate as CI runs it: fixture corpus, then the diff"
+    summary: |
+      The corpus checks the rules themselves rather than the code under review, so
+      it belongs on CI and not on every local commit. Run this before changing a
+      rule, and let CI run it on every pull request.
+    deps: [":frontend:install"]
+    cmds:
+      - node scripts/lint/comment-lint.mjs --selftest
+      - node scripts/lint/comment-lint.mjs {{.CLI_ARGS}}
+
+  comment-lint:hook:
+    desc: "Comment gate for the editor hook: everything this turn changed"
+    summary: |
+      Same scope as `task comment-lint`, kept as its own name so the hook has a
+      stable entry point and the taskfile shows every way the linter is invoked.
+
+      Not in the frontend-install dependency chain on purpose: this runs at the end
+      of every turn, so it stays as short as it can be. If oxlint is missing the TS
+      half warns and skips.
+    cmds:
+      - node scripts/lint/comment-lint.mjs
+
+  comment-lint:all:
+    desc: "Report every comment finding in the tree (never fails)"
+    deps: [":frontend:install"]
+    cmds:
+      - node scripts/lint/comment-lint.mjs --all
+
+  comment-lint:selftest:
+    desc: "Check both comment-lint engines against the fixture corpus"
+    deps: [":frontend:install"]
+    cmds:
+      - node scripts/lint/comment-lint.mjs --selftest
+
  gitleaks-bin:
    internal: true
    desc: "Ensure the pinned, checksum-verified gitleaks binary is cached in .task/bin"
@@ -21,6 +21,43 @@ Task `desc:` fields should describe **what** the task does, not **how** it does
 - `task docker:build` — build standard Docker image
 - `task docker:up` — start Docker compose stack

+## Comments
+
+A comment must carry information the code cannot. If a reader could derive it from the code in front of them, delete it.
+
+Comment the current state. Not what the code used to do, not what changed, not why it changed: git holds that. Where history explains the shape, state the reason instead, so "this used to reimplement the modal internals" becomes "thin wrapper over the shared Modal: duplicating its portal and focus trap is how dialogs drift apart". Future state goes in a TODO with an issue.
+
+Write a comment when it does one of these four jobs:
+
+- **Contract.** What a caller must know that the signature cannot say: preconditions, invariants, units, ownership and lifetime, thread-safety, error semantics, side effects. Document the contract of everything a caller outside the file can reach, and nothing else. Goes on the type/method/module as Javadoc, JSDoc, or a docstring.
+- **Why.** The constraint the code satisfies, the bug it avoids, the alternative rejected and the reason.
+- **Hazard.** "Must stay in sync with X", "order matters because Y", "do not remove, it prevents Z".
+- **Map.** A short orientation at the top of a genuinely complex file: what it owns, and what it deliberately does not.
+
+Never write:
+
+- A comment that restates the next line. `// Handle drag start` above `handleDragStart` is noise.
+- Section banners or position markers: `// --- Types ---`, `// Helpers`, `// =====`.
+- Step narration in a function body (`// Step 1:`, `// Then we`). If the steps need labels they need names: extract functions. Numbering a genuinely numbered thing, like a wizard step, is fine.
+- Commented-out code. Delete it.
+- Doc tags that restate the signature. `@param blob - The blob` says nothing; omit the tag rather than pad it.
+- Docs on self-explanatory members with no constraint to state.
+
+Two tests before keeping a comment:
+
+- **Delete it.** Is any information lost? If not, it stays deleted.
+- **Could a name carry it instead?** A better identifier, an extracted function, or a named constant beats a comment. Prefer the code change.
+
+A comment at the end of a line usually decodes that line, and that is worth keeping: `{0x25, 0x50} // "%PDF"`, `50L * 1024 * 1024 // 50 MB`. The rules that compare a comment against the code below it do not apply there, but a trailing TODO or a trailing bit of history is judged like any other.
+
+A reference is supplementary, never load-bearing: the comment must survive deleting it. `// See #1234` is a dead end; `// saving first loses every annotation (#6865)` is not. Prefer a spec (`RFC 3161`) or CVE where one applies.
+
+A TODO needs an issue, not an owner: `// TODO(#1234): re-enable the gate once account syncing lands`. If it is not worth an issue, it is not worth a TODO. A question is not a TODO.
+
+A comment block over ~12 lines outside a file or type header usually means the code needs restructuring, or that the prose is product documentation and belongs in the docs repo.
+
+`task comment-lint` checks the mechanical part of this on the lines you add, and runs inside `task pre-commit`. Reasoning, worked examples and the linter's own rules: @devGuide/CODE_COMMENTS.md
+
 ## Common Development Commands

 ### Build and Test
@@ -70,7 +107,7 @@ The project structure is defined in `engine/pyproject.toml`. Any new dependencie
 - Avoid nested functions and nested classes unless the language construct requires them.
 - Prefer composition to inheritance when combining concepts.
 - Avoid speculative abstractions. Add a layer only when it removes real duplication or clarifies lifecycle.
- Add comments sparingly and only when they explain non-obvious intent.
+- Comments follow the repo-wide rules in the "Comments" section above.

 #### Python Typing and Models
 - Deserialize into Pydantic models as early as possible.
@@ -42,6 +42,7 @@ Please make sure your Pull Request adheres to the following guidelines:
 - Keep commits atomic. One commit should contain one change. If you want to make multiple changes, submit multiple Pull Requests.
 - Commits should be clear, concise, and easy to understand.
 - References to the Issue number in the Pull Request and/or Commit message.
+- Every comment in the diff should say something the code does not. See [Code comments](devGuide/CODE_COMMENTS.md); `task comment-lint` checks the mechanical part.

 ## Translations

@@ -266,6 +266,20 @@ tasks:
    cmds:
      - task: frontend:lint
      - task: engine:lint
+      - task: comment-lint
+
+  comment-lint:
+    desc: "Check comment quality on the lines this branch adds"
+    aliases: [comments]
+    cmds:
+      - task: pre-commit:comment-lint
+        vars: { CLI_ARGS: '{{.CLI_ARGS}}' }
+
+  comment-lint:branch:
+    desc: "Check comment quality on everything this branch adds over its base"
+    cmds:
+      - task: pre-commit:comment-lint:branch
+        vars: { BASE: '{{.BASE}}' }

  fix:
    desc: "Auto-fix all components"
@@ -48,7 +48,7 @@ public class EndpointConfiguration {
    private final ApplicationProperties applicationProperties;
    @Getter private Map<String, Boolean> endpointStatuses = new ConcurrentHashMap<>();
    private Map<String, Set<String>> endpointGroups = new ConcurrentHashMap<>();
-    private Set<String> disabledGroups = new HashSet<>();
+    private Set<String> disabledGroups = ConcurrentHashMap.newKeySet();
    private Map<String, DisableReason> endpointDisableReasons = new ConcurrentHashMap<>();
    private Map<String, DisableReason> groupDisableReasons = new ConcurrentHashMap<>();
    private Map<String, Set<String>> endpointAlternatives = new ConcurrentHashMap<>();
@@ -237,7 +237,7 @@ public class TabulaTableParser implements TableParser {
            score -= 0.3f;
        }

-        return Math.max(0f, Math.min(1f, score));
+        return Math.clamp(score, 0f, 1f);
    }

    private Bounds tableBounds(Table table) {
@@ -15,7 +15,8 @@ public class StringToMapPropertyEditor extends PropertyEditorSupport {
    @Override
    public void setAsText(String text) throws IllegalArgumentException {
        try {
-            TypeReference<HashMap<String, String>> typeRef = new TypeReference<>() {};
+            TypeReference<HashMap<String, String>> typeRef =
+                    new TypeReference<HashMap<String, String>>() {};
            Map<String, String> map = objectMapper.readValue(text, typeRef);
            setValue(map);
        } catch (Exception e) {
@@ -113,6 +113,16 @@ class RequestUriUtilsTest {
        assertTrue(RequestUriUtils.isFrontendRoute("", "/split-pdf"));
    }

+    @Test
+    void testIsFrontendRoute_editorRouteOwnedByFrontend() {
+        // /editor (and its tool routes) is an SPA route: a direct-nav/refresh must
+        // serve index.html, not the auth filter's 302-to-/login. Regression test for
+        // the editor moving from / to /editor, whose refresh bounced processor users
+        // to the processor because the redirect dropped the return path.
+        assertTrue(RequestUriUtils.isFrontendRoute("", "/editor"));
+        assertTrue(RequestUriUtils.isFrontendRoute("/app", "/app/editor"));
+    }
+
    @Test
    void testIsFrontendRoute_filesRouteOwnedByFrontend() {
        // /files and /files/<folder-uuid> are FileManagerView routes - they
@@ -306,6 +306,9 @@ tasks.register('copyFrontendAssets', Copy) {
        // Exclude files that conflict with backend static resources
        exclude 'robots.txt'  // Backend already has this
        exclude 'favicon.ico' // Backend already has this
+        // Backend ships its own NotoSans-Regular.ttf here and it is git-tracked;
+        // letting the editor's copy win would dirty the source tree on every build.
+        exclude 'fonts/NotoSans-Regular.ttf'
    }
    into resourcesStaticDir
    duplicatesStrategy = DuplicatesStrategy.INCLUDE  // Let frontend overwrite when needed
@@ -237,7 +237,7 @@ public class EditTextController {

        Matcher matcher = edit.pattern().matcher(joined);
        List<MatchSpan> spans = new ArrayList<>();
-        StringBuffer interpolation = new StringBuffer();
+        StringBuilder interpolation = new StringBuilder();
        int previousAppendPosition = 0;
        while (matcher.find()) {
            if (matcher.start() == matcher.end()) {
@@ -0,0 +1,598 @@
+package stirling.software.SPDF.controller.api;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Base64;
+import java.util.List;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDResources;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.annotation.PostMapping;
+import org.springframework.web.bind.annotation.RequestBody;
+
+import com.fasterxml.jackson.annotation.JsonInclude;
+
+import io.swagger.v3.oas.annotations.Operation;
+
+import lombok.Data;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+
+import stirling.software.common.annotations.api.GeneralApi;
+import stirling.software.common.service.CustomPDFDocumentFactory;
+
+/**
+ * Charcode-encode helper for the v2 PDF text editor.
+ *
+ * <p>The frontend editor uses PDFium-WASM, which exposes {@code FPDFText_SetCharcodes} for writing
+ * new text using raw font charcodes (skipping PDFium's broken reverse Unicode→CID lookup for
+ * embedded subset fonts). What PDFium does NOT expose is the byte-encoding side of an existing font
+ * - given a PDFont and a Unicode string, what are the bytes the font's encoding produces? PDFBox
+ * does have that ({@link PDFont#encode}).
+ *
+ * <p>This endpoint accepts the source PDF + a "locator" describing where to find the font in
+ * question (page index + a sample char known to render in the target font, optionally narrowed by
+ * the font's /BaseFont name) + the Unicode text the frontend wants to encode. It returns the
+ * charcode sequence the frontend can pass to {@code FPDFText_SetCharcodes}.
+ *
+ * <p>If the locator can't find a matching text fragment, or if the font can't encode some chars,
+ * the response reports which chars are missing so the frontend can fall back to Helvetica per char.
+ */
+@Slf4j
+@GeneralApi
+@RequiredArgsConstructor
+public class PdfTextEditorCharcodeController {
+
+    /** Reject JSON bodies whose base64 implies a decoded PDF larger than this. */
+    private static final int MAX_PDF_BYTES = 100 * 1024 * 1024;
+
+    /**
+     * Upper bound on {@code request.text} code units. Editor requests are word-sized; an unbounded
+     * text drove a per-code-point encode/exception loop (CPU burn) on crafted requests.
+     */
+    private static final int MAX_TEXT_CHARS = 4096;
+
+    /** Nested form-XObject resource dictionaries visited per lookup (cycle/DoS guard). */
+    private static final int MAX_RESOURCE_DICTS = 32;
+
+    /** Bound on the reverse-map cache so a busy multi-document server can't grow it forever. */
+    private static final int REVERSE_MAP_CACHE_MAX = 32;
+
+    /** Access-ordered LRU bounded at {@link #REVERSE_MAP_CACHE_MAX} entries. */
+    private static final class BoundedReverseMapCache
+            extends java.util.LinkedHashMap<String, java.util.Map<String, Long>> {
+        private static final long serialVersionUID = 1L;
+
+        BoundedReverseMapCache() {
+            super(16, 0.75f, true);
+        }
+
+        @Override
+        protected boolean removeEldestEntry(
+                java.util.Map.Entry<String, java.util.Map<String, Long>> eldest) {
+            return size() > REVERSE_MAP_CACHE_MAX;
+        }
+    }
+
+    private static final java.util.Map<String, java.util.Map<String, Long>> REVERSE_MAP_CACHE =
+            java.util.Collections.synchronizedMap(new BoundedReverseMapCache());
+
+    private final CustomPDFDocumentFactory pdfDocumentFactory;
+
+    // NOTE: PDFBox's PDSimpleFont emits one "No Unicode mapping for .notdef" WARN per probed
+    // charcode when buildReverseUnicodeMap iterates 0..0xFFFF, which once flooded info.log to
+    // ~1.4 GB overnight. That logger is silenced DECLARATIVELY in logback.xml (a config entry ops
+    // can see and revert) rather than by mutating the global logger from a static block here -
+    // mutating it at class-load time hid the same warnings from every other tool in the JVM with
+    // no trace in configuration.
+
+    @Data
+    public static class EncodeCharcodesRequest {
+
+        /** Base64-encoded original PDF. The frontend already has the bytes loaded. */
+        private String pdfBase64;
+
+        /** 0-based page index containing the font sample. */
+        private int pageIndex;
+
+        /**
+         * A char known to exist on the page in the target font. Combined with {@code fontName}
+         * (when supplied) it locates the source PDFont via its ToUnicode CMap.
+         */
+        private String locatorChar;
+
+        /**
+         * Optional /BaseFont name of the target font (as PDFium's FPDFFont_GetBaseFontName reports
+         * it). When a page has TWO fonts that both render {@code locatorChar}, this disambiguates
+         * which one to encode against - otherwise the first font found wins and a cross-font edit
+         * gets the wrong font's charcode. Null = keep the legacy first-match behaviour.
+         */
+        private String fontName;
+
+        /**
+         * Optional SHA-256 (lowercase hex) of the target font's embedded program bytes (what
+         * PDFium's FPDFFont_GetFontData returns = the decoded FontFile/FontFile2/FontFile3 stream).
+         * This is the ONLY unambiguous font identity: PDFium strips the "ABCDEF+" subset tag from
+         * font names, so every subset of one family reports the same {@code fontName} and a
+         * name-based lookup can land on a SIBLING subset whose charcode space is different -
+         * returning valid-but-wrong charcodes that scramble the edited text. When present and a
+         * font on the page matches, it wins over name matching.
+         */
+        private String fontSha256;
+
+        /** Unicode text the frontend wants to encode. */
+        private String text;
+    }
+
+    @Data
+    @JsonInclude(JsonInclude.Include.NON_NULL)
+    public static class EncodeCharcodesResponse {
+        /**
+         * Per-char charcode array (one entry per code point in {@code request.text}). When the
+         * font's encoding produces multi-byte sequences, each char gets the full unsigned int value
+         * of its bytes packed big-endian (so a 2-byte CID like 0x004D becomes 77).
+         */
+        private List<Long> charcodes;
+
+        /** Chars from the request that the font couldn't encode. */
+        private List<String> missing;
+
+        /** Diagnostic note - included so the frontend HUD can show what happened. */
+        private String note;
+
+        /** Set when the request failed entirely (bad pdf bytes, no matching font, etc.). */
+        private String error;
+    }
+
+    @Operation(
+            summary = "Encode Unicode → font charcodes for the v2 PDF text editor",
+            description =
+                    """
+                    Frontend-only helper: takes the source PDF, a locator pointing at an existing
+                    char rendered in the target font, and a Unicode string. Returns the byte
+                    sequence the target font produces for that Unicode, packed as one unsigned
+                    int per char. The frontend then calls FPDFText_SetCharcodes with the
+                    returned ints to inject new text that reuses the embedded font's actual
+                    glyphs. Chars the font can't encode are listed in `missing` so the caller
+                    can fall back per-char.
+                    """)
+    @PostMapping(
+            value = "/pdf-text-editor/encode-charcodes",
+            consumes = "application/json",
+            produces = "application/json")
+    public ResponseEntity<EncodeCharcodesResponse> encodeCharcodes(
+            @RequestBody EncodeCharcodesRequest request) {
+        EncodeCharcodesResponse resp = new EncodeCharcodesResponse();
+        if (request == null
+                || request.getPdfBase64() == null
+                || request.getText() == null
+                || request.getLocatorChar() == null) {
+            resp.setError("missing required fields");
+            return ResponseEntity.badRequest().body(resp);
+        }
+        // length/4*3 bounds the decoded size without decoding, so we reject early before
+        // allocating.
+        String b64 = request.getPdfBase64();
+        if ((long) b64.length() / 4 * 3 > MAX_PDF_BYTES) {
+            resp.setError("pdf too large");
+            return ResponseEntity.status(413).body(resp);
+        }
+        // Reported separately: a combined check names only one cause and misleads the caller.
+        if (request.getText().length() > MAX_TEXT_CHARS) {
+            resp.setError("text too long");
+            return ResponseEntity.badRequest().body(resp);
+        }
+        if (request.getLocatorChar().length() > 4) {
+            resp.setError("locatorChar too long");
+            return ResponseEntity.badRequest().body(resp);
+        }
+        byte[] pdfBytes;
+        try {
+            pdfBytes = Base64.getDecoder().decode(b64);
+        } catch (IllegalArgumentException e) {
+            resp.setError("pdfBase64 is not valid base64");
+            return ResponseEntity.badRequest().body(resp);
+        }
+        try (PDDocument doc = pdfDocumentFactory.load(pdfBytes, true)) {
+            if (request.getPageIndex() < 0 || request.getPageIndex() >= doc.getNumberOfPages()) {
+                resp.setError("pageIndex out of range");
+                return ResponseEntity.badRequest().body(resp);
+            }
+            PDPage page = doc.getPage(request.getPageIndex());
+            // Skip walking the page's content stream (it crashes on Type3 fonts with
+            // UnsupportedOperationException("Not implemented: Type3") before we can do anything
+            // useful). Instead enumerate the page's font resources and pick the one identified by
+            // the request's font-program hash (definitive), falling back to name matching.
+            // For Chrome/Skia-printed PDFs that emit one Type3 font per glyph, this lands on
+            // the exact font that renders the locator char.
+            ResourceFont located =
+                    findFontByToUnicode(
+                            page,
+                            request.getLocatorChar(),
+                            request.getFontName(),
+                            request.getFontSha256(),
+                            doc);
+            if (located == null) {
+                resp.setError(
+                        "no font on page "
+                                + request.getPageIndex()
+                                + " renders locatorChar="
+                                + request.getLocatorChar()
+                                + (request.getFontName() != null
+                                        ? " (fontName=" + request.getFontName() + ")"
+                                        : ""));
+                return ResponseEntity.ok(resp);
+            }
+            // Build a reverse Unicode→charcode map by walking the font's ToUnicode CMap.
+            // This is the ONLY path that works for Type3 fonts (PDFBox's font.encode() throws
+            // "Not implemented: Type3" on them), and it also acts as a more reliable fallback
+            // for subset fonts whose encode() rejects chars not in the original document.
+            //
+            // For Sample.pdf specifically, every embedded font is Type3 (Chrome/Skia output),
+            // but they all carry a ToUnicode CMap mapping CIDs back to Unicode. We iterate
+            // charcodes 0..0xFFFF, call font.toUnicode(cc) for each, and record the inverse
+            // mapping for the chars the user wants to write.
+            PDFont font = located.font();
+            java.util.Map<String, Long> reverseMap =
+                    buildReverseUnicodeMap(pdfBytes, located, request.getPageIndex());
+            List<Long> charcodes = new ArrayList<>();
+            List<String> missing = new ArrayList<>();
+            String text = request.getText();
+            int i = 0;
+            while (i < text.length()) {
+                int cp = text.codePointAt(i);
+                String oneChar = new String(Character.toChars(cp));
+                i += Character.charCount(cp);
+                // Whitespace is NEVER charcode-reused. Subset Type1/LaTeX fonts
+                // usually have no real space glyph, yet font.encode(0x20) still
+                // returns code 0x20 without throwing - and SetCharcodes(0x20)
+                // then paints whatever glyph sits at that subset code (e.g. „
+                // quotedblbase in LMRoman). Report whitespace as missing so the
+                // frontend emits it as a positional gap instead.
+                if (Character.isWhitespace(cp)) {
+                    missing.add(oneChar);
+                    continue;
+                }
+                // 1st try: font.encode() - works for Type0/TrueType/Type1
+                Long packed = null;
+                try {
+                    byte[] encoded = font.encode(oneChar);
+                    long p = 0L;
+                    for (byte b : encoded) p = (p << 8) | (b & 0xff);
+                    packed = p;
+                } catch (IOException
+                        | IllegalArgumentException
+                        | UnsupportedOperationException encodeEx) {
+                    // 2nd try: ToUnicode reverse lookup - works for Type3 + anything with a CMap
+                    packed = reverseMap.get(oneChar);
+                }
+                if (packed != null) charcodes.add(packed);
+                else missing.add(oneChar);
+            }
+            resp.setCharcodes(charcodes);
+            if (!missing.isEmpty()) resp.setMissing(missing);
+            resp.setNote(
+                    "font="
+                            + font.getName()
+                            + " encoded "
+                            + charcodes.size()
+                            + " of "
+                            + (charcodes.size() + missing.size())
+                            + " chars");
+            return ResponseEntity.ok(resp);
+        } catch (IOException e) {
+            log.warn("encodeCharcodes: failed to load PDF", e);
+            resp.setError("failed to load PDF");
+            return ResponseEntity.badRequest().body(resp);
+        } catch (RuntimeException e) {
+            log.warn("encodeCharcodes: unexpected error", e);
+            resp.setError("unexpected error");
+            return ResponseEntity.status(500).body(resp);
+        }
+    }
+
+    /**
+     * Locate the font the request targets. Identity sources, strongest first:
+     *
+     * <ol>
+     *   <li><b>Program hash</b>: SHA-256 of the embedded font program bytes. Definitive - two
+     *       different subsets NEVER share program bytes, and PDFium's FPDFFont_GetFontData returns
+     *       exactly the decoded FontFile stream, so frontend and backend hash the same bytes.
+     *   <li><b>Exact /BaseFont name</b> (subset tag included), then <b>tag-stripped name</b>. Name
+     *       matches are only accepted when UNAMBIGUOUS: PDFium reports subset fonts WITHOUT their
+     *       "ABCDEF+" tag, so a page with several subsets of one family ("AAAAAC+Garamond",
+     *       "AAAAAG+Garamond", ...) has them ALL match the stripped name - and encoding against the
+     *       wrong sibling returns valid-but-wrong charcodes that scramble the edited text ("RUSSELL
+     *       W. MANGUM" rendered "US EEL W. MANGS M"). With 2+ candidates we return null so the
+     *       frontend takes its safe fallback instead of a coin flip.
+     * </ol>
+     *
+     * <p>This avoids running PDFStreamEngine.processPage, which throws
+     * UnsupportedOperationException on Type3 font glyph rendering. The PDFont lookup itself is
+     * purely metadata-driven and works on all subtypes.
+     */
+    private static ResourceFont findFontByToUnicode(
+            PDPage page, String wantChar, String fontName, String fontSha256, PDDocument doc) {
+        try {
+            List<ResourceFont> fonts = collectResourceTreeFonts(page.getResources());
+
+            // 1) Program-hash identity. When several dicts share one program (identical bytes
+            // re-embedded), any of them renders the same glyphs for the same codes; prefer the
+            // one whose ToUnicode covers the locator char so the reverse map is usable.
+            if (fontSha256 != null && !fontSha256.isEmpty()) {
+                List<ResourceFont> hashMatches = new ArrayList<>();
+                for (ResourceFont rf : fonts) {
+                    String sha = fontProgramSha256(rf.font());
+                    if (fontSha256.equalsIgnoreCase(sha)) hashMatches.add(rf);
+                }
+                for (ResourceFont rf : hashMatches) {
+                    if (probesToUnicode(rf.font(), wantChar)) return rf;
+                }
+                if (!hashMatches.isEmpty()) return hashMatches.get(0);
+                // No program on this page hashes to what the frontend is editing (e.g. PDFium
+                // returned a substitute font's bytes for a non-embedded font). Fall through to
+                // name matching rather than failing outright.
+            }
+
+            // 2) Name identity - exact tag-included first, then tag-stripped - each accepted
+            // only when it selects a single font.
+            if (fontName != null && !fontName.isEmpty()) {
+                ResourceFont exact =
+                        selectUnambiguous(
+                                fonts, wantChar, f -> fontName.equals(f.getName()), "exact");
+                if (exact != null) return exact;
+                String wantStripped = stripSubsetTag(fontName);
+                ResourceFont stripped =
+                        selectUnambiguous(
+                                fonts,
+                                wantChar,
+                                f -> wantStripped.equals(stripSubsetTag(f.getName())),
+                                "stripped");
+                if (stripped != null) return stripped;
+                // The frontend NAMED the font it is editing. Falling back to "any font that
+                // renders the char" would hand back a DIFFERENT font's charcodes, which the
+                // frontend then writes into the named font's text object - wrong glyph, and the
+                // backend strategy skips all frontend validation. Report the char missing
+                // instead so the caller takes its own fallback path.
+                return null;
+            }
+
+            // 3) Legacy locator-only behaviour: first font whose ToUnicode renders the char.
+            for (ResourceFont rf : fonts) {
+                if (probesToUnicode(rf.font(), wantChar)) return rf;
+            }
+        } catch (RuntimeException ignore) {
+            // Be defensive: any single bad font shouldn't sink the whole request.
+        }
+        return null;
+    }
+
+    /**
+     * Apply {@code nameFilter}, then decide: exactly one candidate whose ToUnicode covers {@code
+     * wantChar} wins; two+ probe-hits are AMBIGUOUS (null). With zero probe-hits, a single
+     * name-matching font is still returned (font.encode() may handle chars without a ToUnicode -
+     * common for Type0/Identity-H), but two+ name matches are again ambiguous.
+     */
+    private static ResourceFont selectUnambiguous(
+            List<ResourceFont> fonts,
+            String wantChar,
+            java.util.function.Predicate<PDFont> nameFilter,
+            String modeLabel) {
+        List<ResourceFont> named = new ArrayList<>();
+        for (ResourceFont rf : fonts) {
+            try {
+                if (rf.font().getName() != null && nameFilter.test(rf.font())) named.add(rf);
+            } catch (RuntimeException ignore) {
+            }
+        }
+        if (named.isEmpty()) return null;
+        List<ResourceFont> probed = new ArrayList<>();
+        for (ResourceFont rf : named) {
+            if (probesToUnicode(rf.font(), wantChar)) probed.add(rf);
+        }
+        if (probed.size() == 1) return probed.get(0);
+        if (probed.size() > 1) {
+            log.debug(
+                    "encodeCharcodes: {} name match ambiguous ({} fonts render locator '{}') -"
+                            + " refusing cross-subset guess",
+                    modeLabel,
+                    probed.size(),
+                    wantChar);
+            return null;
+        }
+        return named.size() == 1 ? named.get(0) : null;
+    }
+
+    /** True when some charcode in the font's ToUnicode CMap maps to {@code wantChar}. */
+    private static boolean probesToUnicode(PDFont font, String wantChar) {
+        // Cheap inverse-CMap probe: iterate codes until we hit one whose toUnicode is wantChar.
+        // For Type3 with at most ~16 glyphs, this is microseconds. For full Type0 subsets
+        // it's a few-thousand-iteration scan.
+        int upper = font.isStandard14() ? 256 : 0x10000;
+        for (int cc = 0; cc < upper; cc++) {
+            String u;
+            try {
+                u = font.toUnicode(cc);
+            } catch (Exception ignore) {
+                continue;
+            }
+            if (u != null && u.equals(wantChar)) return true;
+        }
+        return false;
+    }
+
+    private record ResourceFont(PDFont font, String path) {}
+
+    private record PendingResources(PDResources resources, String path) {}
+
+    /**
+     * Breadth-first collection of every distinct font reachable from the page's resources AND every
+     * nested form XObject's resources (bounded by {@link #MAX_RESOURCE_DICTS}, cycle-safe, deduped
+     * by COS dictionary identity). The v2 reader surfaces form-XObject text as editable, so its
+     * fonts must be findable too.
+     */
+    private static List<ResourceFont> collectResourceTreeFonts(PDResources resources) {
+        List<ResourceFont> out = new ArrayList<>();
+        java.util.ArrayDeque<PendingResources> queue = new java.util.ArrayDeque<>();
+        java.util.Set<org.apache.pdfbox.cos.COSDictionary> seenDicts =
+                java.util.Collections.newSetFromMap(new java.util.IdentityHashMap<>());
+        java.util.Set<org.apache.pdfbox.cos.COSDictionary> seenFonts =
+                java.util.Collections.newSetFromMap(new java.util.IdentityHashMap<>());
+        if (resources != null) queue.add(new PendingResources(resources, ""));
+        int visited = 0;
+        // Bound a crafted page declaring many fonts none of which match (CPU-DoS guard).
+        final int MAX_FONTS = 64;
+        while (!queue.isEmpty() && visited < MAX_RESOURCE_DICTS) {
+            PendingResources pending = queue.poll();
+            PDResources res = pending.resources();
+            if (!seenDicts.add(res.getCOSObject())) continue;
+            visited++;
+            for (org.apache.pdfbox.cos.COSName name : res.getFontNames()) {
+                if (out.size() >= MAX_FONTS) break;
+                PDFont font;
+                try {
+                    font = res.getFont(name);
+                } catch (IOException | RuntimeException e) {
+                    continue;
+                }
+                if (font == null || !seenFonts.add(font.getCOSObject())) continue;
+                out.add(new ResourceFont(font, pending.path() + "/" + name.getName()));
+            }
+            try {
+                for (org.apache.pdfbox.cos.COSName xn : res.getXObjectNames()) {
+                    try {
+                        org.apache.pdfbox.pdmodel.graphics.PDXObject xo = res.getXObject(xn);
+                        if (xo
+                                instanceof
+                                org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject form) {
+                            PDResources fr = form.getResources();
+                            if (fr != null) {
+                                queue.add(
+                                        new PendingResources(
+                                                fr, pending.path() + "/" + xn.getName()));
+                            }
+                        }
+                    } catch (IOException | RuntimeException ignore) {
+                    }
+                }
+            } catch (RuntimeException ignore) {
+            }
+        }
+        return out;
+    }
+
+    /**
+     * SHA-256 (lowercase hex) of a font's embedded program bytes - the decoded
+     * FontFile/FontFile2/FontFile3 stream, which is byte-identical to what PDFium's
+     * FPDFFont_GetFontData hands the frontend. Null when the font embeds no program.
+     */
+    private static String fontProgramSha256(PDFont font) {
+        try {
+            org.apache.pdfbox.pdmodel.font.PDFontDescriptor fd = font.getFontDescriptor();
+            if (fd == null && font instanceof org.apache.pdfbox.pdmodel.font.PDType0Font type0) {
+                fd = type0.getDescendantFont().getFontDescriptor();
+            }
+            if (fd == null) return null;
+            org.apache.pdfbox.pdmodel.common.PDStream stream = fd.getFontFile2();
+            if (stream == null) stream = fd.getFontFile3();
+            if (stream == null) stream = fd.getFontFile();
+            if (stream == null) return null;
+            return sha256Hex(stream.toByteArray());
+        } catch (IOException | RuntimeException e) {
+            return null;
+        }
+    }
+
+    /** Drop the 6-letter "ABCDEF+" subset prefix PDF puts on subset /BaseFont names. */
+    private static String stripSubsetTag(String fontName) {
+        if (fontName == null) return null;
+        if (fontName.length() > 7
+                && fontName.charAt(6) == '+'
+                && fontName.chars().limit(6).allMatch(c -> c >= 'A' && c <= 'Z')) {
+            return fontName.substring(7);
+        }
+        return fontName;
+    }
+
+    /**
+     * Build a Unicode→charcode map for a font by iterating every charcode in 0..0xFFFF and asking
+     * the font's ToUnicode CMap what Unicode it maps to. Charcodes that aren't in the CMap throw
+     * inside toUnicode (PDFBox returns null or throws depending on font subtype), and those are
+     * skipped silently.
+     *
+     * <p>This is the encoding inverse PDFBox doesn't expose directly. For Type3 fonts (where
+     * font.encode() throws "Not implemented"), this is the ONLY way to write text in the same font
+     * - we look up the user's char in the reverse map and pass that charcode to
+     * FPDFText_SetCharcodes on the frontend.
+     *
+     * <p>The 0..0xFFFF range is sufficient for Type0/CIDFontType2 fonts (CIDs are 16-bit). For
+     * single-byte fonts the loop short-circuits after 256. We don't go higher because no PDF font
+     * has a CID outside that range in practice; the per-font result is memoised in {@link
+     * #REVERSE_MAP_CACHE} so the 65 536-entry probe runs once per document+font, not per request.
+     */
+    private static java.util.Map<String, Long> buildReverseUnicodeMap(
+            byte[] pdfBytes, ResourceFont located, int pageIndex) {
+        String key = sha256Hex(pdfBytes) + "|" + fontCacheIdentity(located, pageIndex);
+        // Compound get/put under the map's own monitor. The 0..0xFFFF probe runs OUTSIDE the
+        // lock so one slow build can't block every other request on the shared cache.
+        java.util.Map<String, Long> cached;
+        synchronized (REVERSE_MAP_CACHE) {
+            cached = REVERSE_MAP_CACHE.get(key);
+        }
+        if (cached != null) return cached;
+        java.util.Map<String, Long> built = computeReverseUnicodeMap(located.font());
+        synchronized (REVERSE_MAP_CACHE) {
+            java.util.Map<String, Long> raced = REVERSE_MAP_CACHE.putIfAbsent(key, built);
+            return raced != null ? raced : built;
+        }
+    }
+
+    private static String fontCacheIdentity(ResourceFont located, int pageIndex) {
+        org.apache.pdfbox.cos.COSObjectKey objectKey = null;
+        try {
+            objectKey = located.font().getCOSObject().getKey();
+        } catch (RuntimeException ignore) {
+        }
+        if (objectKey != null) {
+            return "obj|" + objectKey.getNumber() + "." + objectKey.getGeneration();
+        }
+        return "res|p" + pageIndex + located.path();
+    }
+
+    /** Lowercase hex SHA-256 of the PDF bytes; used as the reverse-map cache key. */
+    private static String sha256Hex(byte[] bytes) {
+        try {
+            byte[] digest = java.security.MessageDigest.getInstance("SHA-256").digest(bytes);
+            StringBuilder sb = new StringBuilder(digest.length * 2);
+            for (byte b : digest) {
+                sb.append(Character.forDigit((b >> 4) & 0xf, 16));
+                sb.append(Character.forDigit(b & 0xf, 16));
+            }
+            return sb.toString();
+        } catch (java.security.NoSuchAlgorithmException e) {
+            // SHA-256 is always present in a JRE; fall back to a length+hash key just in case so
+            // the cache still functions (correctness holds - collisions only cost a rebuild).
+            return bytes.length + ":" + java.util.Arrays.hashCode(bytes);
+        }
+    }
+
+    private static java.util.Map<String, Long> computeReverseUnicodeMap(PDFont font) {
+        java.util.Map<String, Long> out = new java.util.HashMap<>();
+        int upper = font.isStandard14() ? 256 : 0x10000;
+        for (int cc = 0; cc < upper; cc++) {
+            String u;
+            try {
+                u = font.toUnicode(cc);
+            } catch (Exception ignore) {
+                continue;
+            }
+            if (u == null || u.isEmpty()) continue;
+            // First charcode wins for a given Unicode (the canonical mapping).
+            out.putIfAbsent(u, (long) cc);
+        }
+        return out;
+    }
+}
@@ -95,7 +95,8 @@ public class UIDataController {

        try (InputStream is = resource.getInputStream()) {
            Map<String, List<Dependency>> licenseData =
-                    objectMapper.readValue(is, new TypeReference<>() {});
+                    objectMapper.readValue(
+                            is, new TypeReference<Map<String, List<Dependency>>>() {});
            data.setDependencies(licenseData.get("dependencies"));
        } catch (IOException e) {
            log.error("Failed to load licenses data", e);
@@ -25,12 +25,15 @@ final class FormPayloadParser {
    private static final String KEY_VALUE = "value";
    private static final String KEY_DEFAULT_VALUE = "defaultValue";

-    private static final TypeReference<Map<String, Object>> MAP_TYPE = new TypeReference<>() {};
+    private static final TypeReference<Map<String, Object>> MAP_TYPE =
+            new TypeReference<Map<String, Object>>() {};
    private static final TypeReference<List<FormUtils.ModifyFormFieldDefinition>>
-            MODIFY_FIELD_LIST_TYPE = new TypeReference<>() {};
+            MODIFY_FIELD_LIST_TYPE =
+                    new TypeReference<List<FormUtils.ModifyFormFieldDefinition>>() {};
    private static final TypeReference<List<FormUtils.NewFormFieldDefinition>> NEW_FIELD_LIST_TYPE =
            new TypeReference<>() {};
-    private static final TypeReference<List<String>> STRING_LIST_TYPE = new TypeReference<>() {};
+    private static final TypeReference<List<String>> STRING_LIST_TYPE =
+            new TypeReference<List<String>>() {};

    private FormPayloadParser() {}

@@ -96,7 +96,9 @@ public class AddCommentsController {

        List<CommentSpecDto> dtos;
        try {
-            dtos = objectMapper.readValue(commentsJson, new TypeReference<>() {});
+            dtos =
+                    objectMapper.readValue(
+                            commentsJson, new TypeReference<List<CommentSpecDto>>() {});
        } catch (JacksonException e) {
            throw new ResponseStatusException(
                    HttpStatus.BAD_REQUEST, "comments must be a JSON array of CommentSpec objects");
@@ -338,6 +338,19 @@ public class ConfigController {
            // Premium/Enterprise settings
            configData.put("premiumEnabled", applicationProperties.getPremium().isEnabled());

+            // Whether this instance can link a Stirling (SaaS) account at all. The account-link
+            // beans live in :proprietary and are @ConditionalOnProperty on this same key, so when
+            // it is off they are absent and /api/v1/account-link/* returns 404. The frontend cannot
+            // tell that 404 apart from "not linked yet", so it needs this told to it explicitly
+            // before it can prompt anyone to link. Read from the environment rather than
+            // AccountLinkProperties because :core must not depend on :proprietary.
+            configData.put(
+                    "accountLinkAvailable",
+                    applicationContext
+                            .getEnvironment()
+                            .getProperty(
+                                    "stirling.billing.account-link.enabled", Boolean.class, false));
+
            // AI Engine settings
            ApplicationProperties.AiEngine aiEngineConfig = applicationProperties.getAiEngine();
            configData.put("aiEngineEnabled", aiEngineConfig.isEnabled());
@@ -4,7 +4,9 @@ import java.time.Instant;
 import java.time.temporal.ChronoUnit;
 import java.util.Map;
 import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.atomic.AtomicLong;

+import org.springframework.scheduling.annotation.Scheduled;
 import org.springframework.stereotype.Service;

 import lombok.extern.slf4j.Slf4j;
@@ -21,7 +23,7 @@ public class WeeklyActiveUsersService {
    private final Map<String, Instant> activeBrowsers = new ConcurrentHashMap<>();

    // Track total unique browsers seen (overall)
-    private long totalUniqueBrowsers = 0;
+    private final AtomicLong totalUniqueBrowsers = new AtomicLong(0);

    // Application start time
    private final Instant startTime = Instant.now();
@@ -36,12 +38,12 @@ public class WeeklyActiveUsersService {
            return;
        }

-        boolean isNewBrowser = !activeBrowsers.containsKey(browserId);
-        activeBrowsers.put(browserId, Instant.now());
+        Instant now = Instant.now();
+        Instant previous = activeBrowsers.put(browserId, now);

-        if (isNewBrowser) {
-            totalUniqueBrowsers++;
-            log.debug("New browser recorded: {} (Total: {})", browserId, totalUniqueBrowsers);
+        if (previous == null) {
+            long total = totalUniqueBrowsers.incrementAndGet();
+            log.debug("New browser recorded: {} (Total: {})", browserId, total);
        }
    }

@@ -61,7 +63,7 @@ public class WeeklyActiveUsersService {
     * @return Total unique browsers count
     */
    public long getTotalUniqueBrowsers() {
-        return totalUniqueBrowsers;
+        return totalUniqueBrowsers.get();
    }

    /**
@@ -88,7 +90,8 @@ public class WeeklyActiveUsersService {
        activeBrowsers.entrySet().removeIf(entry -> entry.getValue().isBefore(sevenDaysAgo));
    }

-    /** Manual cleanup trigger (can be called by scheduled task if needed) */
+    /** Scheduled cleanup trigger running every hour */
+    @Scheduled(fixedRate = 3600000)
    public void performCleanup() {
        int sizeBefore = activeBrowsers.size();
        cleanupOldEntries();
@@ -154,7 +154,8 @@ public class PdfJsonFontService {
            return "otf";
        }
        if (signature == 0x74746366) {
-            return "cff";
+            log.debug("[FONT-DEBUG] TrueType Collection ('ttcf') font program is unsupported");
+            return null;
        }
        return null;
    }
@@ -175,7 +176,8 @@ public class PdfJsonFontService {
            return "otf";
        }
        if (signature == 0x74746366) {
-            return "cff";
+            log.debug("[FONT-DEBUG] TrueType Collection ('ttcf') FontFile2 is unsupported");
+            return null;
        }
        return null;
    }
@@ -15,26 +15,63 @@
        <encoder>
            <pattern>%d %p %c{1} [%thread] %m%n</pattern>
        </encoder>
-        <rollingPolicy class="ch.qos.logback.core.rolling.TimeBasedRollingPolicy">
-            <fileNamePattern>${LOG_PATH}/auth-%d{yyyy-MM-dd}.log.gz</fileNamePattern>
+        <!-- SizeAndTime, not Time alone: the size trigger is what stops a
+             runaway logger filling the disk (see GENERAL appender note).
+             Archives are gzipped, so 64 MB of them holds far more than a
+             day. Worst case on disk is one 100 MB live file plus the cap. -->
+        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
+            <fileNamePattern>${LOG_PATH}/auth-%d{yyyy-MM-dd}.%i.log.gz</fileNamePattern>
+            <maxFileSize>100MB</maxFileSize>
            <maxHistory>7</maxHistory>
            <totalSizeCap>64MB</totalSizeCap>
        </rollingPolicy>
    </appender>

-    <!-- Rolling File Appender for General Logs -->
+    <!-- Rolling File Appender for General Logs
+
+         Why SizeAndTimeBased + totalSizeCap: a previous build of the v2 PDF
+         text editor's reverse-CMap probe loop triggered PDSimpleFont to emit
+         one "No Unicode mapping for .notdef" WARN per probed charcode per
+         font per request. With TimeBasedRollingPolicy alone there was no
+         size ceiling; info.log grew to 1.4 GB in a single day before the JVM
+         choked. The class-level silencer fixes the specific offender, but
+         this size cap is the defence-in-depth: any future logger that
+         floods unexpectedly will roll + auto-delete instead of starving
+         disk + Jetty threads. -->
    <appender name="GENERAL" class="ch.qos.logback.core.rolling.RollingFileAppender">
        <file>${LOG_PATH}/info.log</file>
        <encoder>
            <pattern>%d %p %c{1} [%thread] %m%n</pattern>
        </encoder>
-        <rollingPolicy class="ch.qos.logback.core.rolling.TimeBasedRollingPolicy">
-            <fileNamePattern>${LOG_PATH}/info-%d{yyyy-MM-dd}.log.gz</fileNamePattern>
+        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
+            <fileNamePattern>${LOG_PATH}/info-%d{yyyy-MM-dd}.%i.log.gz</fileNamePattern>
+            <maxFileSize>100MB</maxFileSize>
            <maxHistory>7</maxHistory>
            <totalSizeCap>256MB</totalSizeCap>
        </rollingPolicy>
    </appender>

+    <!-- Suppress PDFBox PDSimpleFont's per-charcode .notdef WARN.
+
+         Required by the v2 PDF text editor's `buildReverseUnicodeMap`
+         which DELIBERATELY iterates every charcode in 0..0xFFFF to
+         discover the encoding-to-Unicode map of an embedded subset
+         font. For any subset font ~99% of those probes hit .notdef,
+         and the default WARN level for those misses turned info.log
+         into a 1.4 GB monster overnight.
+
+         This declarative logback entry is the SOLE mechanism: it is
+         visible to ops and revertable via configuration. An earlier
+         build also mutated this logger's level from a static block in
+         PdfTextEditorCharcodeController, which silenced the same
+         warnings JVM-wide with no trace in any config file - that
+         static block has been removed in favour of this entry. -->
+    <logger name="org.apache.pdfbox.pdmodel.font.PDSimpleFont"
+            level="ERROR" additivity="false">
+        <appender-ref ref="CONSOLE"/>
+        <appender-ref ref="GENERAL"/>
+    </logger>
+
    <!-- Root Logger -->
    <root level="INFO">
        <appender-ref ref="CONSOLE"/>
@@ -94,7 +94,7 @@
        {
            "moduleName": "com.fasterxml.jackson.core:jackson-core",
            "moduleUrl": "https://github.com/FasterXML/jackson-core",
-            "moduleVersion": "2.22.1",
+            "moduleVersion": "2.22.2",
            "moduleLicense": "The Apache Software License, Version 2.0",
            "moduleLicenseUrl": "https://www.apache.org/licenses/LICENSE-2.0.txt"
        },
@@ -108,7 +108,7 @@
        {
            "moduleName": "com.fasterxml.jackson.core:jackson-databind",
            "moduleUrl": "https://github.com/FasterXML/jackson",
-            "moduleVersion": "2.22.1",
+            "moduleVersion": "2.22.2",
            "moduleLicense": "The Apache Software License, Version 2.0",
            "moduleLicenseUrl": "https://www.apache.org/licenses/LICENSE-2.0.txt"
        },
@@ -143,7 +143,7 @@
        {
            "moduleName": "com.fasterxml.jackson:jackson-bom",
            "moduleUrl": "https://github.com/FasterXML/jackson-bom",
-            "moduleVersion": "2.22.1",
+            "moduleVersion": "2.22.2",
            "moduleLicense": "The Apache Software License, Version 2.0",
            "moduleLicenseUrl": "http://www.apache.org/licenses/LICENSE-2.0.txt"
        },
@@ -440,42 +440,14 @@
        {
            "moduleName": "com.stirling:jpdfium",
            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
-            "moduleLicense": "MIT License",
-            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
-        },
-        {
-            "moduleName": "com.stirling:jpdfium-natives-darwin-arm64",
-            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
-            "moduleLicense": "MIT License",
-            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
-        },
-        {
-            "moduleName": "com.stirling:jpdfium-natives-darwin-x64",
-            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
-            "moduleLicense": "MIT License",
-            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
-        },
-        {
-            "moduleName": "com.stirling:jpdfium-natives-linux-arm64",
-            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
+            "moduleVersion": "1.1.3",
            "moduleLicense": "MIT License",
            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
        },
        {
            "moduleName": "com.stirling:jpdfium-natives-linux-x64",
            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
-            "moduleLicense": "MIT License",
-            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
-        },
-        {
-            "moduleName": "com.stirling:jpdfium-natives-windows-x64",
-            "moduleUrl": "https://github.com/Stirling-Tools/JPDFium",
-            "moduleVersion": "1.0.4",
+            "moduleVersion": "1.1.3",
            "moduleLicense": "MIT License",
            "moduleLicenseUrl": "https://opensource.org/licenses/MIT"
        },
@@ -521,36 +493,18 @@
            "moduleLicense": "GNU General Public License, version 2 with the GNU Classpath Exception",
            "moduleLicenseUrl": "https://www.gnu.org/software/classpath/license.html"
        },
-        {
-            "moduleName": "com.twelvemonkeys.common:common-image",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.common:common-image",
            "moduleVersion": "3.14.0",
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.common:common-io",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.common:common-io",
            "moduleVersion": "3.14.0",
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.common:common-lang",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.common:common-lang",
            "moduleVersion": "3.14.0",
@@ -569,12 +523,6 @@
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.imageio:imageio-core",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.imageio:imageio-core",
            "moduleVersion": "3.14.0",
@@ -587,12 +535,6 @@
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.imageio:imageio-metadata",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.imageio:imageio-metadata",
            "moduleVersion": "3.14.0",
@@ -605,24 +547,12 @@
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.imageio:imageio-tiff",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.imageio:imageio-tiff",
            "moduleVersion": "3.14.0",
            "moduleLicense": "The BSD License",
            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
        },
-        {
-            "moduleName": "com.twelvemonkeys.imageio:imageio-webp",
-            "moduleVersion": "3.13.1",
-            "moduleLicense": "The BSD License",
-            "moduleLicenseUrl": "https://github.com/haraldk/TwelveMonkeys#license"
-        },
        {
            "moduleName": "com.twelvemonkeys.imageio:imageio-webp",
            "moduleVersion": "3.14.0",
@@ -769,13 +699,6 @@
            "moduleLicense": "The Apache Software License, Version 2.0",
            "moduleLicenseUrl": "http://www.apache.org/licenses/LICENSE-2.0.txt"
        },
-        {
-            "moduleName": "commons-beanutils:commons-beanutils",
-            "moduleUrl": "https://commons.apache.org/proper/commons-beanutils",
-            "moduleVersion": "1.11.0",
-            "moduleLicense": "Apache-2.0",
-            "moduleLicenseUrl": "https://www.apache.org/licenses/LICENSE-2.0.txt"
-        },
        {
            "moduleName": "commons-cli:commons-cli",
            "moduleUrl": "http://commons.apache.org/proper/commons-cli/",
@@ -790,13 +713,6 @@
            "moduleLicense": "Apache-2.0",
            "moduleLicenseUrl": "https://www.apache.org/licenses/LICENSE-2.0.txt"
        },
-        {
-            "moduleName": "commons-collections:commons-collections",
-            "moduleUrl": "http://commons.apache.org/collections/",
-            "moduleVersion": "3.2.2",
-            "moduleLicense": "Apache License, Version 2.0",
-            "moduleLicenseUrl": "http://www.apache.org/licenses/LICENSE-2.0.txt"
-        },
        {
            "moduleName": "commons-io:commons-io",
            "moduleUrl": "https://commons.apache.org/proper/commons-io/",
@@ -1360,13 +1276,6 @@
            "moduleLicense": "Apache-2.0",
            "moduleLicenseUrl": "https://www.apache.org/licenses/LICENSE-2.0.txt"
        },
-        {
-            "moduleName": "org.apache.commons:commons-math3",
-            "moduleUrl": "http://commons.apache.org/proper/commons-math/",
-            "moduleVersion": "3.6.1",
-            "moduleLicense": "Apache License, Version 2.0",
-            "moduleLicenseUrl": "http://www.apache.org/licenses/LICENSE-2.0.txt"
-        },
        {
            "moduleName": "org.apache.commons:commons-text",
            "moduleUrl": "https://commons.apache.org/proper/commons-text",
@@ -57,6 +57,8 @@ class ToolIODeclarationCoverageTest {
                    // documents.
                    "/api/v1/convert/pdf/text-editor",
                    "/api/v1/convert/text-editor/pdf",
+                    // Charcode lookup for the v2 editor: returns glyph mappings, not a document.
+                    "/api/v1/general/pdf-text-editor",
                    // Signing sessions, certificate checks and hardware token enumeration; the
                    // signing tool itself is /api/v1/security/cert-sign, which is declared.
                    "/api/v1/security/cert-sign/sessions",
@@ -0,0 +1,516 @@
+package stirling.software.SPDF.controller.api;
+
+import java.io.ByteArrayOutputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.ArrayList;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import javax.imageio.ImageIO;
+
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.cos.COSName;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.PDResources;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.pdmodel.font.PDFontDescriptor;
+import org.apache.pdfbox.pdmodel.font.PDType0Font;
+import org.apache.pdfbox.pdmodel.font.PDType1Font;
+import org.apache.pdfbox.pdmodel.font.PDType3Font;
+import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
+import org.apache.pdfbox.rendering.PDFRenderer;
+import org.junit.jupiter.api.Disabled;
+import org.junit.jupiter.api.Test;
+
+/**
+ * Probe: what can PDFBox actually do for font ENCODING on real-world PDFs. This is a diagnostic
+ * test (not a regression) - run with --tests PdfBoxFontEncodingProbeTest -i to see stdout.
+ *
+ * <p>Answers these questions:
+ *
+ * <ol>
+ *   <li>Type0/CIDFontType2 subset: can we add a new glyph not in the original subset? (no, encode
+ *       throws IllegalArgumentException).
+ *   <li>Type1: same question.
+ *   <li>TrueType: same question.
+ *   <li>Can we load a fresh TTF via PDType0Font.load(doc, file) and write text with it? (yes,
+ *       primary path).
+ *   <li>Round-trip via getFontStream / re-embed - can it rehabilitate Type3? (no - Type3 has no
+ *       FontFile* program at all).
+ *   <li>What fonts ship with PDFBox / fontbox? (only LiberationSans-Regular.ttf + AFM for the 14
+ *       standard fonts; CFF/Type1 binaries are NOT bundled - Standard14Fonts.getMappedFontName
+ *       redirects unmappable ones to LiberationSans).
+ * </ol>
+ */
+@Disabled(
+        "Diagnostic probe: dumps PDFBox font encoding tables to stdout and asserts nothing. Kept for font debugging; run manually.")
+public class PdfBoxFontEncodingProbeTest {
+
+    private static final Path PROJECT_ROOT =
+            Paths.get(System.getProperty("user.dir")).getParent().getParent();
+
+    private static final Path SAMPLE =
+            PROJECT_ROOT.resolve("frontend/editor/public/samples/Sample.pdf");
+
+    private static final Path[] EXTRA_FIXTURES = {
+        PROJECT_ROOT.resolve("frontend/editor/src/core/tests/test-fixtures/stirling-marketing.pdf"),
+        PROJECT_ROOT.resolve("frontend/editor/src/core/tests/test-fixtures/multi-page-sample.pdf"),
+        PROJECT_ROOT.resolve("frontend/editor/src/core/tests/test-fixtures/big-sample.pdf"),
+        PROJECT_ROOT.resolve("frontend/editor/src/core/tests/test-fixtures/paragraph-sample.pdf"),
+        PROJECT_ROOT.resolve("frontend/editor/src/core/tests/test-fixtures/user-sample.pdf"),
+    };
+
+    /**
+     * Rasterize the Q4b output (Sample.pdf with injected Liberation text) to confirm the new text
+     * actually renders on top of the existing Type3 content.
+     */
+    @Test
+    public void probeRenderInjectedSample() throws IOException {
+        Path liberation =
+                PROJECT_ROOT.resolve(
+                        "app/core/src/main/resources/static/fonts/LiberationSans-Regular.ttf");
+        byte[] pdfBytes = Files.readAllBytes(SAMPLE);
+        ByteArrayOutputStream out = new ByteArrayOutputStream();
+        try (PDDocument doc = Loader.loadPDF(pdfBytes)) {
+            PDPage page = doc.getPage(0);
+            PDType0Font ttf;
+            try (InputStream in = Files.newInputStream(liberation)) {
+                ttf = PDType0Font.load(doc, in, true);
+            }
+            try (PDPageContentStream cs =
+                    new PDPageContentStream(
+                            doc, page, PDPageContentStream.AppendMode.APPEND, true, true)) {
+                cs.beginText();
+                cs.setFont(ttf, 24);
+                cs.newLineAtOffset(50, 120);
+                cs.showText("INJECTED via PDType0Font.load - $@#&Z");
+                cs.endText();
+            }
+            doc.save(out);
+        }
+        // Rasterize page 0 to a PNG so we can eyeball it.
+        try (PDDocument check = Loader.loadPDF(out.toByteArray())) {
+            PDFRenderer renderer = new PDFRenderer(check);
+            java.awt.image.BufferedImage img = renderer.renderImageWithDPI(0, 100);
+            // Build dir, not the repo root: this render is a debugging aid and was
+            // twice committed by accident when it landed in the working tree.
+            Path png =
+                    Paths.get(System.getProperty("user.dir"), "build", "probe-output")
+                            .resolve("pdfbox-probe-q4b-rendered.png");
+            Files.createDirectories(png.getParent());
+            ImageIO.write(img, "PNG", png.toFile());
+            System.out.println(
+                    "Rendered injected sample to "
+                            + png
+                            + " - "
+                            + img.getWidth()
+                            + "x"
+                            + img.getHeight());
+        }
+    }
+
+    /**
+     * Build a PDF in memory that uses a Type0/CIDFontType2 subset font (the kind Word / InDesign /
+     * LibreOffice produce), then probe whether encode() can add a glyph that wasn't in the original
+     * subset.
+     */
+    @Test
+    public void probeType0CIDFontType2Subset() throws IOException {
+        System.out.println(
+                "\n##################################################################\n"
+                        + "Q1 probe: Type0/CIDFontType2 SUBSET can/cannot add new glyphs\n"
+                        + "##################################################################\n");
+        Path liberation =
+                PROJECT_ROOT.resolve(
+                        "app/core/src/main/resources/static/fonts/LiberationSans-Regular.ttf");
+
+        // Build a PDF that contains only "abc" subsetted from LiberationSans.
+        ByteArrayOutputStream baos = new ByteArrayOutputStream();
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+            PDType0Font subset;
+            try (InputStream in = Files.newInputStream(liberation)) {
+                subset = PDType0Font.load(doc, in, true /* embedSubset */);
+            }
+            try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
+                cs.beginText();
+                cs.setFont(subset, 12);
+                cs.newLineAtOffset(100, 700);
+                cs.showText("abc");
+                cs.endText();
+            }
+            doc.save(baos);
+        }
+
+        // Reload the produced PDF and try to add a NEW glyph through the embedded subset font.
+        byte[] subsetPdf = baos.toByteArray();
+        try (PDDocument doc = Loader.loadPDF(subsetPdf)) {
+            PDResources res = doc.getPage(0).getResources();
+            for (COSName fn : res.getFontNames()) {
+                PDFont f = res.getFont(fn);
+                System.out.println(
+                        "  Subset font in saved PDF: "
+                                + f.getName()
+                                + " ("
+                                + f.getClass().getSimpleName()
+                                + ", subType="
+                                + f.getSubType()
+                                + ")");
+                for (String ch : new String[] {"a", "b", "c", "Z", "z", "0", "$", "@", "X", " "}) {
+                    try {
+                        byte[] enc = f.encode(ch);
+                        StringBuilder hex = new StringBuilder();
+                        for (byte b : enc) hex.append(String.format("%02X ", b & 0xff));
+                        System.out.println(
+                                "    encode('" + ch + "') -> [" + hex.toString().trim() + "] OK");
+                    } catch (UnsupportedOperationException uoe) {
+                        System.out.println("    encode('" + ch + "') UNSUPPORTED");
+                    } catch (IllegalArgumentException iae) {
+                        System.out.println(
+                                "    encode('" + ch + "') MISSING - " + iae.getMessage());
+                    } catch (IOException ioe) {
+                        System.out.println("    encode('" + ch + "') IO ERR - " + ioe.getMessage());
+                    }
+                }
+            }
+        }
+    }
+
+    @Test
+    public void probeExtraFixtures() throws IOException {
+        System.out.println(
+                "\n##################################################################\n"
+                        + "Extra fixture font-class probe\n"
+                        + "##################################################################\n");
+        for (Path fixture : EXTRA_FIXTURES) {
+            if (!Files.exists(fixture)) {
+                System.out.println("(missing) " + fixture);
+                continue;
+            }
+            System.out.println("\n=== " + fixture.getFileName() + " ===");
+            byte[] bytes = Files.readAllBytes(fixture);
+            try (PDDocument doc = Loader.loadPDF(bytes)) {
+                Set<COSName> seen = new HashSet<>();
+                for (int p = 0; p < doc.getNumberOfPages(); p++) {
+                    PDPage page = doc.getPage(p);
+                    PDResources res = page.getResources();
+                    if (res == null) continue;
+                    for (COSName name : res.getFontNames()) {
+                        if (!seen.add(name)) continue;
+                        try {
+                            PDFont f = res.getFont(name);
+                            if (f == null) continue;
+                            String fontFile = "none";
+                            PDFontDescriptor d = f.getFontDescriptor();
+                            if (d != null) {
+                                if (d.getFontFile() != null) fontFile = "FontFile";
+                                else if (d.getFontFile2() != null) fontFile = "FontFile2";
+                                else if (d.getFontFile3() != null) fontFile = "FontFile3";
+                            }
+                            String z = "?";
+                            try {
+                                f.encode("Z");
+                                z = "OK";
+                            } catch (UnsupportedOperationException ex) {
+                                z = "UNSUPPORTED";
+                            } catch (IllegalArgumentException ex) {
+                                z = "MISSING";
+                            } catch (IOException ex) {
+                                z = "IO_ERR";
+                            }
+                            System.out.println(
+                                    "  page "
+                                            + p
+                                            + " "
+                                            + name.getName()
+                                            + " -> "
+                                            + f.getName()
+                                            + " "
+                                            + f.getClass().getSimpleName()
+                                            + " ("
+                                            + f.getSubType()
+                                            + ", "
+                                            + fontFile
+                                            + ", embed="
+                                            + f.isEmbedded()
+                                            + ") encode('Z')="
+                                            + z);
+                        } catch (IOException e) {
+                            System.out.println(
+                                    "  page "
+                                            + p
+                                            + " "
+                                            + name.getName()
+                                            + " load failed: "
+                                            + e.getMessage());
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    @Test
+    public void probeAllQuestions() throws IOException {
+        System.out.println(
+                "\n##################################################################\n"
+                        + "PDFBox font-encoding probe (Sample.pdf + bundled fallback fonts)\n"
+                        + "##################################################################\n");
+
+        // Discover every font in Sample.pdf so we have a real-world test set.
+        byte[] pdfBytes = Files.readAllBytes(SAMPLE);
+        try (PDDocument doc = Loader.loadPDF(pdfBytes)) {
+            List<PDFont> allFonts = new ArrayList<>();
+            Set<COSName> seen = new HashSet<>();
+            for (int p = 0; p < doc.getNumberOfPages(); p++) {
+                PDPage page = doc.getPage(p);
+                PDResources res = page.getResources();
+                if (res == null) continue;
+                for (COSName name : res.getFontNames()) {
+                    if (!seen.add(name)) continue;
+                    try {
+                        PDFont f = res.getFont(name);
+                        if (f != null) allFonts.add(f);
+                    } catch (Exception e) {
+                        System.out.println(
+                                "  (skipped " + name.getName() + " - " + e.getMessage() + ")");
+                    }
+                }
+            }
+            System.out.println(
+                    "Discovered " + allFonts.size() + " unique fonts across Sample.pdf:");
+            for (PDFont f : allFonts) {
+                System.out.println(
+                        "  - "
+                                + f.getName()
+                                + " ("
+                                + f.getClass().getSimpleName()
+                                + ", subType="
+                                + f.getSubType()
+                                + ", embedded="
+                                + f.isEmbedded()
+                                + ")");
+            }
+
+            // Q1/Q2/Q3
+            // Try encoding a char that is NEVER in Sample.pdf via each font.
+            // 'Z' is unlikely to be in the subset for most marketing pages.
+            // Try several candidates to surface what each font can/can't add.
+            String[] candidates = {"Z", "$", "@", "#", "Q", "&", "A", "0", "M"};
+            for (PDFont f : allFonts) {
+                System.out.println("\n=== Encode-probe for font: " + f.getName() + " ===");
+                for (String ch : candidates) {
+                    try {
+                        byte[] enc = f.encode(ch);
+                        StringBuilder hex = new StringBuilder();
+                        for (byte b : enc) hex.append(String.format("%02X ", b & 0xff));
+                        System.out.println(
+                                "  encode('" + ch + "') -> [" + hex.toString().trim() + "] OK");
+                    } catch (UnsupportedOperationException uoe) {
+                        System.out.println(
+                                "  encode('" + ch + "') UNSUPPORTED: " + uoe.getMessage());
+                    } catch (IllegalArgumentException iae) {
+                        System.out.println("  encode('" + ch + "') MISSING: " + iae.getMessage());
+                    } catch (IOException ioe) {
+                        System.out.println("  encode('" + ch + "') IO ERR: " + ioe.getMessage());
+                    }
+                }
+            }
+
+            // Q5
+            // For each font, see what's in the FontFile* stream - this is what we'd
+            // have to round-trip through to "rehabilitate" a Type3 font.
+            System.out.println("\n=== FontFile stream availability (Q5) ===");
+            for (PDFont f : allFonts) {
+                String kind = "none";
+                int size = 0;
+                PDFontDescriptor d = f.getFontDescriptor();
+                if (d != null) {
+                    if (d.getFontFile() != null) {
+                        kind = "FontFile (Type1)";
+                        size = streamBytes(d.getFontFile().getCOSObject().createInputStream());
+                    } else if (d.getFontFile2() != null) {
+                        kind = "FontFile2 (TTF)";
+                        size = streamBytes(d.getFontFile2().getCOSObject().createInputStream());
+                    } else if (d.getFontFile3() != null) {
+                        kind = "FontFile3 (CFF/OpenType)";
+                        size = streamBytes(d.getFontFile3().getCOSObject().createInputStream());
+                    }
+                }
+                System.out.println(
+                        "  "
+                                + f.getName()
+                                + " ("
+                                + f.getClass().getSimpleName()
+                                + "): "
+                                + kind
+                                + " ("
+                                + size
+                                + " bytes)");
+                if (f instanceof PDType3Font) {
+                    System.out.println(
+                            "    -> Type3 has CharProc streams, NOT a FontFile binary."
+                                    + " getFontStream() returns null. Round-trip rehab is impossible:");
+                    System.out.println(
+                            "       each glyph is a mini content stream, not a glyph outline in a"
+                                    + " standard font format. We'd need to rasterize each CharProc to"
+                                    + " glyph outlines + build a fresh TTF/CFF from scratch.");
+                }
+            }
+        }
+
+        // Q4: PDType0Font.load(doc, file) round-trip
+        System.out.println("\n=== Q4: load fresh TTF and write text to a fresh PDF ===");
+        Path liberation =
+                PROJECT_ROOT.resolve(
+                        "app/core/src/main/resources/static/fonts/LiberationSans-Regular.ttf");
+        if (!Files.exists(liberation)) {
+            System.out.println("  Liberation TTF not found at " + liberation);
+        } else {
+            try (PDDocument out = new PDDocument()) {
+                PDPage page = new PDPage();
+                out.addPage(page);
+                PDType0Font ttf;
+                try (InputStream in = Files.newInputStream(liberation)) {
+                    ttf = PDType0Font.load(out, in, true /* embedSubset */);
+                }
+                System.out.println(
+                        "  Loaded TTF -> "
+                                + ttf.getName()
+                                + " ("
+                                + ttf.getClass().getSimpleName()
+                                + ")");
+                String testText = "Hello world! 0123 Z $ @";
+                byte[] encoded = ttf.encode(testText);
+                System.out.println(
+                        "  Encoded "
+                                + testText.length()
+                                + " chars -> "
+                                + encoded.length
+                                + " bytes (Identity-H = 2 bytes/glyph)");
+                try (PDPageContentStream cs = new PDPageContentStream(out, page)) {
+                    cs.beginText();
+                    cs.setFont(ttf, 12);
+                    cs.newLineAtOffset(100, 700);
+                    cs.showText(testText);
+                    cs.endText();
+                }
+                ByteArrayOutputStream baos = new ByteArrayOutputStream();
+                out.save(baos);
+                Path tmp = Files.createTempFile("pdfbox-probe-q4-", ".pdf");
+                Files.write(tmp, baos.toByteArray());
+                System.out.println(
+                        "  Wrote fresh-TTF PDF to "
+                                + tmp
+                                + " ("
+                                + baos.size()
+                                + " bytes) - opens cleanly.");
+
+                // Re-load to confirm the new font is embedded properly.
+                try (PDDocument check = Loader.loadPDF(baos.toByteArray())) {
+                    PDResources res = check.getPage(0).getResources();
+                    for (COSName fn : res.getFontNames()) {
+                        PDFont f = res.getFont(fn);
+                        System.out.println(
+                                "    embedded font: "
+                                        + f.getName()
+                                        + " ("
+                                        + f.getClass().getSimpleName()
+                                        + ", embedded="
+                                        + f.isEmbedded()
+                                        + ")");
+                    }
+                }
+            }
+        }
+
+        // Q4b: load TTF into an EXISTING PDF (Sample.pdf) and append text
+        System.out.println(
+                "\n=== Q4b: load TTF into EXISTING Sample.pdf and write text on page 0 ===");
+        try (PDDocument doc = Loader.loadPDF(pdfBytes)) {
+            PDPage page = doc.getPage(0);
+            PDType0Font ttf;
+            try (InputStream in = Files.newInputStream(liberation)) {
+                ttf = PDType0Font.load(doc, in, true);
+            }
+            // append-mode content stream so we don't disturb existing graphics
+            try (PDPageContentStream cs =
+                    new PDPageContentStream(
+                            doc,
+                            page,
+                            PDPageContentStream.AppendMode.APPEND,
+                            true /* compress */,
+                            true /* resetContext */)) {
+                cs.beginText();
+                cs.setFont(ttf, 12);
+                cs.newLineAtOffset(50, 50);
+                cs.showText("Injected via PDType0Font.load - $@#&");
+                cs.endText();
+            }
+            ByteArrayOutputStream baos = new ByteArrayOutputStream();
+            doc.save(baos);
+            Path tmp = Files.createTempFile("pdfbox-probe-q4b-", ".pdf");
+            Files.write(tmp, baos.toByteArray());
+            System.out.println(
+                    "  Wrote injected-text PDF to " + tmp + " (" + baos.size() + " bytes).");
+
+            // Verify by re-reading: how many fonts now on page 0?
+            try (PDDocument check = Loader.loadPDF(baos.toByteArray())) {
+                PDResources res = check.getPage(0).getResources();
+                int count = 0;
+                for (COSName fn : res.getFontNames()) {
+                    PDFont f = res.getFont(fn);
+                    count++;
+                    System.out.println(
+                            "    page-0 font: "
+                                    + fn.getName()
+                                    + " -> "
+                                    + f.getName()
+                                    + " ("
+                                    + f.getClass().getSimpleName()
+                                    + ")");
+                }
+                System.out.println("  Total fonts on page 0: " + count);
+            }
+        }
+
+        // Q6: what fonts ship in PDFBox / fontbox
+        System.out.println("\n=== Q6: bundled fonts (Standard14 redirect probe) ===");
+        for (Standard14Fonts.FontName fn : Standard14Fonts.FontName.values()) {
+            PDType1Font f = new PDType1Font(fn);
+            String mapped = "" + Standard14Fonts.getMappedFontName(fn.getName());
+            System.out.println(
+                    "  Standard14 "
+                            + fn.getName()
+                            + " -> mapped='"
+                            + mapped
+                            + "' name="
+                            + f.getName());
+        }
+        System.out.println(
+                "  (PDFBox bundles ONLY LiberationSans-Regular.ttf as a binary; the AFMs cover"
+                        + " metrics for the 14 standard fonts but rendering Helvetica/Times/Courier"
+                        + " glyphs falls back to LiberationSans glyphs at runtime when no system font"
+                        + " is found.)");
+    }
+
+    private static int streamBytes(InputStream is) {
+        try (InputStream it = is) {
+            ByteArrayOutputStream baos = new ByteArrayOutputStream();
+            byte[] buf = new byte[4096];
+            int n;
+            while ((n = it.read(buf)) >= 0) baos.write(buf, 0, n);
+            return baos.size();
+        } catch (IOException e) {
+            return -1;
+        }
+    }
+}
@@ -0,0 +1,755 @@
+package stirling.software.SPDF.controller.api;
+
+import static org.assertj.core.api.Assertions.assertThat;
+import static org.mockito.Mockito.mock;
+
+import java.io.ByteArrayOutputStream;
+import java.io.InputStream;
+import java.util.Base64;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.font.PDType1Font;
+import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
+import org.junit.jupiter.api.Test;
+import org.springframework.http.ResponseEntity;
+
+import stirling.software.SPDF.controller.api.PdfTextEditorCharcodeController.EncodeCharcodesRequest;
+import stirling.software.SPDF.controller.api.PdfTextEditorCharcodeController.EncodeCharcodesResponse;
+import stirling.software.common.service.CustomPDFDocumentFactory;
+import stirling.software.common.service.PdfMetadataService;
+
+/**
+ * Regression coverage for the v2 text editor "spaces render as „" bug.
+ *
+ * <p>mushroom-life.pdf is a LaTeX document whose embedded LMRoman subset font has NO real space
+ * glyph, yet {@code font.encode(" ")} still returns charcode 0x20 without throwing. Reusing that
+ * code via {@code FPDFText_SetCharcodes} paints whatever glyph sits at subset code 0x20 - the
+ * quotedblbase „. The controller must therefore report whitespace as {@code missing} so the
+ * frontend emits it as a positional gap instead of a reused glyph.
+ */
+class PdfTextEditorCharcodeControllerTest {
+
+    private static PdfTextEditorCharcodeController controller() {
+        return new PdfTextEditorCharcodeController(
+                new CustomPDFDocumentFactory(mock(PdfMetadataService.class)));
+    }
+
+    private static String mushroomBase64() throws Exception {
+        try (InputStream in =
+                PdfTextEditorCharcodeControllerTest.class.getResourceAsStream(
+                        "/pdftexteditor/mushroom-life.pdf")) {
+            assertThat(in).as("mushroom-life.pdf test resource").isNotNull();
+            return Base64.getEncoder().encodeToString(in.readAllBytes());
+        }
+    }
+
+    private static EncodeCharcodesRequest request(String text) throws Exception {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(mushroomBase64());
+        req.setPageIndex(0);
+        // findFontByToUnicode locates the font via the ToUnicode CMap - "M" exists on page 0.
+        req.setLocatorChar("M");
+        req.setText(text);
+        return req;
+    }
+
+    @Test
+    void spaceIsReportedMissingNeverEncoded() throws Exception {
+        PdfTextEditorCharcodeController controller = controller();
+        ResponseEntity<EncodeCharcodesResponse> resp = controller.encodeCharcodes(request(" "));
+
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        // The space must be reported missing, NOT handed back as a charcode
+        // (0x20) the frontend would reuse into the „ glyph.
+        assertThat(body.getMissing()).containsExactly(" ");
+        assertThat(body.getCharcodes()).isNullOrEmpty();
+    }
+
+    @Test
+    void realCharsEncodeWhileWhitespaceStaysAGap() throws Exception {
+        PdfTextEditorCharcodeController controller = controller();
+        // "M M" - both M's must encode to real charcodes; only the space is a gap.
+        ResponseEntity<EncodeCharcodesResponse> resp = controller.encodeCharcodes(request("M M"));
+
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getCharcodes()).as("both M glyphs encode").hasSize(2);
+        assertThat(body.getMissing()).containsExactly(" ");
+    }
+
+    @Test
+    void tabAndNewlineAreAlsoTreatedAsGaps() throws Exception {
+        PdfTextEditorCharcodeController controller = controller();
+        ResponseEntity<EncodeCharcodesResponse> resp = controller.encodeCharcodes(request("\t\n"));
+
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getMissing()).containsExactly("\t", "\n");
+        assertThat(body.getCharcodes()).isNullOrEmpty();
+    }
+
+    /**
+     * A page with two fonts that BOTH render 'A'. {@code fontName} must select which one to encode
+     * against - the cross-font fix. Without it the first font in resources order won wins and a
+     * cross-font edit got the wrong font's charcode.
+     */
+    private static String twoFontBase64() throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+            PDType1Font helvetica = new PDType1Font(Standard14Fonts.FontName.HELVETICA);
+            PDType1Font times = new PDType1Font(Standard14Fonts.FontName.TIMES_ROMAN);
+            try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
+                cs.beginText();
+                cs.setFont(helvetica, 12);
+                cs.newLineAtOffset(72, 720);
+                cs.showText("A");
+                cs.endText();
+                cs.beginText();
+                cs.setFont(times, 12);
+                cs.newLineAtOffset(72, 700);
+                cs.showText("A");
+                cs.endText();
+            }
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    private static EncodeCharcodesRequest twoFontRequest(String fontName) throws Exception {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(twoFontBase64());
+        req.setPageIndex(0);
+        req.setLocatorChar("A");
+        req.setFontName(fontName);
+        req.setText("A");
+        return req;
+    }
+
+    @Test
+    void fontNameDisambiguatesBetweenTwoFontsRenderingTheSameChar() throws Exception {
+        PdfTextEditorCharcodeController controller = controller();
+
+        // Targeting Times-Roman must encode against Times-Roman, not whichever
+        // font happens to appear first in the page's font resources.
+        EncodeCharcodesResponse times =
+                controller.encodeCharcodes(twoFontRequest("Times-Roman")).getBody();
+        assertThat(times).isNotNull();
+        assertThat(times.getError()).isNull();
+        assertThat(times.getNote()).contains("Times-Roman");
+        assertThat(times.getCharcodes()).hasSize(1);
+
+        // Targeting Helvetica must encode against Helvetica.
+        EncodeCharcodesResponse helv =
+                controller.encodeCharcodes(twoFontRequest("Helvetica")).getBody();
+        assertThat(helv).isNotNull();
+        assertThat(helv.getError()).isNull();
+        assertThat(helv.getNote()).contains("Helvetica");
+        assertThat(helv.getCharcodes()).hasSize(1);
+    }
+
+    @Test
+    void unknownFontNameReportsNoFontInsteadOfWrongFont() throws Exception {
+        PdfTextEditorCharcodeController controller = controller();
+        // A name that matches no font on the page must NOT silently encode
+        // against a different font: the frontend writes the returned charcodes
+        // into the NAMED font's text object, so a first-match fallback would
+        // bake wrong glyphs. It must report failure so the caller falls back.
+        EncodeCharcodesResponse body =
+                controller.encodeCharcodes(twoFontRequest("DoesNotExist")).getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).contains("no font");
+        assertThat(body.getCharcodes()).isNull();
+    }
+
+    @Test
+    void missingRequiredFieldsReturns400() {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64("AAAA");
+        req.setLocatorChar("M");
+        // text is null
+        ResponseEntity<EncodeCharcodesResponse> resp = controller().encodeCharcodes(req);
+        assertThat(resp.getStatusCode().value()).isEqualTo(400);
+        assertThat(resp.getBody()).isNotNull();
+        assertThat(resp.getBody().getError()).isEqualTo("missing required fields");
+    }
+
+    @Test
+    void invalidBase64Returns400() {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64("!!!notbase64!!!");
+        req.setLocatorChar("M");
+        req.setText("M");
+        ResponseEntity<EncodeCharcodesResponse> resp = controller().encodeCharcodes(req);
+        assertThat(resp.getStatusCode().value()).isEqualTo(400);
+        assertThat(resp.getBody()).isNotNull();
+        assertThat(resp.getBody().getError()).isEqualTo("pdfBase64 is not valid base64");
+    }
+
+    @Test
+    void pageIndexOutOfRangeReturns400() throws Exception {
+        EncodeCharcodesRequest req = request("M");
+        req.setPageIndex(999);
+        ResponseEntity<EncodeCharcodesResponse> resp = controller().encodeCharcodes(req);
+        assertThat(resp.getStatusCode().value()).isEqualTo(400);
+        assertThat(resp.getBody()).isNotNull();
+        assertThat(resp.getBody().getError()).isEqualTo("pageIndex out of range");
+    }
+
+    @Test
+    void nonPdfBytesReturnsGenericError() {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(Base64.getEncoder().encodeToString("not a pdf".getBytes()));
+        req.setLocatorChar("M");
+        req.setText("M");
+        // Must not throw, and must not leak the raw PDFBox parser message.
+        ResponseEntity<EncodeCharcodesResponse> resp = controller().encodeCharcodes(req);
+        assertThat(resp.getStatusCode().is4xxClientError()).isTrue();
+        assertThat(resp.getBody()).isNotNull();
+        assertThat(resp.getBody().getError()).isEqualTo("failed to load PDF");
+    }
+
+    @Test
+    void absentLocatorCharReturns200WithError() throws Exception {
+        // U+FFFF never appears in the document, so no font matches.
+        ResponseEntity<EncodeCharcodesResponse> resp =
+                controller().encodeCharcodes(requestWithLocator(""));
+        assertThat(resp.getStatusCode().value()).isEqualTo(200);
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNotNull();
+        assertThat(body.getCharcodes()).isNull();
+    }
+
+    @Test
+    void oversizePdfRejected() {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        // A base64 string long enough that length/4*3 exceeds the 100MB cap, without
+        // ever allocating the decoded bytes (the guard runs before decode).
+        char[] huge = new char[140 * 1024 * 1024];
+        java.util.Arrays.fill(huge, 'A');
+        req.setPdfBase64(new String(huge));
+        req.setLocatorChar("M");
+        req.setText("M");
+        ResponseEntity<EncodeCharcodesResponse> resp = controller().encodeCharcodes(req);
+        assertThat(resp.getStatusCode().value()).isEqualTo(413);
+        assertThat(resp.getBody()).isNotNull();
+        assertThat(resp.getBody().getError()).isEqualTo("pdf too large");
+    }
+
+    private static EncodeCharcodesRequest requestWithLocator(String locator) throws Exception {
+        EncodeCharcodesRequest req = request("M");
+        req.setLocatorChar(locator);
+        return req;
+    }
+
+    /**
+     * Build a page whose resources declare {@code filler} fonts that do NOT render 'A' (Symbol /
+     * ZapfDingbats have non-Latin encodings) plus, optionally, a trailing Helvetica that does. The
+     * Standard14 probe upper bound is 256 so each scan is cheap.
+     */
+    private static String manyFontsBase64(int filler, boolean trailingTarget) throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+            org.apache.pdfbox.pdmodel.PDResources resources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            for (int n = 0; n < filler; n++) {
+                Standard14Fonts.FontName fn =
+                        (n % 2 == 0)
+                                ? Standard14Fonts.FontName.SYMBOL
+                                : Standard14Fonts.FontName.ZAPF_DINGBATS;
+                resources.put(
+                        org.apache.pdfbox.cos.COSName.getPDFName("Ff" + n), new PDType1Font(fn));
+            }
+            if (trailingTarget) {
+                resources.put(
+                        org.apache.pdfbox.cos.COSName.getPDFName("Target"),
+                        new PDType1Font(Standard14Fonts.FontName.HELVETICA));
+            }
+            page.setResources(resources);
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    private static EncodeCharcodesRequest manyFontsRequest(String base64) {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(base64);
+        req.setPageIndex(0);
+        req.setLocatorChar("A");
+        req.setText("A");
+        return req;
+    }
+
+    @Test
+    void targetFontFoundAmongManyFonts() throws Exception {
+        // 60 non-matching fonts then the Helvetica target, all within the 64-font cap.
+        ResponseEntity<EncodeCharcodesResponse> resp =
+                controller().encodeCharcodes(manyFontsRequest(manyFontsBase64(60, true)));
+        assertThat(resp.getStatusCode().value()).isEqualTo(200);
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getCharcodes()).hasSize(1);
+    }
+
+    @Test
+    void targetBeyondFontCapReturnsGracefulNoFont() throws Exception {
+        // 64 non-matching fonts then the target at position 65 - the scan cap stops
+        // before reaching it, so we get a graceful no-font error rather than a full scan.
+        ResponseEntity<EncodeCharcodesResponse> resp =
+                controller().encodeCharcodes(manyFontsRequest(manyFontsBase64(64, true)));
+        assertThat(resp.getStatusCode().value()).isEqualTo(200);
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNotNull();
+        assertThat(body.getCharcodes()).isNull();
+    }
+
+    // Same-family sibling subsets. One document can embed several subsets of
+    // one family, each re-encoded by order of first glyph use, so a letter has
+    // a different charcode in each ("R" = 0x21 in one, 0x22 in its sibling).
+    // FPDFFont_GetBaseFontName strips the "ABCDEF+" tag, so a name-based
+    // lookup cannot tell them apart and borrows the wrong subset's codes.
+    //
+    // The doc below mirrors that with two TrueType subsets differing only by
+    // subset tag. PUA code points keep it deterministic: font.encode() cannot
+    // resolve them by glyph name, so the charcode can only come from the
+    // selected font's ToUnicode reverse map - proving WHICH font was picked.
+
+    private static final String PUA = "";
+
+    /** ToUnicode CMap mapping each supplied charcode to a BMP code point. */
+    private static byte[] toUnicodeCmap(int[][] codeToUnicode) {
+        StringBuilder sb =
+                new StringBuilder(
+                        """
+                        /CIDInit /ProcSet findresource begin
+                        12 dict begin
+                        begincmap
+                        /CIDSystemInfo << /Registry (Adobe) /Ordering (UCS) /Supplement 0 >> def
+                        /CMapName /Adobe-Identity-UCS def
+                        /CMapType 2 def
+                        1 begincodespacerange
+                        <00><FF>
+                        endcodespacerange
+                        """);
+        sb.append(codeToUnicode.length).append(" beginbfchar\n");
+        for (int[] pair : codeToUnicode) {
+            sb.append(String.format("<%02X><%04X>%n", pair[0], pair[1]));
+        }
+        sb.append(
+                """
+                endbfchar
+                endcmap
+                CMapName currentdict /CMap defineresource pop
+                end
+                end
+                """);
+        return sb.toString().getBytes(java.nio.charset.StandardCharsets.US_ASCII);
+    }
+
+    private static org.apache.pdfbox.cos.COSDictionary subsetFontDict(
+            PDDocument doc, String baseName, byte[] fontProgram, byte[] toUnicode)
+            throws Exception {
+        org.apache.pdfbox.cos.COSDictionary font = new org.apache.pdfbox.cos.COSDictionary();
+        font.setItem(org.apache.pdfbox.cos.COSName.TYPE, org.apache.pdfbox.cos.COSName.FONT);
+        font.setItem(
+                org.apache.pdfbox.cos.COSName.SUBTYPE, org.apache.pdfbox.cos.COSName.TRUE_TYPE);
+        if (baseName != null) {
+            font.setName(org.apache.pdfbox.cos.COSName.BASE_FONT, baseName);
+        }
+        font.setInt(org.apache.pdfbox.cos.COSName.FIRST_CHAR, 0x21);
+        font.setInt(org.apache.pdfbox.cos.COSName.LAST_CHAR, 0x22);
+        org.apache.pdfbox.cos.COSArray widths = new org.apache.pdfbox.cos.COSArray();
+        widths.add(org.apache.pdfbox.cos.COSInteger.get(500));
+        widths.add(org.apache.pdfbox.cos.COSInteger.get(500));
+        font.setItem(org.apache.pdfbox.cos.COSName.WIDTHS, widths);
+
+        org.apache.pdfbox.cos.COSDictionary fd = new org.apache.pdfbox.cos.COSDictionary();
+        fd.setItem(org.apache.pdfbox.cos.COSName.TYPE, org.apache.pdfbox.cos.COSName.FONT_DESC);
+        if (baseName != null) {
+            fd.setName(org.apache.pdfbox.cos.COSName.FONT_NAME, baseName);
+        }
+        fd.setInt(org.apache.pdfbox.cos.COSName.FLAGS, 4);
+        fd.setItem(
+                org.apache.pdfbox.cos.COSName.FONT_BBOX,
+                new org.apache.pdfbox.pdmodel.common.PDRectangle(0, 0, 1000, 1000).getCOSArray());
+        fd.setInt(org.apache.pdfbox.cos.COSName.ITALIC_ANGLE, 0);
+        fd.setInt(org.apache.pdfbox.cos.COSName.ASCENT, 800);
+        fd.setInt(org.apache.pdfbox.cos.COSName.DESCENT, -200);
+        fd.setInt(org.apache.pdfbox.cos.COSName.CAP_HEIGHT, 700);
+        fd.setInt(org.apache.pdfbox.cos.COSName.STEM_V, 80);
+        if (fontProgram != null) {
+            org.apache.pdfbox.pdmodel.common.PDStream ff2 =
+                    new org.apache.pdfbox.pdmodel.common.PDStream(
+                            doc, new java.io.ByteArrayInputStream(fontProgram));
+            ff2.getCOSObject().setInt(org.apache.pdfbox.cos.COSName.LENGTH1, fontProgram.length);
+            fd.setItem(org.apache.pdfbox.cos.COSName.FONT_FILE2, ff2.getCOSObject());
+        }
+        font.setItem(org.apache.pdfbox.cos.COSName.FONT_DESC, fd);
+
+        org.apache.pdfbox.pdmodel.common.PDStream tu =
+                new org.apache.pdfbox.pdmodel.common.PDStream(
+                        doc, new java.io.ByteArrayInputStream(toUnicode));
+        font.setItem(org.apache.pdfbox.cos.COSName.getPDFName("ToUnicode"), tu.getCOSObject());
+        return font;
+    }
+
+    // Distinct fake font programs - hashing distinguishes the subsets by these bytes.
+    private static final byte[] PROGRAM_A =
+            "fake-ttf-program-A".getBytes(java.nio.charset.StandardCharsets.US_ASCII);
+    private static final byte[] PROGRAM_B =
+            "fake-ttf-program-B".getBytes(java.nio.charset.StandardCharsets.US_ASCII);
+
+    /**
+     * Two sibling subsets of "FakeGaramond" whose ToUnicode maps give U+E000 DIFFERENT charcodes:
+     * 0x22 in subset A (AAAAAC+), 0x21 in subset B (AAAAAG+) - exactly the CV's shifted-code
+     * layout. {@code includeSecond=false} keeps only subset A for the unambiguous-fallback case.
+     */
+    private static String siblingSubsetsBase64(boolean includeSecond) throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+            org.apache.pdfbox.cos.COSDictionary fonts = new org.apache.pdfbox.cos.COSDictionary();
+            fonts.setItem(
+                    org.apache.pdfbox.cos.COSName.getPDFName("TTA"),
+                    subsetFontDict(
+                            doc,
+                            "AAAAAC+FakeGaramond",
+                            PROGRAM_A,
+                            toUnicodeCmap(new int[][] {{0x21, 0xE001}, {0x22, 0xE000}})));
+            if (includeSecond) {
+                fonts.setItem(
+                        org.apache.pdfbox.cos.COSName.getPDFName("TTB"),
+                        subsetFontDict(
+                                doc,
+                                "AAAAAG+FakeGaramond",
+                                PROGRAM_B,
+                                toUnicodeCmap(new int[][] {{0x21, 0xE000}, {0x22, 0xE002}})));
+            }
+            org.apache.pdfbox.pdmodel.PDResources resources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            resources.getCOSObject().setItem(org.apache.pdfbox.cos.COSName.FONT, fonts);
+            page.setResources(resources);
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    private static String sha256Hex(byte[] bytes) throws Exception {
+        byte[] digest = java.security.MessageDigest.getInstance("SHA-256").digest(bytes);
+        StringBuilder sb = new StringBuilder();
+        for (byte b : digest) sb.append(String.format("%02x", b));
+        return sb.toString();
+    }
+
+    private static EncodeCharcodesRequest siblingRequest(
+            String base64, String fontName, String fontSha256) {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(base64);
+        req.setPageIndex(0);
+        req.setLocatorChar(PUA);
+        req.setFontName(fontName);
+        req.setFontSha256(fontSha256);
+        req.setText(PUA);
+        return req;
+    }
+
+    @Test
+    void fontProgramHashSelectsTheExactSubset() throws Exception {
+        String base64 = siblingSubsetsBase64(true);
+        PdfTextEditorCharcodeController controller = controller();
+
+        // Both requests carry the SAME tag-stripped name PDFium reports ("FakeGaramond"),
+        // so only the program hash can tell the subsets apart.
+        EncodeCharcodesResponse viaA =
+                controller
+                        .encodeCharcodes(
+                                siblingRequest(base64, "FakeGaramond", sha256Hex(PROGRAM_A)))
+                        .getBody();
+        assertThat(viaA).isNotNull();
+        assertThat(viaA.getError()).isNull();
+        assertThat(viaA.getNote()).contains("AAAAAC+FakeGaramond");
+        assertThat(viaA.getCharcodes()).containsExactly(0x22L);
+
+        EncodeCharcodesResponse viaB =
+                controller
+                        .encodeCharcodes(
+                                siblingRequest(base64, "FakeGaramond", sha256Hex(PROGRAM_B)))
+                        .getBody();
+        assertThat(viaB).isNotNull();
+        assertThat(viaB.getError()).isNull();
+        assertThat(viaB.getNote()).contains("AAAAAG+FakeGaramond");
+        assertThat(viaB.getCharcodes()).containsExactly(0x21L);
+    }
+
+    @Test
+    void ambiguousStrippedNameRefusesToGuessBetweenSiblingSubsets() throws Exception {
+        // No hash, and the tag-stripped name matches BOTH subsets which both render the
+        // locator char. Guessing here is what scrambled "RUSSELL W. MANGUM III" into
+        // "US EEL W. MANGS M III" - the sibling's codes hit different glyphs. The
+        // backend must refuse so the frontend takes its safe fallback.
+        EncodeCharcodesResponse body =
+                controller()
+                        .encodeCharcodes(
+                                siblingRequest(siblingSubsetsBase64(true), "FakeGaramond", null))
+                        .getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).contains("no font");
+        assertThat(body.getCharcodes()).isNull();
+    }
+
+    @Test
+    void exactTaggedNameStillSelectsItsSubset() throws Exception {
+        // A caller that DOES know the full tagged /BaseFont name keeps working.
+        EncodeCharcodesResponse body =
+                controller()
+                        .encodeCharcodes(
+                                siblingRequest(
+                                        siblingSubsetsBase64(true), "AAAAAG+FakeGaramond", null))
+                        .getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getNote()).contains("AAAAAG+FakeGaramond");
+        assertThat(body.getCharcodes()).containsExactly(0x21L);
+    }
+
+    @Test
+    void strippedNameStillWorksWhenUnambiguous() throws Exception {
+        // With a SINGLE subset on the page, the tag-stripped name (what PDFium
+        // reports) must keep resolving - the ambiguity guard only bites when
+        // two+ siblings could answer.
+        EncodeCharcodesResponse body =
+                controller()
+                        .encodeCharcodes(
+                                siblingRequest(siblingSubsetsBase64(false), "FakeGaramond", null))
+                        .getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getNote()).contains("AAAAAC+FakeGaramond");
+        assertThat(body.getCharcodes()).containsExactly(0x22L);
+    }
+
+    @Test
+    void staleHashFallsBackToNameMatching() throws Exception {
+        // A hash matching NO font on the page (e.g. PDFium handed back a substitute
+        // font's bytes) must not brick the request: name matching still runs, and an
+        // exact tagged name resolves.
+        EncodeCharcodesResponse body =
+                controller()
+                        .encodeCharcodes(
+                                siblingRequest(
+                                        siblingSubsetsBase64(true),
+                                        "AAAAAC+FakeGaramond",
+                                        "0000000000000000000000000000000000000000000000000000000000000000"))
+                        .getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getNote()).contains("AAAAAC+FakeGaramond");
+        assertThat(body.getCharcodes()).containsExactly(0x22L);
+    }
+
+    private static final String PUA_E000 = "";
+    private static final String PUA_E002 = "";
+
+    private static final byte[] SHARED_PROGRAM =
+            "fake-ttf-program-shared".getBytes(java.nio.charset.StandardCharsets.US_ASCII);
+
+    private static String cacheIdentityPairBase64(String baseName, byte[] program)
+            throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+            org.apache.pdfbox.cos.COSDictionary fonts = new org.apache.pdfbox.cos.COSDictionary();
+            fonts.setItem(
+                    org.apache.pdfbox.cos.COSName.getPDFName("C1"),
+                    subsetFontDict(
+                            doc,
+                            baseName,
+                            program,
+                            toUnicodeCmap(new int[][] {{0x21, 0xE001}, {0x22, 0xE000}})));
+            fonts.setItem(
+                    org.apache.pdfbox.cos.COSName.getPDFName("C2"),
+                    subsetFontDict(
+                            doc,
+                            baseName,
+                            program,
+                            toUnicodeCmap(new int[][] {{0x21, 0xE002}, {0x22, 0xE003}})));
+            org.apache.pdfbox.pdmodel.PDResources resources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            resources.getCOSObject().setItem(org.apache.pdfbox.cos.COSName.FONT, fonts);
+            page.setResources(resources);
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    private static EncodeCharcodesRequest cacheIdentityRequest(
+            String base64, String locator, String fontName, String fontSha256) {
+        EncodeCharcodesRequest req = new EncodeCharcodesRequest();
+        req.setPdfBase64(base64);
+        req.setPageIndex(0);
+        req.setLocatorChar(locator);
+        req.setFontName(fontName);
+        req.setFontSha256(fontSha256);
+        req.setText(locator);
+        return req;
+    }
+
+    @Test
+    void unnamedFontsSharingOneProgramDoNotShareACachedMap() throws Exception {
+        String base64 = cacheIdentityPairBase64(null, SHARED_PROGRAM);
+        String sha = sha256Hex(SHARED_PROGRAM);
+        PdfTextEditorCharcodeController controller = controller();
+
+        EncodeCharcodesResponse first =
+                controller
+                        .encodeCharcodes(cacheIdentityRequest(base64, PUA_E000, null, sha))
+                        .getBody();
+        assertThat(first).isNotNull();
+        assertThat(first.getError()).isNull();
+        assertThat(first.getCharcodes()).containsExactly(0x22L);
+
+        EncodeCharcodesResponse second =
+                controller
+                        .encodeCharcodes(cacheIdentityRequest(base64, PUA_E002, null, sha))
+                        .getBody();
+        assertThat(second).isNotNull();
+        assertThat(second.getError()).isNull();
+        assertThat(second.getMissing()).isNullOrEmpty();
+        assertThat(second.getCharcodes())
+                .as("second font must not be served the first font's cached map")
+                .containsExactly(0x21L);
+    }
+
+    @Test
+    void fontsSharingOneNameDoNotShareACachedMap() throws Exception {
+        String base64 = cacheIdentityPairBase64("SharedName", null);
+        PdfTextEditorCharcodeController controller = controller();
+
+        EncodeCharcodesResponse first =
+                controller
+                        .encodeCharcodes(cacheIdentityRequest(base64, PUA_E000, "SharedName", null))
+                        .getBody();
+        assertThat(first).isNotNull();
+        assertThat(first.getError()).isNull();
+        assertThat(first.getCharcodes()).containsExactly(0x22L);
+
+        EncodeCharcodesResponse second =
+                controller
+                        .encodeCharcodes(cacheIdentityRequest(base64, PUA_E002, "SharedName", null))
+                        .getBody();
+        assertThat(second).isNotNull();
+        assertThat(second.getError()).isNull();
+        assertThat(second.getMissing()).isNullOrEmpty();
+        assertThat(second.getCharcodes())
+                .as("same-name fonts must not share one cached map")
+                .containsExactly(0x21L);
+    }
+
+    private static String formXObjectFontBase64() throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+
+            org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject outer =
+                    new org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject(doc);
+            outer.setBBox(new org.apache.pdfbox.pdmodel.common.PDRectangle(0, 0, 200, 200));
+            org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject inner =
+                    new org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject(doc);
+            inner.setBBox(new org.apache.pdfbox.pdmodel.common.PDRectangle(0, 0, 100, 100));
+
+            org.apache.pdfbox.pdmodel.PDResources innerResources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            innerResources.put(
+                    org.apache.pdfbox.cos.COSName.getPDFName("F1"),
+                    new PDType1Font(Standard14Fonts.FontName.HELVETICA));
+            inner.setResources(innerResources);
+
+            org.apache.pdfbox.pdmodel.PDResources outerResources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            outerResources.put(org.apache.pdfbox.cos.COSName.getPDFName("Fm1"), inner);
+            outer.setResources(outerResources);
+
+            org.apache.pdfbox.pdmodel.PDResources pageResources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            pageResources.put(org.apache.pdfbox.cos.COSName.getPDFName("Fm0"), outer);
+            page.setResources(pageResources);
+
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    private static String cyclicFormXObjectsBase64() throws Exception {
+        try (PDDocument doc = new PDDocument()) {
+            PDPage page = new PDPage();
+            doc.addPage(page);
+
+            org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject formA =
+                    new org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject(doc);
+            formA.setBBox(new org.apache.pdfbox.pdmodel.common.PDRectangle(0, 0, 100, 100));
+            org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject formB =
+                    new org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject(doc);
+            formB.setBBox(new org.apache.pdfbox.pdmodel.common.PDRectangle(0, 0, 100, 100));
+
+            org.apache.pdfbox.pdmodel.PDResources resA =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            org.apache.pdfbox.pdmodel.PDResources resB =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            resA.put(org.apache.pdfbox.cos.COSName.getPDFName("Self"), formA);
+            resA.put(org.apache.pdfbox.cos.COSName.getPDFName("Fb"), formB);
+            resB.put(org.apache.pdfbox.cos.COSName.getPDFName("Fa"), formA);
+            resB.put(
+                    org.apache.pdfbox.cos.COSName.getPDFName("F1"),
+                    new PDType1Font(Standard14Fonts.FontName.HELVETICA));
+            formA.setResources(resA);
+            formB.setResources(resB);
+
+            org.apache.pdfbox.pdmodel.PDResources pageResources =
+                    new org.apache.pdfbox.pdmodel.PDResources();
+            pageResources.put(org.apache.pdfbox.cos.COSName.getPDFName("Fm0"), formA);
+            page.setResources(pageResources);
+
+            ByteArrayOutputStream bos = new ByteArrayOutputStream();
+            doc.save(bos);
+            return Base64.getEncoder().encodeToString(bos.toByteArray());
+        }
+    }
+
+    @Test
+    void fontReachableOnlyThroughAFormXObjectIsFound() throws Exception {
+        ResponseEntity<EncodeCharcodesResponse> resp =
+                controller().encodeCharcodes(manyFontsRequest(formXObjectFontBase64()));
+        assertThat(resp.getStatusCode().value()).isEqualTo(200);
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getNote()).contains("Helvetica");
+        assertThat(body.getCharcodes()).containsExactly((long) 'A');
+    }
+
+    @Test
+    @org.junit.jupiter.api.Timeout(60)
+    void cyclicFormXObjectResourcesTerminate() throws Exception {
+        ResponseEntity<EncodeCharcodesResponse> resp =
+                controller().encodeCharcodes(manyFontsRequest(cyclicFormXObjectsBase64()));
+        assertThat(resp.getStatusCode().value()).isEqualTo(200);
+        EncodeCharcodesResponse body = resp.getBody();
+        assertThat(body).isNotNull();
+        assertThat(body.getError()).isNull();
+        assertThat(body.getCharcodes()).containsExactly((long) 'A');
+    }
+}
@@ -0,0 +1,340 @@
+package stirling.software.SPDF.controller.api;
+
+import java.io.ByteArrayInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.HashSet;
+import java.util.Set;
+import java.util.TreeSet;
+
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.contentstream.PDFStreamEngine;
+import org.apache.pdfbox.contentstream.operator.state.Concatenate;
+import org.apache.pdfbox.contentstream.operator.state.Restore;
+import org.apache.pdfbox.contentstream.operator.state.Save;
+import org.apache.pdfbox.contentstream.operator.state.SetGraphicsStateParameters;
+import org.apache.pdfbox.contentstream.operator.state.SetMatrix;
+import org.apache.pdfbox.contentstream.operator.text.BeginText;
+import org.apache.pdfbox.contentstream.operator.text.EndText;
+import org.apache.pdfbox.contentstream.operator.text.SetFontAndSize;
+import org.apache.pdfbox.contentstream.operator.text.SetTextHorizontalScaling;
+import org.apache.pdfbox.contentstream.operator.text.SetTextLeading;
+import org.apache.pdfbox.contentstream.operator.text.SetTextRenderingMode;
+import org.apache.pdfbox.contentstream.operator.text.SetTextRise;
+import org.apache.pdfbox.contentstream.operator.text.SetWordSpacing;
+import org.apache.pdfbox.contentstream.operator.text.ShowText;
+import org.apache.pdfbox.cos.COSBase;
+import org.apache.pdfbox.cos.COSDictionary;
+import org.apache.pdfbox.cos.COSName;
+import org.apache.pdfbox.cos.COSStream;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDResources;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.pdmodel.font.PDFontDescriptor;
+import org.apache.pdfbox.pdmodel.font.PDType3CharProc;
+import org.apache.pdfbox.pdmodel.font.PDType3Font;
+import org.junit.jupiter.api.Disabled;
+import org.junit.jupiter.api.Test;
+
+/**
+ * Diagnostic test: enumerate every font referenced by Sample.pdf and dump its subtype, encoding,
+ * ToUnicode, and embedded font program info. For Type3 fonts also dump CharProcs glyph names and
+ * the content stream of one glyph (the 'M' if present).
+ *
+ * <p>Not a real regression test - run with --tests SamplePdfFontDumpTest -i to see the stdout
+ * output.
+ */
+@Disabled(
+        "Diagnostic probe: dumps Sample.pdf font internals to stdout and asserts nothing. Kept for font debugging; run manually.")
+public class SamplePdfFontDumpTest {
+
+    private static final Path SAMPLE =
+            Paths.get(System.getProperty("user.dir"))
+                    .getParent()
+                    .getParent()
+                    .resolve("frontend/editor/public/samples/Sample.pdf");
+
+    @Test
+    public void dumpFonts() throws IOException {
+        byte[] pdfBytes = Files.readAllBytes(SAMPLE);
+        try (PDDocument doc = Loader.loadPDF(pdfBytes)) {
+            int numPages = doc.getNumberOfPages();
+            System.out.println("Sample.pdf has " + numPages + " pages.");
+            Set<COSDictionary> seenFontDicts = new HashSet<>();
+            for (int p = 0; p < numPages; p++) {
+                PDPage page = doc.getPage(p);
+                System.out.println("\n=== Page " + p + " ===");
+                PDResources resources = page.getResources();
+                if (resources == null) {
+                    System.out.println("  (no resources)");
+                    continue;
+                }
+                for (COSName fontName : resources.getFontNames()) {
+                    PDFont font;
+                    try {
+                        font = resources.getFont(fontName);
+                    } catch (IOException e) {
+                        System.out.println(
+                                "  Font "
+                                        + fontName.getName()
+                                        + ": failed to load - "
+                                        + e.getMessage());
+                        continue;
+                    }
+                    if (font == null) continue;
+                    COSDictionary dict = font.getCOSObject();
+                    if (!seenFontDicts.add(dict)) {
+                        System.out.println(
+                                "  Font " + fontName.getName() + " -> already seen above");
+                        continue;
+                    }
+                    dumpFont(fontName.getName(), font);
+                }
+            }
+            // Scan: for every text-show operation, record per-font (charcode, unicode) pairs.
+            System.out.println("\n=== All (font, charcode, unicode) seen on page ===");
+            for (int p = 0; p < numPages; p++) {
+                PDPage page = doc.getPage(p);
+                AllCharsScanner scanner = new AllCharsScanner();
+                scanner.processPage(page);
+                System.out.println("\nPage " + p + ":");
+                for (var entry : scanner.perFont.entrySet()) {
+                    PDFont font = entry.getKey();
+                    var seen = entry.getValue();
+                    System.out.println("  Font " + font.getName() + " " + font.getSubType() + ":");
+                    var sortedSeen = new java.util.TreeMap<Integer, String>(seen);
+                    for (var s : sortedSeen.entrySet()) {
+                        System.out.println(
+                                "    charcode 0x"
+                                        + Integer.toHexString(s.getKey())
+                                        + " ("
+                                        + s.getKey()
+                                        + ") -> '"
+                                        + s.getValue()
+                                        + "'");
+                    }
+                }
+            }
+
+            // Confirm font.encode() works for Type3 fonts.
+            System.out.println("\n=== Can we encode existing chars in F27/F28? ===");
+            PDPage page0 = doc.getPage(0);
+            PDResources r0 = page0.getResources();
+            for (String fname : new String[] {"F27", "F28"}) {
+                PDFont f = r0.getFont(COSName.getPDFName(fname));
+                if (f == null) {
+                    System.out.println("  " + fname + ": NOT FOUND on page 0");
+                    continue;
+                }
+                System.out.println("  " + fname + ": " + f.getClass().getSimpleName());
+                for (String ch : new String[] {"M", "0", "1", "+", "Z", "a"}) {
+                    try {
+                        byte[] enc = f.encode(ch);
+                        StringBuilder sb = new StringBuilder();
+                        for (byte b : enc) sb.append(String.format("%02X ", b & 0xff));
+                        System.out.println(
+                                "    encode('" + ch + "') -> [" + sb.toString().trim() + "]");
+                    } catch (Exception e) {
+                        System.out.println(
+                                "    encode('"
+                                        + ch
+                                        + "') FAILED: "
+                                        + e.getClass().getSimpleName()
+                                        + " "
+                                        + e.getMessage());
+                    }
+                }
+            }
+
+            // Dump page 0 content stream so we can see how "10M+" is composed.
+            System.out.println("\n=== Page 0 RAW content stream (first 4kb) ===");
+            try (InputStream is = doc.getPage(0).getContents()) {
+                byte[] bytes = is.readAllBytes();
+                System.out.println("Total content stream size: " + bytes.length + " bytes");
+                String asStr = new String(bytes, StandardCharsets.ISO_8859_1);
+                int idx = asStr.indexOf("F27");
+                if (idx >= 0) {
+                    int start = Math.max(0, idx - 100);
+                    int end = Math.min(asStr.length(), idx + 2500);
+                    System.out.println("--- F27 context ---");
+                    System.out.println(asStr.substring(start, end));
+                    System.out.println("---");
+                }
+                int idx2 = asStr.indexOf("F28");
+                if (idx2 >= 0) {
+                    int start = Math.max(0, idx2 - 200);
+                    int end = Math.min(asStr.length(), idx2 + 600);
+                    System.out.println("--- F28 context ---");
+                    System.out.println(asStr.substring(start, end));
+                    System.out.println("---");
+                }
+            }
+
+            // Dump a CharProc for each font's first non-zero glyph, with focus on any 'M' or "0".
+            System.out.println("\n=== Sample CharProc dumps for Type3 fonts ===");
+            Set<COSDictionary> printed = new HashSet<>();
+            for (int p = 0; p < numPages; p++) {
+                PDPage page = doc.getPage(p);
+                PDResources resources = page.getResources();
+                if (resources == null) continue;
+                for (COSName fn : resources.getFontNames()) {
+                    PDFont font = resources.getFont(fn);
+                    if (!(font instanceof PDType3Font)) continue;
+                    if (!printed.add(font.getCOSObject())) continue;
+                    PDType3Font t3 = (PDType3Font) font;
+                    // Iterate charcodes 0..255 looking for any that map to 'M' or '0' or '+'.
+                    for (int cc = 0; cc < 256; cc++) {
+                        String u = null;
+                        try {
+                            u = t3.toUnicode(cc);
+                        } catch (Exception e) {
+                            /* */
+                        }
+                        if (u == null) continue;
+                        if (u.equals("M") || u.equals("0") || u.equals("+") || u.equals("1")) {
+                            System.out.println(
+                                    "Page "
+                                            + p
+                                            + " font '"
+                                            + fn.getName()
+                                            + "' charcode "
+                                            + cc
+                                            + " maps to '"
+                                            + u
+                                            + "':");
+                            dumpType3Glyph(t3, cc);
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    private void dumpFont(String resourceName, PDFont font) {
+        COSDictionary dict = font.getCOSObject();
+        String subtype = dict.getNameAsString(COSName.SUBTYPE);
+        String baseFont = dict.getNameAsString(COSName.BASE_FONT);
+        boolean hasEncoding = dict.containsKey(COSName.ENCODING);
+        boolean hasToUnicode = dict.containsKey(COSName.TO_UNICODE);
+        PDFontDescriptor descriptor = font.getFontDescriptor();
+        boolean hasEmbedded = false;
+        String embeddedKind = "none";
+        if (descriptor != null) {
+            COSDictionary dDict = descriptor.getCOSObject();
+            if (dDict.containsKey(COSName.FONT_FILE)) {
+                hasEmbedded = true;
+                embeddedKind = "FontFile (Type1)";
+            } else if (dDict.containsKey(COSName.FONT_FILE2)) {
+                hasEmbedded = true;
+                embeddedKind = "FontFile2 (TrueType)";
+            } else if (dDict.containsKey(COSName.FONT_FILE3)) {
+                hasEmbedded = true;
+                COSBase ff3 = dDict.getDictionaryObject(COSName.FONT_FILE3);
+                if (ff3 instanceof COSStream) {
+                    String ff3Subtype = ((COSStream) ff3).getNameAsString(COSName.SUBTYPE);
+                    embeddedKind = "FontFile3 (" + ff3Subtype + ")";
+                } else {
+                    embeddedKind = "FontFile3";
+                }
+            }
+        }
+        System.out.println(
+                "  Font resource '"
+                        + resourceName
+                        + "': base='"
+                        + baseFont
+                        + "' subtype="
+                        + subtype
+                        + " hasEncoding="
+                        + hasEncoding
+                        + " hasToUnicode="
+                        + hasToUnicode
+                        + " embedded="
+                        + hasEmbedded
+                        + " ("
+                        + embeddedKind
+                        + ")");
+
+        if (font instanceof PDType3Font) {
+            PDType3Font t3 = (PDType3Font) font;
+            COSDictionary charProcs = t3.getCharProcs();
+            int count = charProcs == null ? 0 : charProcs.size();
+            System.out.println("    Type3 CharProcs count = " + count);
+            if (charProcs != null) {
+                TreeSet<String> names = new TreeSet<>();
+                for (COSName k : charProcs.keySet()) names.add(k.getName());
+                System.out.println("    glyph names: " + names);
+            }
+        }
+    }
+
+    private void dumpType3Glyph(PDType3Font font, int charcode) throws IOException {
+        String name = font.getEncoding() != null ? font.getEncoding().getName(charcode) : null;
+        System.out.println("  Type3 charcode " + charcode + " -> glyph name '" + name + "'");
+        PDType3CharProc proc = font.getCharProc(charcode);
+        if (proc == null) {
+            System.out.println("    (no CharProc for that charcode)");
+            return;
+        }
+        COSStream stream = proc.getCOSObject();
+        byte[] raw;
+        try (InputStream is = stream.createInputStream()) {
+            raw = is.readAllBytes();
+        }
+        System.out.println("    CharProc content stream (" + raw.length + " bytes):");
+        System.out.println("---");
+        System.out.println(new String(raw, StandardCharsets.ISO_8859_1));
+        System.out.println("---");
+    }
+
+    /** Records every (font, charcode -> unicode) tuple seen on a page. */
+    static final class AllCharsScanner extends PDFStreamEngine {
+        final java.util.LinkedHashMap<PDFont, java.util.Map<Integer, String>> perFont =
+                new java.util.LinkedHashMap<>();
+
+        AllCharsScanner() {
+            addOperator(new BeginText(this));
+            addOperator(new EndText(this));
+            addOperator(new SetFontAndSize(this));
+            addOperator(new SetTextHorizontalScaling(this));
+            addOperator(new SetTextLeading(this));
+            addOperator(new SetTextRenderingMode(this));
+            addOperator(new SetTextRise(this));
+            addOperator(new SetWordSpacing(this));
+            addOperator(new SetMatrix(this));
+            addOperator(new Save(this));
+            addOperator(new Restore(this));
+            addOperator(new Concatenate(this));
+            addOperator(new SetGraphicsStateParameters(this));
+            addOperator(new ShowText(this));
+        }
+
+        @Override
+        protected void showText(byte[] string) throws IOException {
+            PDFont font = getGraphicsState().getTextState().getFont();
+            if (font == null) return;
+            var seen = perFont.computeIfAbsent(font, k -> new java.util.LinkedHashMap<>());
+            ByteArrayInputStream in = new ByteArrayInputStream(string);
+            while (in.available() > 0) {
+                int code;
+                try {
+                    code = font.readCode(in);
+                } catch (IOException e) {
+                    break;
+                }
+                String u;
+                try {
+                    u = font.toUnicode(code);
+                } catch (RuntimeException e) {
+                    u = null;
+                }
+                seen.putIfAbsent(code, u);
+            }
+        }
+    }
+}
@@ -485,9 +485,9 @@ class PdfJsonFontServiceMoreTest {
    class DetectExtra {

        @Test
-        @DisplayName("detectFontFlavor recognises ttcf as cff and otf via OTTO")
+        @DisplayName("detectFontFlavor rejects ttcf collections and recognises otf via OTTO")
        void detectFlavorExtra() {
-            assertEquals("cff", service.detectFontFlavor(new byte[] {0x74, 0x74, 0x63, 0x66}));
+            assertNull(service.detectFontFlavor(new byte[] {0x74, 0x74, 0x63, 0x66}));
            List<byte[]> otfVariants = List.of(new byte[] {0x4F, 0x54, 0x54, 0x4F});
            for (byte[] otf : otfVariants) {
                assertEquals("otf", service.detectFontFlavor(otf));
@@ -57,10 +57,9 @@ class PdfJsonFontServiceTest {
    }

    @Test
-    void detectFontFlavor_cffSignature_returnsCff() {
-        // 0x74746366 = "ttcf"
-        byte[] cff = {0x74, 0x74, 0x63, 0x66};
-        assertEquals("cff", service.detectFontFlavor(cff));
+    void detectFontFlavor_ttcSignature_returnsNull() {
+        byte[] ttc = {0x74, 0x74, 0x63, 0x66};
+        assertNull(service.detectFontFlavor(ttc));
    }

    @Test
@@ -94,9 +93,9 @@ class PdfJsonFontServiceTest {
    }

    @Test
-    void detectTrueTypeFormat_cffSignature_returnsCff() {
-        byte[] cff = {0x74, 0x74, 0x63, 0x66};
-        assertEquals("cff", service.detectTrueTypeFormat(cff));
+    void detectTrueTypeFormat_ttcSignature_returnsNull() {
+        byte[] ttc = {0x74, 0x74, 0x63, 0x66};
+        assertNull(service.detectTrueTypeFormat(ttc));
    }

    @Test
@@ -59,7 +59,7 @@ public enum AuditLevel {
     */
    public static AuditLevel fromInt(int level) {
        // Ensure level is within valid bounds
-        int boundedLevel = Math.min(Math.max(level, 0), 3);
+        int boundedLevel = Math.clamp(level, 0, 3);

        for (AuditLevel auditLevel : values()) {
            if (auditLevel.level == boundedLevel) {
@@ -17,16 +17,16 @@ import org.springframework.data.redis.core.ScanOptions;
 import org.springframework.data.redis.core.StringRedisTemplate;
 import org.springframework.stereotype.Component;

-import com.fasterxml.jackson.core.JsonProcessingException;
-import com.fasterxml.jackson.core.type.TypeReference;
-import com.fasterxml.jackson.databind.ObjectMapper;
-
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;

 import stirling.software.common.cluster.JobStore;
 import stirling.software.common.cluster.JobStoreEntry;

+import tools.jackson.core.JacksonException;
+import tools.jackson.core.type.TypeReference;
+import tools.jackson.databind.ObjectMapper;
+
 /**
 * Valkey-backed {@link JobStore}. Each job is one hash; a reverse index maps fileId to jobId.
 *
@@ -44,8 +44,10 @@ public class ValkeyJobStore implements JobStore {
    private static final String FILE_INDEX_PREFIX = "stirling:file2job:";

    private static final ObjectMapper MAPPER = new ObjectMapper();
-    private static final TypeReference<List<String>> LIST_STRING = new TypeReference<>() {};
-    private static final TypeReference<Map<String, String>> MAP_STRING = new TypeReference<>() {};
+    private static final TypeReference<List<String>> LIST_STRING =
+            new TypeReference<List<String>>() {};
+    private static final TypeReference<Map<String, String>> MAP_STRING =
+            new TypeReference<Map<String, String>>() {};

    private final StringRedisTemplate template;

@@ -265,7 +267,7 @@ public class ValkeyJobStore implements JobStore {
        }
        try {
            return MAPPER.readValue(v.toString(), MAP_STRING);
-        } catch (JsonProcessingException e) {
+        } catch (JacksonException e) {
            log.warn(
                    "JobStore {} field 'resultMeta' is not valid JSON '{}' - treating as empty",
                    key,
@@ -277,7 +279,7 @@ public class ValkeyJobStore implements JobStore {
    private static String writeJson(Object value) {
        try {
            return MAPPER.writeValueAsString(value);
-        } catch (JsonProcessingException e) {
+        } catch (JacksonException e) {
            throw new IllegalStateException("Failed to JSON-serialize JobStore field", e);
        }
    }
@@ -286,7 +288,7 @@ public class ValkeyJobStore implements JobStore {
        try {
            List<String> parsed = MAPPER.readValue(json, LIST_STRING);
            return parsed == null ? new ArrayList<>() : parsed;
-        } catch (JsonProcessingException e) {
+        } catch (JacksonException e) {
            log.warn(
                    "JobStore {} field 'fileIds' is not valid JSON '{}' - treating as empty",
                    key,
@@ -35,7 +35,7 @@ public class AuditConfigurationProperties {

        // Ensure level is within valid bounds (0-3)
        int configLevel = auditConfig.getLevel();
-        this.level = Math.min(Math.max(configLevel, 0), 3);
+        this.level = Math.clamp(configLevel, 0, 3);

        // Retention days (0 means infinite)
        this.retentionDays = auditConfig.getRetentionDays();
@@ -48,7 +48,7 @@ public class UsageRestController {
            @RequestParam(value = "dataType", defaultValue = "all") String dataType,
            @RequestParam(value = "days", defaultValue = "30") Integer days) {

-        int lookbackDays = Math.max(1, Math.min(days, 365));
+        int lookbackDays = Math.clamp(days, 1, 365);

        // Get audit events filtered by type
        List<PersistentAuditEvent> events = getEventsByDataType(dataType, lookbackDays);
@@ -18,6 +18,19 @@ public enum FailureActionId {

    DISMISS(Execution.SERVER, "Dismiss"),

+    /**
+     * Open the failed operation in the client with its document, for the owner to run again
+     * themselves. Not a re-run: the settings are theirs to check first.
+     */
+    OPEN_IN_TOOL(Execution.CLIENT, "Retry"),
+
+    /**
+     * Ask the owner for the password and unlock the document in their client. Re-running is implied
+     * rather than named: an id says what a caller must supply, and a {@link
+     * FailureActionSlot#RESOLUTION} runs the failed work again once it has it.
+     */
+    DECRYPT(Execution.CLIENT, "Decrypt and retry"),
+
    /** Open the document behind the incident, in whichever client can resolve its id. */
    VIEW_FILE(Execution.CLIENT, "View file"),

@@ -0,0 +1,14 @@
+package stirling.software.proprietary.failure;
+
+/** Placement intent, not layout: the client promotes, knowing what it can actually run. */
+public enum FailureActionSlot {
+
+    /** The action that resolves the failure. At most one per kind. */
+    RESOLUTION,
+
+    /** Offered alongside the resolution, for a caller the resolution is not aimed at. */
+    SECONDARY,
+
+    /** Available but folded away: correct, rarely what anyone wants to press next. */
+    OVERFLOW
+}
@@ -1,8 +1,12 @@
 package stirling.software.proprietary.failure;

+import static stirling.software.proprietary.failure.FailureActionId.DECRYPT;
 import static stirling.software.proprietary.failure.FailureActionId.DISMISS;
+import static stirling.software.proprietary.failure.FailureActionId.OPEN_IN_TOOL;
 import static stirling.software.proprietary.failure.FailureActionId.VIEW_FILE;
 import static stirling.software.proprietary.failure.FailureActionId.VIEW_IN_PROCESSOR;
+import static stirling.software.proprietary.failure.FailureActionSlot.OVERFLOW;
+import static stirling.software.proprietary.failure.FailureActionSlot.SECONDARY;
 import static stirling.software.proprietary.failure.FailureAudience.ANYONE_WHO_SEES;
 import static stirling.software.proprietary.failure.FailureAudience.OWNER;
 import static stirling.software.proprietary.failure.FailureAudience.TEAM_REVIEWER;
@@ -21,11 +25,8 @@ import lombok.AccessLevel;
 import lombok.Getter;

 /**
- * The registry of failure kinds, described as data: a stable id, i18n keys and an English fallback
- * like {@code ExceptionUtils.ErrorCode}, plus the facets a review surface needs.
- *
- * <p>A new kind ships as a registry entry plus copy. Each offer also says who it is for, since one
- * incident is read both by whoever hit it and by whoever reviews after them.
+ * The registry of failure kinds as data: id, i18n keys, English fallback, plus the facets a review
+ * surface needs. A new kind ships as an entry plus copy; each offer says who it is for and where.
 */
@Getter
 public enum FailureKind {
@@ -36,9 +37,12 @@ public enum FailureKind {
            FailureScope.FILE,
            errorCodes("E004"),
            fallback("This document is password-protected, so the pipeline could not read it."),
-            offer(VIEW_FILE, OWNER),
-            offer(VIEW_IN_PROCESSOR, TEAM_REVIEWER),
-            offer(DISMISS, ANYONE_WHO_SEES)),
+            // The password is the fix; the owner's own document is the runner-up.
+            resolution(DECRYPT, OWNER),
+            global(VIEW_FILE, OWNER, SECONDARY),
+            global(VIEW_IN_PROCESSOR, TEAM_REVIEWER, OVERFLOW),
+            global(OPEN_IN_TOOL, OWNER, OVERFLOW),
+            global(DISMISS, ANYONE_WHO_SEES, OVERFLOW)),

    UNKNOWN(
            FailureStage.INTERNAL,
@@ -47,11 +51,11 @@ public enum FailureKind {
            FailureScope.RUN,
            noErrorCodes(),
            fallback("This run failed for a reason Stirling does not yet recognise."),
-            // Same order as every other kind: declaration order is display order, so the document
-            // leads wherever it is offered rather than moving between failures.
-            offer(VIEW_FILE, OWNER),
-            offer(VIEW_IN_PROCESSOR, TEAM_REVIEWER),
-            offer(DISMISS, ANYONE_WHO_SEES));
+            // No known fix to declare, so a plain retry leads: these are often one-offs.
+            global(OPEN_IN_TOOL, OWNER, SECONDARY),
+            global(VIEW_FILE, OWNER, SECONDARY),
+            global(VIEW_IN_PROCESSOR, TEAM_REVIEWER, OVERFLOW),
+            global(DISMISS, ANYONE_WHO_SEES, OVERFLOW));

    private static final String KEY_PREFIX = "portal.failures.kind.";
    private static final String ACTION_KEY_PREFIX = "portal.failures.action.";
@@ -98,27 +102,37 @@ public enum FailureKind {
        this.offers = List.of(offers);
    }

-    /**
-     * One ordered list rather than ids plus parallel maps of audiences and labels, which could
-     * disagree with each other.
-     *
-     * @param labelKeySuffix key under {@code portal.failures.action.}, or null for the generic
-     *     label
-     */
-    private record Offer(FailureActionId id, FailureAudience audience, String labelKeySuffix) {}
+    /** One ordered list, not parallel maps of audiences, slots and labels that could disagree. */
+    private record Offer(
+            FailureActionId id,
+            FailureAudience audience,
+            FailureActionSlot slot,
+            String labelKeySuffix) {}

-    /** Declaration order is display order. */
-    private static Offer offer(FailureActionId id, FailureAudience audience) {
-        return new Offer(id, audience, null);
+    /** The action that fixes this kind. One per kind: needing two would make it two kinds. */
+    private static Offer resolution(FailureActionId id, FailureAudience audience) {
+        return new Offer(id, audience, FailureActionSlot.RESOLUTION, null);
    }

-    /**
-     * As {@link #offer(FailureActionId, FailureAudience)}, but labelled by this kind's own wording
-     * where the shared one reads badly.
-     */
-    private static Offer offer(
+    /** As {@link #resolution(FailureActionId, FailureAudience)}, with this kind's own wording. */
+    private static Offer resolution(
            FailureActionId id, FailureAudience audience, String labelKeySuffix) {
-        return new Offer(id, audience, labelKeySuffix);
+        return new Offer(id, audience, FailureActionSlot.RESOLUTION, labelKeySuffix);
+    }
+
+    /** Not this kind's fix: an offer any kind can make, with the shared wording. */
+    private static Offer global(
+            FailureActionId id, FailureAudience audience, FailureActionSlot slot) {
+        return new Offer(id, audience, slot, null);
+    }
+
+    /** As above, with this kind's own wording where the shared one reads badly. */
+    private static Offer global(
+            FailureActionId id,
+            FailureAudience audience,
+            FailureActionSlot slot,
+            String labelKeySuffix) {
+        return new Offer(id, audience, slot, labelKeySuffix);
    }

    /**
@@ -157,21 +171,25 @@ public enum FailureKind {
        return offers.stream().map(Offer::id).toList();
    }

-    /**
-     * What this kind offers, in declaration order, each with its label resolved. What a review
-     * surface reads, so it never has to ask two separate questions about one offer.
-     */
+    /** What this kind offers, in declaration order, each with label and placement resolved. */
    public List<OfferedAction> getOfferedActions() {
        return offers.stream()
                .map(
                        offer ->
                                new OfferedAction(
-                                        offer.id(), labelKeyFor(offer.id()), offer.audience()))
+                                        offer.id(),
+                                        labelKeyFor(offer.id()),
+                                        offer.audience(),
+                                        offer.slot()))
                .toList();
    }

-    /** One action as a kind declares it: what to call it and who it is for. */
-    public record OfferedAction(FailureActionId id, String labelKey, FailureAudience audience) {}
+    /** One action as a kind declares it: what to call it, who it is for, where it wants to sit. */
+    public record OfferedAction(
+            FailureActionId id,
+            String labelKey,
+            FailureAudience audience,
+            FailureActionSlot slot) {}

    /** Whether this kind offers {@code action}. The dispatch guard: see {@code FailureActionId}. */
    public boolean declares(FailureActionId action) {
@@ -64,16 +64,17 @@ public interface FileRunEventRepository extends JpaRepository<FileRunEventEntity
    int fold(@Param("id") String id, @Param("now") Instant now, @Param("detail") String detail);

    /**
-     * Reopen a resolved incident whose failure has recurred. Guarded on the current status so only
-     * {@code RESOLVED} flips; a concurrent dismiss is never overwritten back to {@code NEW}.
+     * A recurrence reopens {@code RESOLVED} (the fix did not hold) and {@code FILE_REMOVED} (the
+     * document is back). Guarded, so a reviewer's {@code DISMISSED} is never overwritten.
     */
    @Modifying(clearAutomatically = true)
    @Transactional
    @Query(
            "update FileRunEventEntity e set"
                    + " e.status = stirling.software.proprietary.failure.FileRunEventStatus.NEW,"
-                    + " e.statusActor = null, e.statusAt = null where e.id = :id and e.status ="
-                    + " stirling.software.proprietary.failure.FileRunEventStatus.RESOLVED")
+                    + " e.statusActor = null, e.statusAt = null where e.id = :id and e.status in"
+                    + " (stirling.software.proprietary.failure.FileRunEventStatus.RESOLVED,"
+                    + " stirling.software.proprietary.failure.FileRunEventStatus.FILE_REMOVED)")
    int reopenIfResolved(@Param("id") String id);

    /**
@@ -1,5 +1,9 @@
 package stirling.software.proprietary.failure;

+import java.nio.charset.StandardCharsets;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
+import java.util.HexFormat;
 import java.util.List;
 import java.util.Map;

@@ -171,6 +175,18 @@ public class FileRunEventService {
        return action.execute(event, inputs == null ? Map.of() : inputs, currentActor());
    }

+    /** Mark an incident resolved after a client's own retry worked. Idempotent. */
+    public FileRunEvent resolve(String eventId) {
+        FileRunEvent event = requireVisible(eventId);
+        // No terminal pre-check: the store's guarded UPDATE decides, rather than racing a read.
+        return store.applyStatusOnce(
+                event.id(),
+                event.teamId(),
+                FileRunEventStatus.RESOLVED,
+                currentActor(),
+                FileRunEventStatus.open());
+    }
+
    /** "No such event" rather than a refusal, so trying does not confirm a colleague's exists. */
    private FileRunEvent requireVisible(String eventId) {
        ReadScope scope = readScope();
@@ -222,7 +238,8 @@ public class FileRunEventService {
            boolean unattended,
            boolean documentless) {
        String reason = disabledReasonFor(offer.audience(), closed, unattended, documentless);
-        return new AvailableAction(offer.id(), offer.labelKey(), reason == null, reason);
+        return new AvailableAction(
+                offer.id(), offer.labelKey(), offer.slot(), reason == null, reason);
    }

    /** Closed wins over everything, then the owner-only reasons, most specific first. */
@@ -251,11 +268,38 @@ public class FileRunEventService {
        };
    }

-    /** Login disabled has no roles, so its one operator triages everything. */
-    private boolean reviewsTeam() {
+    /** Whether the caller triages the team's incidents, not only their own. Login disabled: all. */
+    public boolean reviewsTeam() {
        return !enforced() || policyManagementAuthority.canEditPolicies();
    }

+    /**
+     * An opaque, stable discriminator for the calling viewer, for a client scoping per-browser read
+     * state. Hashed rather than the username itself: a client only needs to tell one viewer from
+     * another, and the value ends up in that browser's own storage.
+     *
+     * <p>{@code "anonymous"} with login disabled, where the one operator is every viewer.
+     */
+    public String viewerKey() {
+        String actor = currentActor();
+        return actor == null || actor.isBlank() ? "anonymous" : sha256Prefix(actor);
+    }
+
+    /** First 8 bytes of SHA-256 as hex: stable, one-way, and collision-safe enough to key on. */
+    private static String sha256Prefix(String value) {
+        try {
+            byte[] digest =
+                    MessageDigest.getInstance("SHA-256")
+                            .digest(value.getBytes(StandardCharsets.UTF_8));
+            return HexFormat.of().formatHex(digest, 0, 8);
+        } catch (NoSuchAlgorithmException e) {
+            // Every JVM ships SHA-256; a constant here would silently merge two viewers' read
+            // state, so the caller gets no key and the client falls back to showing everything.
+            log.warn("SHA-256 unavailable, so notifications cannot be scoped to a viewer", e);
+            return "";
+        }
+    }
+
    private FailureActionId parseActionId(String actionId) {
        for (FailureActionId candidate : FailureActionId.values()) {
            if (candidate.name().equals(actionId)) {
@@ -326,6 +370,11 @@ public class FileRunEventService {
        return applicationProperties.getSecurity().isEnableLogin();
    }

+    /** One action offered to one caller, availability resolved. */
    public record AvailableAction(
-            FailureActionId id, String labelKey, boolean enabled, String disabledReasonKey) {}
+            FailureActionId id,
+            String labelKey,
+            FailureActionSlot slot,
+            boolean enabled,
+            String disabledReasonKey) {}
 }
@@ -3,10 +3,7 @@ package stirling.software.proprietary.failure;
 import java.util.Arrays;
 import java.util.List;

-/**
- * Disposition of one recorded failure. {@code RESOLVED} is declared but not set yet (it becomes
- * system-set later); the rollup already defines what a repeat means for it, which is to reopen.
- */
+/** Disposition of one recorded failure. {@code RESOLVED} is system-set; a repeat reopens it. */
 public enum FileRunEventStatus {
    NEW(false),
    ACKNOWLEDGED(false),
@@ -14,9 +11,8 @@ public enum FileRunEventStatus {
    RESOLVED(true),

    /**
-     * The document this incident was about was deleted from its owner's editor, so there is nothing
-     * left to act on. Distinct from {@code DISMISSED}, which is a reviewer's decision, and from
-     * {@code RESOLVED}, which reopens on recurrence: this one cannot recur, the file is gone.
+     * The document was deleted, so there is nothing left to act on. A recurrence reopens it like
+     * {@code RESOLVED}: a fresh failure is proof the document is back.
     */
    FILE_REMOVED(true);

@@ -61,14 +61,15 @@ public record FileRunEventView(
    }

    /**
-     * {@code defaultLabel} and {@code execution} let a client render and route an action it was
-     * never built with. Declaration order is display order.
+     * {@code defaultLabel} and {@code execution} let a client render an action it was never built
+     * with; {@code slot} is placement intent. See {@link FailureActionSlot}.
     */
    public record ActionView(
            String id,
            String labelKey,
            String defaultLabel,
            FailureActionId.Execution execution,
+            FailureActionSlot slot,
            boolean enabled,
            String disabledReasonKey) {

@@ -78,6 +79,7 @@ public record FileRunEventView(
                    action.labelKey(),
                    action.id().getDefaultLabel(),
                    action.id().getExecution(),
+                    action.slot(),
                    action.enabled(),
                    action.disabledReasonKey());
        }
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.model;

+import java.io.Serial;
 import java.io.Serializable;

 import jakarta.persistence.*;
@@ -19,7 +20,7 @@ import lombok.*;
@ToString
 public class UserLicenseSettings implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    public static final Long SINGLETON_ID = 1L;

@@ -25,4 +25,7 @@ public class AiWorkflowRequest {
                    "Prior chat messages exchanged between the user and the assistant, ordered"
                            + " oldest-first. Excludes the current userMessage.")
    private List<AiConversationMessage> conversationHistory = new ArrayList<>();
+
+    @Schema(description = "IETF language tag the reply should be written in", example = "fr-FR")
+    private String locale;
 }
@@ -2,10 +2,14 @@ package stirling.software.proprietary.notification;

 import java.util.List;

+import org.springframework.http.HttpStatus;
 import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.PathVariable;
+import org.springframework.web.bind.annotation.PostMapping;
 import org.springframework.web.bind.annotation.RequestMapping;
 import org.springframework.web.bind.annotation.RequestParam;
 import org.springframework.web.bind.annotation.RestController;
+import org.springframework.web.server.ResponseStatusException;

 import io.swagger.v3.oas.annotations.Hidden;
 import io.swagger.v3.oas.annotations.Operation;
@@ -13,9 +17,11 @@ import io.swagger.v3.oas.annotations.tags.Tag;

 import lombok.RequiredArgsConstructor;

+import stirling.software.proprietary.failure.FailureActionException;
+
 /**
- * Open to any authenticated user, unlike the failure endpoints it draws on: each source scopes its
- * own rows. Read-only, because every action a notification offers runs on the client's own device.
+ * Open to any authenticated user: each source scopes its own rows. Every action runs on the
+ * client's own device, so the only write is it reporting a fix.
 */
@RestController
@RequestMapping("/api/v1/notifications")
@@ -40,9 +46,34 @@ public class NotificationController {
                            + " to mark read here yet: the client tracks what it has shown.")
    public NotificationsResponse list(@RequestParam(required = false) Integer limit) {
        int capped = Math.min(limit == null ? DEFAULT_LIMIT : Math.max(1, limit), MAX_LIMIT);
-        return new NotificationsResponse(notifications.list(capped));
+        return new NotificationsResponse(
+                notifications.list(capped),
+                notifications.callerReviewsTeam(),
+                notifications.callerViewerKey());
+    }
+
+    @PostMapping("/{notificationId}/resolved")
+    @Operation(
+            summary = "Record that a client-side retry fixed what a notification was about",
+            description =
+                    "Takes the prefixed notification id, not the producing row's id. Not an action:"
+                            + " nobody is offered a resolve button, and a recurrence brings the"
+                            + " notification back.")
+    public NotificationView resolved(@PathVariable String notificationId) {
+        try {
+            return notifications.resolve(notificationId);
+        } catch (IllegalArgumentException e) {
+            throw new ResponseStatusException(HttpStatus.BAD_REQUEST, e.getMessage(), e);
+        } catch (FailureActionException e) {
+            throw new ResponseStatusException(
+                    FailureActionException.statusOf(e.getReason()), e.getMessage(), e);
+        }
    }

    /** Wrapped so paging or a total can be added without breaking clients. */
-    public record NotificationsResponse(List<NotificationView> notifications) {}
+    public record NotificationsResponse(
+            List<NotificationView> notifications,
+            boolean viewerReviewsTeam,
+            /** Opaque; the client scopes its own read state on it. Empty means "cannot scope". */
+            String viewerKey) {}
 }
@@ -20,9 +20,47 @@ public class NotificationService {

    private final FileRunEventService fileRunEvents;

-    /** Newest first, and only open failures: one already dealt with is not news. */
+    /**
+     * Newest first, and only open failures about a document: one already dealt with is not news,
+     * and a row naming no file has nothing the bell can offer beyond saying so.
+     *
+     * <p>Filtered on the named file rather than the kind's scope, because a RUN-scoped kind still
+     * names one when the editor reported it: a failed tool run belongs here. Applied after the
+     * limit, so a page can come back short while unattributed rows exist - the review surface is
+     * where those are meant to be read, and it lists them unfiltered.
+     */
    public List<NotificationView> list(int limit) {
-        return fileRunEvents.list(null, null, limit).stream().map(this::fromFailure).toList();
+        return fileRunEvents.list(null, null, limit).stream()
+                .filter(event -> event.fileId() != null && !event.fileId().isBlank())
+                .map(this::fromFailure)
+                .toList();
+    }
+
+    /** Whether the caller sees the whole team's incidents rather than only their own. */
+    public boolean callerReviewsTeam() {
+        return fileRunEvents.reviewsTeam();
+    }
+
+    /** Opaque and stable, so a shared browser can keep one viewer's read state off another's. */
+    public String callerViewerKey() {
+        return fileRunEvents.viewerKey();
+    }
+
+    /** Takes the prefixed id, so the bell cannot reach a failure endpoint even by accident. */
+    public NotificationView resolve(String notificationId) {
+        NotificationSource.QualifiedId qualified = qualify(notificationId);
+        return switch (qualified.source()) {
+            case FAILURE -> fromFailure(fileRunEvents.resolve(qualified.rowId()));
+        };
+    }
+
+    /** The source and row id behind a notification id, refusing anything that is not one. */
+    private static NotificationSource.QualifiedId qualify(String notificationId) {
+        return NotificationSource.parse(notificationId)
+                .orElseThrow(
+                        () ->
+                                new IllegalArgumentException(
+                                        "Not a notification id: " + notificationId));
    }

    /** Prefixes the row id on the way out, so it is never sent bare. */
@@ -1,6 +1,8 @@
 package stirling.software.proprietary.notification;

+import java.util.Arrays;
 import java.util.Locale;
+import java.util.Optional;

 /**
 * Which subsystem produced a notification. Every id is prefixed with it, so a client never holds
@@ -18,4 +20,24 @@ public enum NotificationSource {
    public String qualify(String sourceRowId) {
        return prefix() + sourceRowId;
    }
+
+    /** Empty rather than throwing for an unprefixed or unknown id: both arrive from clients. */
+    public static Optional<QualifiedId> parse(String notificationId) {
+        if (notificationId == null) {
+            return Optional.empty();
+        }
+        int separator = notificationId.indexOf(SEPARATOR);
+        if (separator <= 0 || separator == notificationId.length() - 1) {
+            return Optional.empty();
+        }
+        String prefix = notificationId.substring(0, separator);
+        String rowId = notificationId.substring(separator + 1);
+        return Arrays.stream(values())
+                .filter(source -> source.name().equalsIgnoreCase(prefix))
+                .findFirst()
+                .map(source -> new QualifiedId(source, rowId));
+    }
+
+    /** A notification id split into the source that owns it and that source's own row id. */
+    public record QualifiedId(NotificationSource source, String rowId) {}
 }
@@ -336,6 +336,15 @@ public class PolicyController {
     * nothing to check.
     */
    private void requireAccessibleOutput(Policy policy) {
+        // An editor policy hands its results back to the workspace the file came from. A stored
+        // destination would send the run to a folder or bucket instead, leaving the editor's copy
+        // untouched - and the editor's import would then have nothing to collect.
+        if (policy.editor().allowed() && !policy.outputIds().isEmpty()) {
+            throw new ResponseStatusException(
+                    HttpStatus.BAD_REQUEST,
+                    "An editor policy delivers back to the editor and can't also have a"
+                            + " destination");
+        }
        for (String outputId : policy.outputIds()) {
            Source destination =
                    sourceStore
@@ -389,11 +398,14 @@ public class PolicyController {
                policy.name(),
                owner,
                policy.enabled(),
+                policy.required(),
+                policy.icon(),
                policy.inputs(),
                policy.steps(),
                policy.output(),
                policy.outputIds(),
-                teamId);
+                teamId,
+                policy.editor());
    }

    /** Output secrets never leave the server: reads return the redaction sentinel instead. */
@@ -0,0 +1,34 @@
+package stirling.software.proprietary.policy.model;
+
+/**
+ * How a policy participates in the editor: it fires in the browser as each file passes through,
+ * rather than being swept from a stored {@code Source} on a trigger.
+ *
+ * <p>An object rather than a bare flag so the moment it fires ({@code runOn}) travels with the
+ * decision, and so later editor-only settings have somewhere to live.
+ *
+ * @param allowed whether the editor may run this policy at all
+ * @param runOn which moment it fires on: {@code "upload"} or {@code "export"}
+ */
+public record EditorConfig(boolean allowed, String runOn) {
+
+    public static final String UPLOAD = "upload";
+    public static final String EXPORT = "export";
+
+    public EditorConfig {
+        runOn = EXPORT.equals(runOn) ? EXPORT : UPLOAD;
+    }
+
+    /** Not an editor policy: swept server-side, or run only on demand. */
+    public static EditorConfig disabled() {
+        return new EditorConfig(false, UPLOAD);
+    }
+
+    public static EditorConfig onUpload() {
+        return new EditorConfig(true, UPLOAD);
+    }
+
+    public static EditorConfig onExport() {
+        return new EditorConfig(true, EXPORT);
+    }
+}
@@ -1,6 +1,7 @@
 package stirling.software.proprietary.policy.model;

 import java.util.List;
+import java.util.Optional;

 /**
 * A stored automation: ordered tool steps, input bindings, and output destinations.
@@ -20,17 +21,59 @@ public record Policy(
        String name,
        String owner,
        boolean enabled,
+        boolean required,
+        String icon,
        List<PipelineInput> inputs,
        List<PipelineStep> steps,
        OutputSpec output,
        List<String> outputIds,
-        Long teamId) {
+        Long teamId,
+        EditorConfig editor) {

    public Policy {
+        icon = icon == null ? "" : icon;
        inputs = inputs == null ? List.of() : List.copyOf(inputs);
        steps = steps == null ? List.of() : steps;
        output = output == null ? OutputSpec.inline() : output;
        outputIds = outputIds == null ? List.of() : List.copyOf(outputIds);
+        editor = editor == null ? EditorConfig.disabled() : editor;
+    }
+
+    /**
+     * Without the {@code required} flag, {@code icon}, or editor participation: defaults to not
+     * org-required, no icon, and a swept/on-demand policy. Kept for the many callers and tests
+     * written before those fields; the frontend and stores that care use the full constructor.
+     */
+    public Policy(
+            String id,
+            String name,
+            String owner,
+            boolean enabled,
+            List<PipelineInput> inputs,
+            List<PipelineStep> steps,
+            OutputSpec output,
+            List<String> outputIds,
+            Long teamId) {
+        this(id, name, owner, enabled, false, "", inputs, steps, output, outputIds, teamId, null);
+    }
+
+    /**
+     * Without the {@code required} flag or {@code icon} but with explicit editor participation: the
+     * seeded Classification policy runs on the editor, so it must set {@link EditorConfig} even
+     * though it predates the org-required and icon fields.
+     */
+    public Policy(
+            String id,
+            String name,
+            String owner,
+            boolean enabled,
+            List<PipelineInput> inputs,
+            List<PipelineStep> steps,
+            OutputSpec output,
+            List<String> outputIds,
+            Long teamId,
+            EditorConfig editor) {
+        this(id, name, owner, enabled, false, "", inputs, steps, output, outputIds, teamId, editor);
    }

    /**
@@ -70,6 +113,14 @@ public record Policy(
        return inputs.stream().map(PipelineInput::sourceId).toList();
    }

+    /**
+     * The moment this policy fires in the editor ("upload" / "export"), or empty when the editor
+     * does not run it. Legacy blobs are lifted onto {@link EditorConfig} when they are read.
+     */
+    public Optional<String> editorRunOn() {
+        return editor.allowed() ? Optional.of(editor.runOn()) : Optional.empty();
+    }
+
    /** The distinct trigger types configured across this policy's inputs (manual inputs aside). */
    public List<String> triggerTypes() {
        return inputs.stream()
@@ -82,17 +133,33 @@ public record Policy(

    /** A copy with the inline output replaced (e.g. resolved for the engine, or migrated). */
    public Policy withOutput(OutputSpec resolved) {
-        return new Policy(id, name, owner, enabled, inputs, steps, resolved, outputIds, teamId);
+        return new Policy(
+                id, name, owner, enabled, required, icon, inputs, steps, resolved, outputIds,
+                teamId, editor);
    }

    /** A copy under a different owner (e.g. moving a seed off a placeholder name). */
    public Policy withOwner(String newOwner) {
-        return new Policy(id, name, newOwner, enabled, inputs, steps, output, outputIds, teamId);
+        return new Policy(
+                id, name, newOwner, enabled, required, icon, inputs, steps, output, outputIds,
+                teamId, editor);
    }

    /** A copy referencing the given saved output destinations. */
    public Policy withOutputIds(List<String> newOutputIds) {
-        return new Policy(id, name, owner, enabled, inputs, steps, output, newOutputIds, teamId);
+        return new Policy(
+                id,
+                name,
+                owner,
+                enabled,
+                required,
+                icon,
+                inputs,
+                steps,
+                output,
+                newOutputIds,
+                teamId,
+                editor);
    }

    /**
@@ -20,28 +20,23 @@ import stirling.software.proprietary.policy.source.SourceStore;
 import stirling.software.proprietary.policy.store.PolicyStore;

 /**
- * Builds the Pipelines overview: one row per policy the caller's team built on the Pipelines page,
- * with its sources resolved to live display names, its steps, and a trigger/output summary.
- * Frontend/catalogue policies (marked by a {@code categoryId} in their output options) belong to
- * the user-facing Policies page and are excluded; a folder-watch trigger is not a signal.
+ * Builds the unified Pipelines overview: one row per policy the caller's team owns, with its
+ * sources resolved to live display names, its steps, and a trigger/output summary. This lists EVERY
+ * policy - both pipelines built in the full builder and the friendly "suggested" policies - since
+ * the two surfaces were merged (a policy is a pipeline the org requires). No catalogue filter any
+ * more.
 */
@Service
@RequiredArgsConstructor
 public class PolicyOverviewService {

-    // Output-options key marking a frontend/catalogue policy (set by the Policies page and seeder).
-    private static final String CATEGORY_OPTION = "categoryId";
-
    private final PolicyStore policyStore;
    private final SourceStore sourceStore;
    private final PolicyAccessGuard policyAccessGuard;
    private final SourceAccessGuard sourceAccessGuard;

    public PoliciesOverviewResponse overview() {
-        List<Policy> policies =
-                policyAccessGuard.visibleFrom(policyStore).stream()
-                        .filter(PolicyOverviewService::isPipeline)
-                        .toList();
+        List<Policy> policies = policyAccessGuard.visibleFrom(policyStore).stream().toList();
        Map<String, String> sourceNames = sourceNames();

        List<PolicyView> views =
@@ -55,18 +50,6 @@ public class PolicyOverviewService {
        return new PoliciesOverviewResponse(buildKpis(policies), views);
    }

-    private static boolean isPipeline(Policy policy) {
-        return !isCataloguePolicy(policy);
-    }
-
-    /** A frontend/catalogue policy, marked by a {@code categoryId} in its output options. */
-    private static boolean isCataloguePolicy(Policy policy) {
-        OutputSpec output = policy.output();
-        return output != null
-                && output.options().get(CATEGORY_OPTION) instanceof String category
-                && !category.isBlank();
-    }
-
    /** Display names for every source the caller's team can see, keyed by source id. */
    private Map<String, String> sourceNames() {
        Map<String, String> names = new HashMap<>();
@@ -88,6 +71,8 @@ public class PolicyOverviewService {
                policy.id(),
                policy.name(),
                policy.enabled(),
+                policy.required(),
+                iconKey(policy),
                policy.enabled() ? "active" : "paused",
                triggerSummary(policy),
                sources,
@@ -111,13 +96,36 @@ public class PolicyOverviewService {
        return outputSummary(policy.output());
    }

+    /**
+     * The list-row icon key. The policy's first-class {@code icon} wins; otherwise a
+     * template-derived policy falls back to its {@code categoryId} (the template-identity marker
+     * the frontend maps to the category glyph). Empty when neither is set, so the frontend shows
+     * its default.
+     */
+    private static String iconKey(Policy policy) {
+        if (!policy.icon().isBlank()) {
+            return policy.icon();
+        }
+        OutputSpec output = policy.output();
+        if (output != null
+                && output.options().get("categoryId") instanceof String category
+                && !category.isBlank()) {
+            return category;
+        }
+        return "";
+    }
+
    /**
     * Summarise a policy's triggers for the overview row: "manual" when no input is triggered,
     * otherwise the distinct trigger types across its inputs (e.g. "folder-watch, schedule").
+     *
+     * <p>An editor policy has no wire input to trigger, but it is not manual either - it fires in
+     * the editor on every upload or export, so it reports that rather than reading as on-demand.
     */
    private static String triggerSummary(Policy policy) {
        List<String> types = policy.triggerTypes();
-        return types.isEmpty() ? "manual" : String.join(", ", types);
+        if (!types.isEmpty()) return String.join(", ", types);
+        return policy.editorRunOn().map(runOn -> "editor-" + runOn).orElse("manual");
    }

    private static String outputSummary(OutputSpec output) {
@@ -3,15 +3,18 @@ package stirling.software.proprietary.policy.overview;
 import java.util.List;

 /**
- * One row in the Pipelines overview: a stored policy shown for the admin portal, with its
- * referenced sources resolved to names and its pipeline summarised. The portal's "all pipelines"
- * surface lists every backend policy (the user-facing Policies page builds only a friendly subset
- * of these).
+ * One row in the unified Pipelines overview: a stored policy shown for the admin portal, with its
+ * referenced sources resolved to names and its pipeline summarised. This surface lists every
+ * backend policy - both the pipelines built in the full builder and the friendly "suggested"
+ * policies - so a {@code required} policy (one the org mandates) reads the same as any other
+ * pipeline here.
 */
 public record PolicyView(
        String id,
        String name,
        boolean enabled,
+        boolean required,
+        String icon,
        String status,
        String trigger,
        List<SourceRef> sources,
@@ -14,6 +14,7 @@ import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;

 import stirling.software.proprietary.model.TeamCreatedEvent;
+import stirling.software.proprietary.policy.model.EditorConfig;
 import stirling.software.proprietary.policy.model.OutputSpec;
 import stirling.software.proprietary.policy.model.PipelineStep;
 import stirling.software.proprietary.policy.model.Policy;
@@ -98,9 +99,8 @@ public class DefaultClassificationPolicySeeder {
    static Policy defaultPolicy(Long teamId) {
        Map<String, Object> options = new HashMap<>();
        options.put("categoryId", CATEGORY);
-        options.put("runOn", "upload");
        options.put("mode", "new_version");
-        options.put("sources", List.of("editor"));
+        options.put("sources", List.of());
        options.put("scopeTypes", List.of());
        options.put("reviewerEmail", "");
        return new Policy(
@@ -113,6 +113,9 @@ public class DefaultClassificationPolicySeeder {
                List.of(),
                List.of(new PipelineStep(CLASSIFY_ENDPOINT, Map.of())),
                new OutputSpec("inline", options),
-                teamId);
+                List.of(),
+                teamId,
+                // Classification runs in the editor on every upload.
+                EditorConfig.onUpload());
    }
 }
@@ -107,14 +107,12 @@ public class SourceOverviewService {
    }

    /**
-     * Whether a policy runs from the editor. Editor membership is carried in the policy's output
-     * metadata ({@code output.options.sources}) - a client-side list the editor writes when a
-     * policy targets it - rather than as a persisted {@code sourceId}, because the editor is
-     * virtual and has no stored source to reference.
+     * Whether a policy runs from the editor. Read from the policy's first-class {@link
+     * stirling.software.proprietary.policy.model.EditorConfig}, never inferred from a sources list
+     * (the editor is not a real source).
     */
    private static boolean runsFromEditor(Policy policy) {
-        Object sources = policy.output().options().get("sources");
-        return sources instanceof List<?> list && list.contains(EditorSource.ID);
+        return policy.editor().allowed();
    }

    /**
@@ -34,11 +34,14 @@ public class InProcessPolicyStore implements PolicyStore {
                        policy.name(),
                        policy.owner(),
                        policy.enabled(),
+                        policy.required(),
+                        policy.icon(),
                        policy.inputs(),
                        policy.steps(),
                        policy.output(),
                        policy.outputIds(),
-                        policy.teamId());
+                        policy.teamId(),
+                        policy.editor());
        policies.put(id, stored);
        // Existing policy keeps its position; a new one appends to the end of its team's queue.
        sortOrders.computeIfAbsent(id, key -> nextSortOrder(stored.teamId()));
@@ -3,6 +3,7 @@ package stirling.software.proprietary.policy.store;
 import java.util.List;
 import java.util.Objects;
 import java.util.Optional;
+import java.util.Set;
 import java.util.UUID;

 import org.springframework.stereotype.Service;
@@ -11,9 +12,12 @@ import org.springframework.transaction.annotation.Transactional;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;

+import stirling.software.proprietary.policy.model.EditorConfig;
 import stirling.software.proprietary.policy.model.Policy;
 import stirling.software.proprietary.policy.model.PolicyBinding;
+import stirling.software.proprietary.policy.source.EditorSource;

+import tools.jackson.databind.DeserializationFeature;
 import tools.jackson.databind.JsonNode;
 import tools.jackson.databind.ObjectMapper;
 import tools.jackson.databind.node.ArrayNode;
@@ -44,11 +48,14 @@ public class JpaPolicyStore implements PolicyStore {
                        policy.name(),
                        policy.owner(),
                        policy.enabled(),
+                        policy.required(),
+                        policy.icon(),
                        policy.inputs(),
                        policy.steps(),
                        policy.output(),
                        policy.outputIds(),
-                        policy.teamId());
+                        policy.teamId(),
+                        policy.editor());

        PolicyEntity entity = new PolicyEntity();
        entity.setId(id);
@@ -148,8 +155,17 @@ public class JpaPolicyStore implements PolicyStore {
    // One unreadable row must never abort a bulk read or crash startup.
    private Optional<Policy> toPolicy(PolicyEntity entity) {
        try {
-            JsonNode node = upgradeLegacyShape(objectMapper.readTree(entity.getPolicyJson()));
-            return Optional.of(objectMapper.treeToValue(node, Policy.class));
+            JsonNode node =
+                    liftEditorConfig(
+                            upgradeLegacyShape(objectMapper.readTree(entity.getPolicyJson())));
+            // A blob written by an older version won't carry fields added since (e.g. required,
+            // icon). Default absent primitives rather than rejecting the whole policy, so upgrades
+            // don't drop existing pipelines.
+            return Optional.of(
+                    objectMapper
+                            .readerFor(Policy.class)
+                            .without(DeserializationFeature.FAIL_ON_NULL_FOR_PRIMITIVES)
+                            .readValue(node));
        } catch (Exception e) {
            log.error(
                    "Skipping unreadable policy id={} name={}: stored JSON could not be parsed"
@@ -191,4 +207,61 @@ public class JpaPolicyStore implements PolicyStore {
        obj.remove("sourceIds");
        return obj;
    }
+
+    /** Categories whose editor moment defaulted to export before it was stored (see runOn.ts). */
+    private static final Set<String> EXPORT_BY_DEFAULT = Set.of("security");
+
+    /**
+     * Derive {@code editor} for a blob written before editor participation had its own field, from
+     * its {@code output.options}: allowed when {@code sources} lists {@code "editor"}, or - for a
+     * catalogue policy - when there is no {@code sources} list at all (an unnarrowed catalogue
+     * policy runs in the editor).
+     *
+     * <p>Runs on every read, deliberately outside {@link #upgradeLegacyShape}'s early return: a
+     * blob written after triggers moved onto {@code inputs} but before this field existed still
+     * needs lifting, and that early return would skip exactly those rows.
+     */
+    private JsonNode liftEditorConfig(JsonNode root) {
+        if (!(root instanceof ObjectNode obj) || obj.hasNonNull("editor")) {
+            return root;
+        }
+        JsonNode options = obj.path("output").path("options");
+        String categoryId = text(options, "categoryId");
+        JsonNode sources = options.get("sources");
+        boolean listed = sources != null && sources.isArray() && !sources.isEmpty();
+        boolean allowed;
+        if (listed) {
+            // An explicit scope list decides: only the editor's own id puts it on the editor.
+            allowed = false;
+            for (JsonNode source : sources) {
+                if (source.isValueNode() && EditorSource.ID.equals(source.asString())) {
+                    allowed = true;
+                    break;
+                }
+            }
+        } else {
+            // No list: a catalogue policy ran in the editor by default, but a builder pipeline
+            // (no category) could not reach the editor at all, so silence is not consent there.
+            allowed = !categoryId.isBlank();
+        }
+        ObjectNode editor = objectMapper.createObjectNode();
+        editor.put("allowed", allowed);
+        editor.put("runOn", legacyRunOn(options, categoryId));
+        obj.set("editor", editor);
+        return obj;
+    }
+
+    /** The stored moment, or the category default the client applied when none was stored. */
+    private static String legacyRunOn(JsonNode options, String categoryId) {
+        String stored = text(options, "runOn");
+        if (EditorConfig.EXPORT.equals(stored) || EditorConfig.UPLOAD.equals(stored)) {
+            return stored;
+        }
+        return EXPORT_BY_DEFAULT.contains(categoryId) ? EditorConfig.EXPORT : EditorConfig.UPLOAD;
+    }
+
+    private static String text(JsonNode parent, String field) {
+        JsonNode node = parent.path(field);
+        return node.isValueNode() ? node.asString() : "";
+    }
 }
@@ -70,21 +70,23 @@ public class CustomLogoutSuccessHandler extends SimpleUrlLogoutSuccessHandler {

        if (!response.isCommitted()) {
            if (authentication != null) {
-                if (authentication instanceof Saml2Authentication samlAuthentication) {
-                    // Handle SAML2 logout redirection
-                    getRedirect_saml2(request, response, samlAuthentication);
-                } else if (authentication instanceof OAuth2AuthenticationToken oAuthToken) {
-                    // Handle OAuth2 logout redirection
-                    getRedirect_oauth2(request, response, oAuthToken);
-                } else if (authentication instanceof UsernamePasswordAuthenticationToken) {
-                    // Handle Username/Password logout
-                    getRedirectStrategy().sendRedirect(request, response, LOGOUT_PATH);
-                } else {
-                    // Handle unknown authentication types
-                    log.error(
-                            "Authentication class unknown: {}",
-                            authentication.getClass().getSimpleName());
-                    getRedirectStrategy().sendRedirect(request, response, LOGOUT_PATH);
+                switch (authentication) {
+                    case Saml2Authentication samlAuthentication ->
+                            // Handle SAML2 logout redirection
+                            getRedirect_saml2(request, response, samlAuthentication);
+                    case OAuth2AuthenticationToken oAuthToken ->
+                            // Handle OAuth2 logout redirection
+                            getRedirect_oauth2(request, response, oAuthToken);
+                    case UsernamePasswordAuthenticationToken usernamePasswordAuthenticationToken ->
+                            // Handle Username/Password logout
+                            getRedirectStrategy().sendRedirect(request, response, LOGOUT_PATH);
+                    default -> {
+                        // Handle unknown authentication types
+                        log.error(
+                                "Authentication class unknown: {}",
+                                authentication.getClass().getSimpleName());
+                        getRedirectStrategy().sendRedirect(request, response, LOGOUT_PATH);
+                    }
                }
            } else {
                if (jwtService != null) {
@@ -357,12 +357,12 @@ public class SecurityConfiguration {
                                            req -> {
                                                String uri = req.getRequestURI();
                                                String contextPath = req.getContextPath();
-                                                // Check if it's a public auth endpoint or static
-                                                // resource
                                                return RequestUriUtils.isStaticResource(
                                                                contextPath, uri)
                                                        || RequestUriUtils.isPublicAuthEndpoint(
-                                                                uri, contextPath);
+                                                                uri, contextPath)
+                                                        || RequestUriUtils.isFrontendRoute(
+                                                                contextPath, uri);
                                            })
                                    .permitAll()
                                    .anyRequest()
@@ -392,40 +392,40 @@ public class SecurityConfiguration {
            // Handle OAUTH2 Logins
            if (securityProperties.isOauth2Active()) {
                http.oauth2Login(
-                        oauth2 -> {
-                            oauth2.loginPage("/login")
-                                    .authorizationEndpoint(
-                                            authorizationEndpoint -> {
-                                                if (clientRegistrationRepository != null) {
-                                                    authorizationEndpoint
-                                                            .authorizationRequestResolver(
-                                                                    new TauriAuthorizationRequestResolver(
-                                                                            clientRegistrationRepository));
-                                                }
-                                            })
-                                    .successHandler(
-                                            new CustomOAuth2AuthenticationSuccessHandler(
-                                                    loginAttemptService,
-                                                    securityProperties.getOauth2(),
-                                                    userService,
-                                                    jwtService,
-                                                    licenseSettingsService,
-                                                    applicationProperties))
-                                    .failureHandler(new CustomOAuth2AuthenticationFailureHandler())
-                                    // Add existing Authorities from the database
-                                    .userInfoEndpoint(
-                                            userInfoEndpoint ->
-                                                    userInfoEndpoint
-                                                            .oidcUserService(
-                                                                    new CustomOAuth2UserService(
-                                                                            securityProperties
-                                                                                    .getOauth2(),
-                                                                            userService,
-                                                                            loginAttemptService))
-                                                            .userAuthoritiesMapper(
-                                                                    oAuth2userAuthoritiesMapper))
-                                    .permitAll();
-                        });
+                        oauth2 ->
+                                oauth2.loginPage("/login")
+                                        .authorizationEndpoint(
+                                                authorizationEndpoint -> {
+                                                    if (clientRegistrationRepository != null) {
+                                                        authorizationEndpoint
+                                                                .authorizationRequestResolver(
+                                                                        new TauriAuthorizationRequestResolver(
+                                                                                clientRegistrationRepository));
+                                                    }
+                                                })
+                                        .successHandler(
+                                                new CustomOAuth2AuthenticationSuccessHandler(
+                                                        loginAttemptService,
+                                                        securityProperties.getOauth2(),
+                                                        userService,
+                                                        jwtService,
+                                                        licenseSettingsService,
+                                                        applicationProperties))
+                                        .failureHandler(
+                                                new CustomOAuth2AuthenticationFailureHandler())
+                                        // Add existing Authorities from the database
+                                        .userInfoEndpoint(
+                                                userInfoEndpoint ->
+                                                        userInfoEndpoint
+                                                                .oidcUserService(
+                                                                        new CustomOAuth2UserService(
+                                                                                securityProperties
+                                                                                        .getOauth2(),
+                                                                                userService,
+                                                                                loginAttemptService))
+                                                                .userAuthoritiesMapper(
+                                                                        oAuth2userAuthoritiesMapper))
+                                        .permitAll());
            }
            // Handle SAML
            if (securityProperties.isSaml2Active() && runningProOrHigher) {
@@ -703,17 +703,18 @@ public class AuthController {
    }

    private long extractEpochMillis(Object claimValue) {
-        if (claimValue == null) {
-            return -1L;
-        }
-
-        if (claimValue instanceof java.util.Date date) {
-            return date.getTime();
-        }
-
-        if (claimValue instanceof Number number) {
-            long epochSeconds = number.longValue();
-            return epochSeconds * 1000L;
+        switch (claimValue) {
+            case null -> {
+                return -1L;
+            }
+            case java.util.Date date -> {
+                return date.getTime();
+            }
+            case Number number -> {
+                long epochSeconds = number.longValue();
+                return epochSeconds * 1000L;
+            }
+            default -> {}
        }

        return -1L;
@@ -760,14 +760,14 @@ public class UserController {
            for (Object principal : principals) {
                List<SessionInformation> sessionsInformation =
                        sessionRegistry.getAllSessions(principal, false);
-                if (principal instanceof UserDetails detailsUser) {
-                    userNameP = detailsUser.getUsername();
-                } else if (principal instanceof OAuth2User oAuth2User) {
-                    userNameP = oAuth2User.getName();
-                } else if (principal instanceof CustomSaml2AuthenticatedPrincipal saml2User) {
-                    userNameP = saml2User.name();
-                } else if (principal instanceof String stringUser) {
-                    userNameP = stringUser;
+                switch (principal) {
+                    case null -> {}
+                    case UserDetails detailsUser -> userNameP = detailsUser.getUsername();
+                    case OAuth2User oAuth2User -> userNameP = oAuth2User.getName();
+                    case CustomSaml2AuthenticatedPrincipal saml2User ->
+                            userNameP = saml2User.name();
+                    case String stringUser -> userNameP = stringUser;
+                    default -> {}
                }
                if (userNameP.equalsIgnoreCase(username)) {
                    for (SessionInformation sessionInfo : sessionsInformation) {
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.security.model;

+import java.io.Serial;
 import java.io.Serializable;

 import org.springframework.security.core.GrantedAuthority;
@@ -28,7 +29,7 @@ import lombok.Setter;
@Setter
 public class Authority implements GrantedAuthority, Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.security.model;

+import java.io.Serial;
 import java.io.Serializable;
 import java.time.LocalDateTime;

@@ -18,7 +19,7 @@ import lombok.Setter;
@Setter
 public class InviteToken implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -36,57 +36,62 @@ public class CustomOAuth2AuthenticationFailureHandler
            AuthenticationException exception)
            throws IOException, ServletException {

-        if (exception instanceof BadCredentialsException) {
-            log.error("BadCredentialsException", exception);
-            getRedirectStrategy().sendRedirect(request, response, "/login?error=badCredentials");
-            return;
-        }
-        if (exception instanceof DisabledException) {
-            log.error("User is deactivated: ", exception);
-            getRedirectStrategy().sendRedirect(request, response, "/logout?userIsDisabled=true");
-            return;
-        }
-        if (exception instanceof LockedException) {
-            log.error("Account locked: ", exception);
-            getRedirectStrategy().sendRedirect(request, response, "/logout?error=locked");
-            return;
-        }
-        if (exception instanceof OAuth2AuthenticationException oAuth2Exception) {
-            OAuth2Error error = oAuth2Exception.getError();
-
-            String errorCode = error.getErrorCode();
-
-            if ("Password must not be null".equals(error.getErrorCode())) {
-                errorCode = "userAlreadyExistsWeb";
+        switch (exception) {
+            case BadCredentialsException badCredentialsException -> {
+                log.error("BadCredentialsException", exception);
+                getRedirectStrategy()
+                        .sendRedirect(request, response, "/login?error=badCredentials");
+                return;
            }
+            case DisabledException disabledException -> {
+                log.error("User is deactivated: ", exception);
+                getRedirectStrategy()
+                        .sendRedirect(request, response, "/logout?userIsDisabled=true");
+                return;
+            }
+            case LockedException lockedException -> {
+                log.error("Account locked: ", exception);
+                getRedirectStrategy().sendRedirect(request, response, "/logout?error=locked");
+                return;
+            }
+            case OAuth2AuthenticationException oAuth2Exception -> {
+                OAuth2Error error = oAuth2Exception.getError();

-            log.error(
-                    "OAuth2 Authentication error: {}",
-                    errorCode != null ? errorCode : exception.getMessage(),
-                    exception);
-            String errorValue = errorCode != null ? errorCode : "oauth2AuthenticationError";
-            clearRedirectCookie(response);
-            boolean tauriState = TauriOAuthUtils.isTauriState(request);
-            String redirectUrl;
-            if (tauriState) {
-                String basePath =
-                        TauriOAuthUtils.defaultTauriCallbackPath(request.getContextPath());
-                redirectUrl = basePath;
-                String stateParam = request.getParameter("state");
-                if (stateParam != null && !stateParam.isBlank()) {
-                    redirectUrl = appendQueryParam(redirectUrl, "state", stateParam);
-                    // Extract and pass nonce for CSRF validation
-                    String nonce = TauriOAuthUtils.extractNonceFromState(stateParam);
-                    if (nonce != null) {
-                        redirectUrl = appendQueryParam(redirectUrl, "nonce", nonce);
-                    }
+                String errorCode = error.getErrorCode();
+
+                if ("Password must not be null".equals(error.getErrorCode())) {
+                    errorCode = "userAlreadyExistsWeb";
                }
-                redirectUrl = appendQueryParam(redirectUrl, "errorOAuth", errorValue);
-            } else {
-                redirectUrl = buildFailureRedirectUrl(request, errorValue);
+
+                log.error(
+                        "OAuth2 Authentication error: {}",
+                        errorCode != null ? errorCode : exception.getMessage(),
+                        exception);
+                String errorValue = errorCode != null ? errorCode : "oauth2AuthenticationError";
+                clearRedirectCookie(response);
+                boolean tauriState = TauriOAuthUtils.isTauriState(request);
+                String redirectUrl;
+                if (tauriState) {
+                    String basePath =
+                            TauriOAuthUtils.defaultTauriCallbackPath(request.getContextPath());
+                    redirectUrl = basePath;
+                    String stateParam = request.getParameter("state");
+                    if (stateParam != null && !stateParam.isBlank()) {
+                        redirectUrl = appendQueryParam(redirectUrl, "state", stateParam);
+                        // Extract and pass nonce for CSRF validation
+                        String nonce = TauriOAuthUtils.extractNonceFromState(stateParam);
+                        if (nonce != null) {
+                            redirectUrl = appendQueryParam(redirectUrl, "nonce", nonce);
+                        }
+                    }
+                    redirectUrl = appendQueryParam(redirectUrl, "errorOAuth", errorValue);
+                } else {
+                    redirectUrl = buildFailureRedirectUrl(request, errorValue);
+                }
+                getRedirectStrategy().sendRedirect(request, response, redirectUrl);
+                return;
            }
-            getRedirectStrategy().sendRedirect(request, response, redirectUrl);
-            return;
+            default -> {}
        }
        log.error("Unhandled authentication exception", exception);
        super.onAuthenticationFailure(request, response, exception);
@@ -61,7 +61,12 @@ public class CustomSaml2ResponseAuthenticationConverter

    @Override
    public Saml2Authentication convert(ResponseToken responseToken) {
-        Assertion assertion = responseToken.getResponse().getAssertions().getFirst();
+        List<Assertion> assertions = responseToken.getResponse().getAssertions();
+        if (assertions == null || assertions.isEmpty()) {
+            log.error("SAML response contains no assertions");
+            return null;
+        }
+        Assertion assertion = assertions.getFirst();
        Map<String, List<Object>> attributes = extractAttributes(assertion);

        // Debug log with actual values
@@ -213,8 +213,11 @@ public class CustomOAuth2UserService implements OAuth2UserService<OidcUserReques
        }

        sb.append(
-                "\nWARNING: this block contains PII. Set security.oauth2.debugLogging=false once"
-                        + " troubleshooting is complete.\n");
+                """
+
+                WARNING: this block contains PII. Set security.oauth2.debugLogging=false once\
+                 troubleshooting is complete.
+                """);
        sb.append("========== [/OAUTH2 DEBUG] ==========");

        if (failure) {
@@ -132,7 +132,9 @@ public class KeyPersistenceService implements KeyPersistenceServiceInterface {
            verifyingKeyCache.put(
                    key.getKeyId(), new JwtVerificationKey(key.getKeyId(), key.getVerifyingKey()));
        }
-        activeKey = new JwtVerificationKey(keys.get(0).getKeyId(), keys.get(0).getVerifyingKey());
+        activeKey =
+                new JwtVerificationKey(
+                        keys.getFirst().getKeyId(), keys.getFirst().getVerifyingKey());
        log.info("Loaded {} JWT key(s) from DB, active key: {}", keys.size(), activeKey.getKeyId());
    }

@@ -640,14 +640,14 @@ public class UserService implements UserServiceInterface {
        for (Object principal : sessionRegistry.getAllPrincipals()) {
            for (SessionInformation sessionsInformation :
                    sessionRegistry.getAllSessions(principal, false)) {
-                if (principal instanceof UserDetails detailsUser) {
-                    usernameP = detailsUser.getUsername();
-                } else if (principal instanceof OAuth2User oAuth2User) {
-                    usernameP = oAuth2User.getName();
-                } else if (principal instanceof CustomSaml2AuthenticatedPrincipal saml2User) {
-                    usernameP = saml2User.name();
-                } else if (principal instanceof String stringUser) {
-                    usernameP = stringUser;
+                switch (principal) {
+                    case null -> {}
+                    case UserDetails detailsUser -> usernameP = detailsUser.getUsername();
+                    case OAuth2User oAuth2User -> usernameP = oAuth2User.getName();
+                    case CustomSaml2AuthenticatedPrincipal saml2User ->
+                            usernameP = saml2User.name();
+                    case String stringUser -> usernameP = stringUser;
+                    default -> {}
                }
                if (usernameP.equalsIgnoreCase(username)) {
                    sessionRegistry.expireSession(sessionsInformation.getSessionId());
@@ -47,14 +47,13 @@ public class SessionPersistentRegistry implements SessionRegistry {
        List<SessionInformation> sessionInformations = new ArrayList<>();
        String principalName = null;

-        if (principal instanceof UserDetails detailsUser) {
-            principalName = detailsUser.getUsername();
-        } else if (principal instanceof OAuth2User oAuth2User) {
-            principalName = oAuth2User.getName();
-        } else if (principal instanceof CustomSaml2AuthenticatedPrincipal saml2User) {
-            principalName = saml2User.name();
-        } else if (principal instanceof String stringUser) {
-            principalName = stringUser;
+        switch (principal) {
+            case null -> {}
+            case UserDetails detailsUser -> principalName = detailsUser.getUsername();
+            case OAuth2User oAuth2User -> principalName = oAuth2User.getName();
+            case CustomSaml2AuthenticatedPrincipal saml2User -> principalName = saml2User.name();
+            case String stringUser -> principalName = stringUser;
+            default -> {}
        }

        if (principalName != null) {
@@ -78,14 +77,13 @@ public class SessionPersistentRegistry implements SessionRegistry {
    public void registerNewSession(String sessionId, Object principal) {
        String principalName = null;

-        if (principal instanceof UserDetails detailsUser) {
-            principalName = detailsUser.getUsername();
-        } else if (principal instanceof OAuth2User oAuth2User) {
-            principalName = oAuth2User.getName();
-        } else if (principal instanceof CustomSaml2AuthenticatedPrincipal saml2User) {
-            principalName = saml2User.name();
-        } else if (principal instanceof String stringUser) {
-            principalName = stringUser;
+        switch (principal) {
+            case null -> {}
+            case UserDetails detailsUser -> principalName = detailsUser.getUsername();
+            case OAuth2User oAuth2User -> principalName = oAuth2User.getName();
+            case CustomSaml2AuthenticatedPrincipal saml2User -> principalName = saml2User.name();
+            case String stringUser -> principalName = stringUser;
+            default -> {}
        }

        if (principalName != null) {
@@ -185,6 +185,7 @@ public class AiWorkflowService {
            initialRequest.setConversationHistory(
                    new ArrayList<>(request.getConversationHistory()));
            initialRequest.setEnabledEndpoints(endpointResolver.getEnabledEndpointUrls());
+            initialRequest.setLocale(request.getLocale());
            listener.onProgress(AiWorkflowProgressEvent.of(AiWorkflowPhase.ANALYZING));

            WorkflowState state = new WorkflowState.Pending(initialRequest);
@@ -287,6 +288,7 @@ public class AiWorkflowService {
            nextRequest.setArtifacts(pdfContentExtractor.buildArtifacts(contentResults));
            nextRequest.setResumeWith(response.getResumeWith());
            nextRequest.setEnabledEndpoints(request.getEnabledEndpoints());
+            nextRequest.setLocale(request.getLocale());
            return new WorkflowState.Pending(nextRequest);
        } finally {
            for (LoadedFile lf : loadedFiles) {
@@ -338,6 +340,7 @@ public class AiWorkflowService {
        nextRequest.setFiles(request.getFiles());
        nextRequest.setConversationHistory(request.getConversationHistory());
        nextRequest.setResumeWith(response.getResumeWith());
+        nextRequest.setLocale(request.getLocale());
        return new WorkflowState.Pending(nextRequest);
    }

@@ -530,6 +533,7 @@ public class AiWorkflowService {
                                new PdfContentExtractor.ToolReportArtifact(
                                        result.reportTool(), result.report()));
                resumeRequest.setResumeWith(resumeWith);
+                resumeRequest.setLocale(previousRequest.getLocale());
                return new WorkflowState.Pending(resumeRequest);
            }

@@ -802,5 +806,6 @@ public class AiWorkflowService {
        private List<WorkflowArtifact> artifacts = new ArrayList<>();
        private String resumeWith;
        private List<String> enabledEndpoints = new ArrayList<>();
+        private String locale;
    }
 }
@@ -17,6 +17,7 @@ import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.PDResources;
 import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.text.PDFTextStripper;
 import org.springframework.stereotype.Service;

 import lombok.extern.slf4j.Slf4j;
@@ -158,6 +159,9 @@ public class FontEmbeddingService {
            if (after.getNumberOfPages() != before.getNumberOfPages()) {
                return false;
            }
+            if (lostText(before, after)) {
+                return false;
+            }
            long beforeBytes = contentBytes(before);
            long afterBytes = contentBytes(after);
            if (beforeBytes == 0) {
@@ -170,6 +174,45 @@ public class FontEmbeddingService {
        }
    }

+    /**
+     * Fraction of the original's extracted text a rewrite must still carry. The embedder re-encodes
+     * text, so a few characters either way mean nothing; a tenth of the document going missing is
+     * content loss.
+     */
+    private static final double TEXT_RETENTION_FLOOR = 0.9;
+
+    /**
+     * True when the rewrite dropped a meaningful share of the document's text.
+     *
+     * <p>Content-stream bytes cannot answer this on their own: the embedder recompresses, so they
+     * move for reasons unrelated to the page keeping its content. An 80-page document measured here
+     * came back with each page truncated to its first half - 422070 characters down to 211230 -
+     * while its content streams stayed well inside the byte ratio below.
+     *
+     * <p>Growth is not loss: flattening a widget annotation into the page legitimately adds text.
+     * Only a shortfall fails.
+     */
+    private static boolean lostText(PDDocument before, PDDocument after) {
+        String textBefore = extractText(before);
+        String textAfter = extractText(after);
+        if (textBefore == null || textAfter == null || textBefore.isBlank()) {
+            return false;
+        }
+        return textAfter.length() < textBefore.length() * TEXT_RETENTION_FLOOR;
+    }
+
+    /** Extracted text, or null when the document cannot be read - never a partial read. */
+    private static String extractText(PDDocument document) {
+        try {
+            PDFTextStripper stripper = new PDFTextStripper();
+            stripper.setSortByPosition(false);
+            return stripper.getText(document);
+        } catch (IOException | RuntimeException e) {
+            log.debug("Could not extract text while checking the rewrite: {}", e.getMessage());
+            return null;
+        }
+    }
+
    private static long contentBytes(PDDocument document) {
        long total = 0;
        for (PDPage page : document.getPages()) {
@@ -3,16 +3,16 @@ package stirling.software.proprietary.storage.converter;
 import java.util.HashMap;
 import java.util.Map;

-import com.fasterxml.jackson.core.JsonProcessingException;
-import com.fasterxml.jackson.core.type.TypeReference;
-import com.fasterxml.jackson.databind.JsonNode;
-import com.fasterxml.jackson.databind.ObjectMapper;
-
 import jakarta.persistence.AttributeConverter;
 import jakarta.persistence.Converter;

 import lombok.extern.slf4j.Slf4j;

+import tools.jackson.core.JacksonException;
+import tools.jackson.core.type.TypeReference;
+import tools.jackson.databind.JsonNode;
+import tools.jackson.databind.ObjectMapper;
+
 /**
 * JPA AttributeConverter for storing Map<String, Object> as JSON in database columns.
 *
@@ -33,7 +33,7 @@ public class JsonMapConverter implements AttributeConverter<Map<String, Object>,

        try {
            return objectMapper.writeValueAsString(attribute);
-        } catch (JsonProcessingException e) {
+        } catch (JacksonException e) {
            log.error("Failed to convert map to JSON", e);
            throw new RuntimeException("Failed to convert map to JSON", e);
        }
@@ -48,7 +48,7 @@ public class JsonMapConverter implements AttributeConverter<Map<String, Object>,
        try {
            // Try normal parsing first
            return objectMapper.readValue(dbData, new TypeReference<Map<String, Object>>() {});
-        } catch (JsonProcessingException e) {
+        } catch (JacksonException e) {
            // Fallback: try double-parsing for legacy double-encoded data
            // This handles data that was stored as JSON strings instead of JSON objects
            log.debug("Attempting double-decode fallback for legacy metadata format");
@@ -69,7 +69,7 @@ public class JsonMapConverter implements AttributeConverter<Map<String, Object>,
                    return objectMapper.readValue(
                            node.asText(), new TypeReference<Map<String, Object>>() {});
                }
-            } catch (JsonProcessingException e2) {
+            } catch (JacksonException e2) {
                log.error("Failed to parse metadata even with double-decode fallback", e2);
            }

@@ -1,5 +1,6 @@
 package stirling.software.proprietary.storage.model;

+import java.io.Serial;
 import java.io.Serializable;
 import java.time.LocalDateTime;

@@ -46,7 +47,7 @@ import stirling.software.proprietary.security.model.User;
@Setter
 public class FileShare implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.storage.model;

+import java.io.Serial;
 import java.io.Serializable;
 import java.time.LocalDateTime;

@@ -39,7 +40,7 @@ import stirling.software.proprietary.security.model.User;
@Setter
 public class FileShareAccess implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.storage.model;

+import java.io.Serial;
 import java.io.Serializable;
 import java.time.LocalDateTime;

@@ -24,7 +25,7 @@ import lombok.Setter;
@Setter
 public class StorageCleanupEntry implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.storage.model;

+import java.io.Serial;
 import java.io.Serializable;
 import java.time.LocalDateTime;
 import java.util.HashSet;
@@ -45,7 +46,7 @@ import stirling.software.proprietary.workflow.model.WorkflowSession;
@Setter
 public class StoredFile implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
@@ -1,5 +1,6 @@
 package stirling.software.proprietary.storage.model;

+import java.io.Serial;
 import java.io.Serializable;

 import jakarta.persistence.Column;
@@ -19,7 +20,7 @@ import lombok.Setter;
@Setter
 public class StoredFileBlob implements Serializable {

-    private static final long serialVersionUID = 1L;
+    @Serial private static final long serialVersionUID = 1L;

    @Id
    @Column(name = "storage_key", nullable = false, length = 128)
@@ -7,6 +7,7 @@ import org.slf4j.MDC;
 import org.springframework.core.Ordered;
 import org.springframework.core.annotation.Order;
 import org.springframework.security.core.Authentication;
+import org.springframework.security.core.GrantedAuthority;
 import org.springframework.security.core.context.SecurityContextHolder;
 import org.springframework.stereotype.Component;
 import org.springframework.web.filter.OncePerRequestFilter;
@@ -64,7 +65,7 @@ public class AuditWebFilter extends OncePerRequestFilter {
            if (auth != null && auth.getAuthorities() != null) {
                String roles =
                        auth.getAuthorities().stream()
-                                .map(a -> a.getAuthority())
+                                .map(GrantedAuthority::getAuthority)
                                .reduce((a, b) -> a + "," + b)
                                .orElse("");
                MDC.put("userRoles", roles);
--- a/Show More
+++ b/Show More