WIP

* Non blocking improvements

* Test fix

* Background thread re-queue

* Nonblockimg improvements, run tasks in background, add warning about CPU cores

* Misc fixes

.github/nginx-reverse-proxy-test.conf

@@ -0,0 +1,33 @@
+server {
+    listen 80;
+    server_name localhost;
+
+    # Test basic reverse proxy to changedetection.io
+    location / {
+        proxy_pass http://changedet-app:5000;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+
+        # WebSocket support
+        proxy_http_version 1.1;
+        proxy_set_header Upgrade $http_upgrade;
+        proxy_set_header Connection "upgrade";
+    }
+
+    # Test subpath deployment with X-Forwarded-Prefix
+    location /changedet-sub/ {
+        proxy_pass http://changedet-app:5000/;
+        proxy_set_header X-Forwarded-Prefix /changedet-sub;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+
+        # WebSocket support
+        proxy_http_version 1.1;
+        proxy_set_header Upgrade $http_upgrade;
+        proxy_set_header Connection "upgrade";
+    }
+}

.github/workflows/containers.yml

@@ -15,7 +15,6 @@ on:
   push:
     branches:
       - master
-      - dev
 
 jobs:
   metadata:
@@ -67,43 +66,54 @@ jobs:
           echo ${{ github.ref }} > changedetectionio/tag.txt
 
       - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
+        uses: docker/setup-qemu-action@v4
         with:
           image: tonistiigi/binfmt:latest
           platforms: all
 
       - name: Login to GitHub Container Registry
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           registry: ghcr.io
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Login to Docker Hub Container Registry
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           username: ${{ secrets.DOCKER_HUB_USERNAME }}
           password: ${{ secrets.DOCKER_HUB_ACCESS_TOKEN }}
 
       - name: Set up Docker Buildx
         id: buildx
-        uses: docker/setup-buildx-action@v3
+        uses: docker/setup-buildx-action@v4
         with:
           install: true
           version: latest
           driver-opts: image=moby/buildkit:master
 
-      # dev branch -> :dev container tag
+      # master branch -> :dev container tag
+      - name: Docker meta :dev
+        if: ${{ github.ref == 'refs/heads/master' && github.event_name != 'release' }}
+        uses: docker/metadata-action@v6
+        id: meta_dev
+        with:
+          images: |
+            ${{ secrets.DOCKER_HUB_USERNAME }}/changedetection.io
+            ghcr.io/${{ github.repository }}
+          tags: |
+            type=raw,value=dev
+
       - name: Build and push :dev
         id: docker_build
-        if: ${{ github.ref == 'refs/heads/dev' }}
-        uses: docker/build-push-action@v6
+        if: ${{ github.ref == 'refs/heads/master' && github.event_name != 'release' }}
+        uses: docker/build-push-action@v7
         with:
           context: ./
           file: ./Dockerfile
           push: true
-          tags: |
-            ${{ secrets.DOCKER_HUB_USERNAME }}/changedetection.io:dev,ghcr.io/${{ github.repository }}:dev
+          tags: ${{ steps.meta_dev.outputs.tags }}
+          labels: ${{ steps.meta_dev.outputs.labels }}
           platforms: linux/amd64,linux/arm64,linux/arm/v7,linux/arm/v8
           cache-from: type=gha
           cache-to: type=gha,mode=max
@@ -121,7 +131,7 @@ jobs:
           
       - name: Docker meta :tag
         if: github.event_name == 'release' && startsWith(github.event.release.tag_name, '0.')
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         id: meta
         with:
             images: |
@@ -136,12 +146,13 @@ jobs:
       - name: Build and push :tag
         id: docker_build_tag_release
         if: github.event_name == 'release' && startsWith(github.event.release.tag_name, '0.')
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         with:
           context: ./
           file: ./Dockerfile
           push: true
           tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
           platforms: linux/amd64,linux/arm64,linux/arm/v7,linux/arm/v8
           cache-from: type=gha
           cache-to: type=gha,mode=max

.github/workflows/pypi-release.yml

@@ -21,7 +21,7 @@ jobs:
     - name: Build a binary wheel and a source tarball
       run: python3 -m build
     - name: Store the distribution packages
-      uses: actions/upload-artifact@v6
+      uses: actions/upload-artifact@v7
       with:
         name: python-package-distributions
         path: dist/
@@ -34,7 +34,7 @@ jobs:
     - build
     steps:
     - name: Download all the dists
-      uses: actions/download-artifact@v7
+      uses: actions/download-artifact@v8
       with:
         name: python-package-distributions
         path: dist/
@@ -61,8 +61,8 @@ jobs:
         
         # --- API test ---
         # This also means that the docs/api-spec.yml was shipped and could be read
-        test -f /tmp/url-watches.json
-        API_KEY=$(jq -r '.. | .api_access_token? // empty' /tmp/url-watches.json)
+        test -f /tmp/changedetection.json
+        API_KEY=$(jq -r '.. | .api_access_token? // empty' /tmp/changedetection.json)
         echo Test API KEY is $API_KEY
         curl -X POST "http://127.0.0.1:10000/api/v1/watch" \
           -H "x-api-key: ${API_KEY}" \
@@ -93,7 +93,7 @@ jobs:
 
     steps:
     - name: Download all the dists
-      uses: actions/download-artifact@v7
+      uses: actions/download-artifact@v8
       with:
         name: python-package-distributions
         path: dist/

.github/workflows/test-container-build.yml

@@ -60,14 +60,14 @@ jobs:
 
         # Just test that the build works, some libraries won't compile on ARM/rPi etc
         - name: Set up QEMU
-          uses: docker/setup-qemu-action@v3
+          uses: docker/setup-qemu-action@v4
           with:
             image: tonistiigi/binfmt:latest
             platforms: all
 
         - name: Set up Docker Buildx
           id: buildx
-          uses: docker/setup-buildx-action@v3
+          uses: docker/setup-buildx-action@v4
           with:
             install: true
             version: latest
@@ -75,7 +75,7 @@ jobs:
 
         - name: Test that the docker containers can build (${{ matrix.platform }} - ${{ matrix.dockerfile }})
           id: docker_build
-          uses: docker/build-push-action@v6
+          uses: docker/build-push-action@v7
           # https://github.com/docker/build-push-action#customizing
           with:
             context: ./

.github/workflows/test-only.yml

@@ -52,4 +52,13 @@ jobs:
     uses: ./.github/workflows/test-stack-reusable-workflow.yml
     with:
       python-version: '3.13'
-      skip-pypuppeteer: true
+      skip-pypuppeteer: true
+
+
+  test-application-3-14:
+    #if: github.event_name == 'push' && github.ref == 'refs/heads/master'
+    needs: lint-code
+    uses: ./.github/workflows/test-stack-reusable-workflow.yml
+    with:
+      python-version: '3.14'
+      skip-pypuppeteer: false

.github/workflows/test-stack-reusable-workflow.yml

@@ -37,10 +37,29 @@ jobs:
             ${{ runner.os }}-pip-py${{ env.PYTHON_VERSION }}-
             ${{ runner.os }}-pip-
 
+      - name: Get current date for cache key
+        id: date
+        run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v4
+
       - name: Build changedetection.io container for testing under Python ${{ env.PYTHON_VERSION }}
+        uses: docker/build-push-action@v7
+        with:
+          context: ./
+          file: ./Dockerfile
+          build-args: |
+            PYTHON_VERSION=${{ env.PYTHON_VERSION }}
+            LOGGER_LEVEL=TRACE
+          tags: test-changedetectionio
+          load: true
+          cache-from: type=gha,scope=build-${{ github.ref_name }}-py${{ env.PYTHON_VERSION }}-${{ hashFiles('requirements.txt', 'Dockerfile') }}-${{ steps.date.outputs.date }}
+          cache-to: type=gha,mode=max,scope=build-${{ github.ref_name }}-py${{ env.PYTHON_VERSION }}-${{ hashFiles('requirements.txt', 'Dockerfile') }}-${{ steps.date.outputs.date }}
+
+      - name: Verify build
         run: |
-          echo "---- Building for Python ${{ env.PYTHON_VERSION }} -----"
-          docker build --build-arg PYTHON_VERSION=${{ env.PYTHON_VERSION }} --build-arg LOGGER_LEVEL=TRACE -t test-changedetectionio .
+          echo "---- Built for Python ${{ env.PYTHON_VERSION }} -----"
           docker run test-changedetectionio bash -c 'pip list'
 
       - name: We should be Python ${{ env.PYTHON_VERSION }} ...
@@ -52,7 +71,7 @@ jobs:
           docker save test-changedetectionio -o /tmp/test-changedetectionio.tar
 
       - name: Upload Docker image artifact
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp/test-changedetectionio.tar
@@ -69,7 +88,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -84,6 +103,7 @@ jobs:
           docker run test-changedetectionio bash -c 'python3 -m unittest changedetectionio.tests.unit.test_watch_model'
           docker run test-changedetectionio bash -c 'python3 -m unittest changedetectionio.tests.unit.test_jinja2_security'
           docker run test-changedetectionio bash -c 'python3 -m unittest changedetectionio.tests.unit.test_semver'
+          docker run test-changedetectionio bash -c 'python3 -m unittest changedetectionio.tests.unit.test_html_to_text'         
 
   # Basic pytest tests with ancillary services
   basic-tests:
@@ -96,7 +116,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -110,6 +130,32 @@ jobs:
           docker network inspect changedet-network >/dev/null 2>&1 || docker network create changedet-network
           docker run --name test-cdio-basic-tests --network changedet-network test-changedetectionio bash -c 'cd changedetectionio && ./run_basic_tests.sh'
 
+      - name: Test CLI options
+        run: |
+          docker network inspect changedet-network >/dev/null 2>&1 || docker network create changedet-network
+          docker run --name test-cdio-cli-opts --network changedet-network test-changedetectionio bash -c 'changedetectionio/test_cli_opts.sh' &> cli-opts-output.txt
+          echo "=== CLI Options Test Output ==="
+          cat cli-opts-output.txt
+
+      - name: CLI Memory Test
+        run: |
+          echo "=== Checking CLI batch mode memory usage ==="
+          # Extract RSS memory value from output
+          RSS_MB=$(grep -oP "Memory consumption before worker shutdown: RSS=\K[\d.]+" cli-opts-output.txt | head -1 || echo "0")
+          echo "RSS Memory: ${RSS_MB} MB"
+
+          # Check if RSS is less than 100MB
+          if [ -n "$RSS_MB" ]; then
+            if (( $(echo "$RSS_MB < 100" | bc -l) )); then
+              echo "✓ Memory usage is acceptable: ${RSS_MB} MB < 100 MB"
+            else
+              echo "✗ Memory usage too high: ${RSS_MB} MB >= 100 MB"
+              exit 1
+            fi
+          else
+            echo "⚠ Could not extract memory usage, skipping check"
+          fi
+
       - name: Extract memory report and logs
         if: always()
         uses: ./.github/actions/extract-memory-report
@@ -119,11 +165,18 @@ jobs:
 
       - name: Store test artifacts
         if: always()
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: test-cdio-basic-tests-output-py${{ env.PYTHON_VERSION }}
           path: output-logs
 
+      - name: Store CLI test output
+        if: always()
+        uses: actions/upload-artifact@v7
+        with:
+          name: test-cdio-cli-opts-output-py${{ env.PYTHON_VERSION }}
+          path: cli-opts-output.txt
+
   # Playwright tests
   playwright-tests:
     runs-on: ubuntu-latest
@@ -135,7 +188,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -177,7 +230,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -217,7 +270,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -253,7 +306,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -271,6 +324,175 @@ jobs:
         run: |
           docker run --rm --network changedet-network test-changedetectionio bash -c 'cd changedetectionio;pytest tests/smtp/test_notification_smtp.py'
 
+  nginx-reverse-proxy:
+    runs-on: ubuntu-latest
+    needs: build
+    timeout-minutes: 10
+    env:
+      PYTHON_VERSION: ${{ inputs.python-version }}
+    steps:
+      - uses: actions/checkout@v6
+
+      - name: Download Docker image artifact
+        uses: actions/download-artifact@v8
+        with:
+          name: test-changedetectionio-${{ env.PYTHON_VERSION }}
+          path: /tmp
+
+      - name: Load Docker image
+        run: |
+          docker load -i /tmp/test-changedetectionio.tar
+
+      - name: Spin up services
+        run: |
+          docker network create changedet-network
+
+          # Start changedetection.io container with X-Forwarded headers support
+          docker run --name changedet-app --hostname changedet-app --network changedet-network \
+            -e USE_X_SETTINGS=true \
+            -d test-changedetectionio
+          sleep 3
+
+      - name: Start nginx reverse proxy
+        run: |
+          # Start nginx with our test configuration
+          docker run --name nginx-proxy --network changedet-network -d -p 8080:80 --rm \
+            -v ${{ github.workspace }}/.github/nginx-reverse-proxy-test.conf:/etc/nginx/conf.d/default.conf:ro \
+            nginx:alpine
+          sleep 2
+
+      - name: Test reverse proxy - root path
+        run: |
+          echo "=== Testing nginx reverse proxy at root path ==="
+          curl --retry-connrefused --retry 6 -s http://localhost:8080/ > /tmp/nginx-test-root.html
+
+          # Check for changedetection.io UI elements
+          if grep -q "checkbox-uuid" /tmp/nginx-test-root.html; then
+            echo "✓ Found checkbox-uuid in response"
+          else
+            echo "ERROR: checkbox-uuid not found in response"
+            cat /tmp/nginx-test-root.html
+            exit 1
+          fi
+
+          # Check for watchlist content
+          if grep -q -i "watch" /tmp/nginx-test-root.html; then
+            echo "✓ Found watch/watchlist content in response"
+          else
+            echo "ERROR: watchlist content not found"
+            cat /tmp/nginx-test-root.html
+            exit 1
+          fi
+
+          echo "✓ Root path reverse proxy working correctly"
+
+      - name: Test reverse proxy - subpath with X-Forwarded-Prefix
+        run: |
+          echo "=== Testing nginx reverse proxy at subpath /changedet-sub/ ==="
+          curl --retry-connrefused --retry 6 -s http://localhost:8080/changedet-sub/ > /tmp/nginx-test-subpath.html
+
+          # Check for changedetection.io UI elements
+          if grep -q "checkbox-uuid" /tmp/nginx-test-subpath.html; then
+            echo "✓ Found checkbox-uuid in subpath response"
+          else
+            echo "ERROR: checkbox-uuid not found in subpath response"
+            cat /tmp/nginx-test-subpath.html
+            exit 1
+          fi
+
+          echo "✓ Subpath reverse proxy working correctly"
+
+      - name: Test API through reverse proxy subpath
+        run: |
+          echo "=== Testing API endpoints through nginx subpath /changedet-sub/ ==="
+
+          # Extract API key from the changedetection.io datastore
+          API_KEY=$(docker exec changedet-app cat /datastore/changedetection.json | grep -o '"api_access_token": *"[^"]*"' | cut -d'"' -f4)
+
+          if [ -z "$API_KEY" ]; then
+            echo "ERROR: Could not extract API key from datastore"
+            docker exec changedet-app cat /datastore/changedetection.json
+            exit 1
+          fi
+
+          echo "✓ Extracted API key: ${API_KEY:0:8}..."
+
+          # Create a watch via API through nginx proxy subpath
+          echo "Creating watch via POST to /changedet-sub/api/v1/watch"
+          RESPONSE=$(curl -s -w "\n%{http_code}" -X POST "http://localhost:8080/changedet-sub/api/v1/watch" \
+            -H "x-api-key: ${API_KEY}" \
+            -H "Content-Type: application/json" \
+            -d '{
+              "url": "https://example.com/test-nginx-proxy",
+              "tag": "nginx-test"
+            }')
+
+          HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
+          BODY=$(echo "$RESPONSE" | head -n-1)
+
+          if [ "$HTTP_CODE" != "201" ]; then
+            echo "ERROR: Expected HTTP 201, got $HTTP_CODE"
+            echo "Response: $BODY"
+            exit 1
+          fi
+
+          echo "✓ Watch created successfully (HTTP 201)"
+
+          # Extract the watch UUID from response
+          WATCH_UUID=$(echo "$BODY" | grep -o '"uuid": *"[^"]*"' | cut -d'"' -f4)
+          echo "✓ Watch UUID: $WATCH_UUID"
+
+          # Update the watch via PUT through nginx proxy subpath
+          echo "Updating watch via PUT to /changedet-sub/api/v1/watch/${WATCH_UUID}"
+          RESPONSE=$(curl -s -w "\n%{http_code}" -X PUT "http://localhost:8080/changedet-sub/api/v1/watch/${WATCH_UUID}" \
+            -H "x-api-key: ${API_KEY}" \
+            -H "Content-Type: application/json" \
+            -d '{
+              "paused": true
+            }')
+
+          HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
+          BODY=$(echo "$RESPONSE" | head -n-1)
+
+          if [ "$HTTP_CODE" != "200" ]; then
+            echo "ERROR: Expected HTTP 200, got $HTTP_CODE"
+            echo "Response: $BODY"
+            exit 1
+          fi
+
+          if echo "$BODY" | grep -q 'OK'; then
+            echo "✓ Watch updated successfully (HTTP 200, response: OK)"
+          else
+            echo "ERROR: Expected response 'OK', got: $BODY"
+            echo "Response: $BODY"
+            exit 1
+          fi
+
+          # Verify the watch is paused via GET
+          echo "Verifying watch is paused via GET"
+          RESPONSE=$(curl -s "http://localhost:8080/changedet-sub/api/v1/watch/${WATCH_UUID}" \
+            -H "x-api-key: ${API_KEY}")
+
+          if echo "$RESPONSE" | grep -q '"paused": *true'; then
+            echo "✓ Watch is paused as expected"
+          else
+            echo "ERROR: Watch paused state not confirmed"
+            echo "Response: $RESPONSE"
+            exit 1
+          fi
+
+          echo "✓ API tests through nginx subpath completed successfully"
+
+      - name: Cleanup nginx test
+        if: always()
+        run: |
+          docker logs nginx-proxy || true
+          docker logs changedet-app || true
+          docker stop nginx-proxy changedet-app || true
+          docker rm nginx-proxy changedet-app || true
+
+
+
   # Proxy tests
   proxy-tests:
     runs-on: ubuntu-latest
@@ -282,7 +504,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -322,7 +544,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -342,6 +564,29 @@ jobs:
           cd changedetectionio
           ./run_custom_browser_url_tests.sh
 
+  processor-plugin-tests:
+    runs-on: ubuntu-latest
+    needs: build
+    timeout-minutes: 20
+    env:
+      PYTHON_VERSION: ${{ inputs.python-version }}
+    steps:
+      - uses: actions/checkout@v6
+
+      - name: Download Docker image artifact
+        uses: actions/download-artifact@v8
+        with:
+          name: test-changedetectionio-${{ env.PYTHON_VERSION }}
+          path: /tmp
+
+      - name: Load Docker image
+        run: |
+          docker load -i /tmp/test-changedetectionio.tar
+
+      - name: Basic processor plugin registration and checks
+        run: |
+          docker run  -e EXTRA_PACKAGES=changedetection.io-osint-processor test-changedetectionio bash -c 'cd changedetectionio;pytest -vvv -s  tests/plugins/test_processor.py::test_check_plugin_processor'
+
   # Container startup tests
   container-tests:
     runs-on: ubuntu-latest
@@ -353,7 +598,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -398,7 +643,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Download Docker image artifact
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: test-changedetectionio-${{ env.PYTHON_VERSION }}
           path: /tmp
@@ -440,3 +685,154 @@ jobs:
             exit 1
           fi
           docker rm sig-test
+
+  # Upgrade path test
+  upgrade-path-test:
+    runs-on: ubuntu-latest
+    needs: build
+    timeout-minutes: 25
+    env:
+      PYTHON_VERSION: ${{ inputs.python-version }}
+    steps:
+      - uses: actions/checkout@v6
+        with:
+          fetch-depth: 0  # Fetch all history and tags for upgrade testing
+
+      - name: Set up Python ${{ env.PYTHON_VERSION }}
+        uses: actions/setup-python@v6
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Check upgrade works without error
+        run: |
+          echo "=== Testing upgrade path from 0.49.1 to ${{ github.ref_name }} (${{ github.sha }}) ==="
+          sudo apt-get update && sudo apt-get install -y --no-install-recommends \
+              g++ \
+              gcc \
+              libc-dev \
+              libffi-dev \
+              libjpeg-dev \
+              libssl-dev \
+              libxslt-dev \
+              make \
+              patch \
+              pkg-config \
+              zlib1g-dev
+          
+          # Checkout old version and create datastore
+          git checkout 0.49.1
+          python3 -m venv .venv
+          source .venv/bin/activate
+          pip install -r requirements.txt
+          pip install 'pyOpenSSL>=23.2.0'
+
+          echo "=== Running version 0.49.1 to create datastore ==="
+          ALLOW_IANA_RESTRICTED_ADDRESSES=true python3 ./changedetection.py -C -d /tmp/data &
+          APP_PID=$!
+
+          # Wait for app to be ready
+          echo "Waiting for 0.49.1 to be ready..."
+          sleep 6
+
+          # Extract API key from datastore (0.49.1 uses url-watches.json)
+          API_KEY=$(jq -r '.settings.application.api_access_token // empty' /tmp/data/url-watches.json)
+          echo "API Key: ${API_KEY:0:8}..."
+
+          # Create a watch with tag "github-group-test" via API
+          echo "Creating test watch with tag via API..."
+          curl -X POST "http://127.0.0.1:5000/api/v1/watch" \
+            -H "x-api-key: ${API_KEY}" \
+            -H "Content-Type: application/json" \
+            --show-error --fail \
+            --retry 6 --retry-delay 1 --retry-connrefused \
+            -d '{
+              "url": "https://example.com/upgrade-test",
+              "tag": "github-group-test"
+            }'
+
+          echo "✓ Created watch with tag 'github-group-test'"
+
+          # Create a specific test URL watch
+          echo "Creating test URL watch via API..."
+          curl -X POST "http://127.0.0.1:5000/api/v1/watch" \
+            -H "x-api-key: ${API_KEY}" \
+            -H "Content-Type: application/json" \
+            --show-error --fail \
+            -d '{
+              "url": "http://localhost/test.txt"
+            }'
+
+          echo "✓ Created watch for 'http://localhost/test.txt' in version 0.49.1"
+
+          # Stop the old version gracefully
+          kill $APP_PID
+          wait $APP_PID || true
+          echo "✓ Version 0.49.1 stopped"
+
+          # Upgrade to current version (use commit SHA since we're in detached HEAD)
+          echo "Upgrading to commit ${{ github.sha }}"
+          git checkout ${{ github.sha }}
+          pip install -r requirements.txt
+
+          echo "=== Running current version (commit ${{ github.sha }}) with old datastore (testing mode) ==="
+          ALLOW_IANA_RESTRICTED_ADDRESSES=true TESTING_SHUTDOWN_AFTER_DATASTORE_LOAD=1 python3 ./changedetection.py -d /tmp/data > /tmp/upgrade-test.log 2>&1
+
+          echo "=== Upgrade test output ==="
+          cat /tmp/upgrade-test.log
+          echo "✓ Datastore upgraded successfully"
+
+          # Now start the current version normally to verify the tag survived
+          echo "=== Starting current version to verify tag exists after upgrade ==="
+          ALLOW_IANA_RESTRICTED_ADDRESSES=true timeout 20 python3 ./changedetection.py -d /tmp/data > /tmp/ui-test.log 2>&1 &
+          APP_PID=$!
+
+          # Wait for app to be ready and fetch UI
+          echo "Waiting for current version to be ready..."
+          sleep 5
+          curl --retry 6 --retry-delay 1 --retry-connrefused --silent http://127.0.0.1:5000 > /tmp/ui-output.html
+
+          # Verify tag exists in UI
+          if grep -q "github-group-test" /tmp/ui-output.html; then
+            echo "✓ Tag 'github-group-test' found in UI after upgrade"
+          else
+            echo "ERROR: Tag 'github-group-test' not found in UI after upgrade"
+            echo "=== UI Output ==="
+            cat /tmp/ui-output.html
+            echo "=== App Log ==="
+            cat /tmp/ui-test.log
+            kill $APP_PID || true
+            exit 1
+          fi
+
+          # Verify test URL exists in UI
+          if grep -q "http://localhost/test.txt" /tmp/ui-output.html; then
+            echo "✓ Watch URL 'http://localhost/test.txt' found in UI after upgrade"
+          else
+            echo "ERROR: Watch URL 'http://localhost/test.txt' not found in UI after upgrade"
+            echo "=== UI Output ==="
+            cat /tmp/ui-output.html
+            echo "=== App Log ==="
+            cat /tmp/ui-test.log
+            kill $APP_PID || true
+            exit 1
+          fi
+
+          # Cleanup
+          kill $APP_PID || true
+          wait $APP_PID || true
+
+          echo ""
+          echo "✓✓✓ Upgrade test passed: 0.49.1 → ${{ github.ref_name }} ✓✓✓"
+          echo "    - Commit: ${{ github.sha }}"
+          echo "    - Datastore migrated successfully"
+          echo "    - Tag 'github-group-test' survived upgrade"
+          echo "    - Watch URL 'http://localhost/test.txt' survived upgrade"
+
+          echo "✓ Upgrade test passed: 0.49.1 → ${{ github.ref_name }}"
+
+      - name: Upload upgrade test logs
+        if: always()
+        uses: actions/upload-artifact@v7
+        with:
+          name: upgrade-test-logs-py${{ env.PYTHON_VERSION }}
+          path: /tmp/upgrade-test.log

.gitignore

@@ -29,3 +29,4 @@ test-datastore/
 
 # Memory consumption log
 test-memory.log
+tests/logs/

Dockerfile

@@ -78,6 +78,12 @@ RUN --mount=type=cache,id=pip,sharing=locked,target=/tmp/pip-cache \
 # Final image stage
 FROM python:${PYTHON_VERSION}-slim-bookworm
 LABEL org.opencontainers.image.source="https://github.com/dgtlmoon/changedetection.io"
+LABEL org.opencontainers.image.url="https://changedetection.io"
+LABEL org.opencontainers.image.documentation="https://changedetection.io/tutorials"
+LABEL org.opencontainers.image.title="changedetection.io"
+LABEL org.opencontainers.image.description="Self-hosted web page change monitoring and notification service"
+LABEL org.opencontainers.image.licenses="Apache-2.0"
+LABEL org.opencontainers.image.vendor="changedetection.io"
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
     libxslt1.1 \
@@ -132,6 +138,15 @@ ENV LOGGER_LEVEL="$LOGGER_LEVEL"
 ENV LC_ALL=en_US.UTF-8
 
 WORKDIR /app
+
+# Copy and set up entrypoint script for installing extra packages
+COPY docker-entrypoint.sh /docker-entrypoint.sh
+RUN chmod +x /docker-entrypoint.sh
+
+# Set entrypoint to handle EXTRA_PACKAGES env var
+ENTRYPOINT ["/docker-entrypoint.sh"]
+
+# Default command (can be overridden in docker-compose.yml)
 CMD ["python", "./changedetection.py", "-d", "/datastore"]
 
 

MANIFEST.in

@@ -9,12 +9,15 @@ recursive-include changedetectionio/notification *
 recursive-include changedetectionio/processors *
 recursive-include changedetectionio/realtime *
 recursive-include changedetectionio/static *
+recursive-include changedetectionio/store *
 recursive-include changedetectionio/templates *
 recursive-include changedetectionio/tests *
+recursive-include changedetectionio/translations *
 recursive-include changedetectionio/widgets *
 prune changedetectionio/static/package-lock.json
 prune changedetectionio/static/styles/node_modules
 prune changedetectionio/static/styles/package-lock.json
+include changedetectionio/favicon_utils.py
 include changedetection.py
 include requirements.txt
 include README-pip.md

babel.cfg

@@ -0,0 +1,5 @@
+[python: **.py]
+keywords = _:1,_l:1,gettext:1
+
+[jinja2: **/templates/**.html]
+encoding = utf-8

changedetectionio/__init__.py

@@ -2,23 +2,24 @@
 
 # Read more https://github.com/dgtlmoon/changedetection.io/wiki
 # Semver means never use .01, or 00. Should be .1.
-__version__ = '0.52.1'
+__version__ = '0.54.4'
 
 from changedetectionio.strtobool import strtobool
 from json.decoder import JSONDecodeError
+
+from loguru import logger
+import getopt
 import logging
 import os
-import getopt
 import platform
 import signal
-
-import sys
+import threading
+import time
 
 # Eventlet completely removed - using threading mode for SocketIO
 # This provides better Python 3.12+ compatibility and eliminates eventlet/asyncio conflicts
-from changedetectionio import store
-from changedetectionio.flask_app import changedetection_app
-from loguru import logger
+# Note: store and changedetection_app are imported inside main() to avoid
+# initialization before argument parsing (allows --help to work without loading everything)
 
 # ==============================================================================
 # Multiprocessing Configuration - CRITICAL for Thread Safety
@@ -41,9 +42,10 @@ from loguru import logger
 #
 # IMPLEMENTATION:
 #   1. Explicit contexts everywhere (primary protection):
-#      - Watch.py: ctx = multiprocessing.get_context('spawn')
 #      - playwright.py: ctx = multiprocessing.get_context('spawn')
 #      - puppeteer.py: ctx = multiprocessing.get_context('spawn')
+#      - isolated_opencv.py: ctx = multiprocessing.get_context('spawn')
+#      - isolated_libvips.py: ctx = multiprocessing.get_context('spawn')
 #
 #   2. Global default (defense-in-depth, below):
 #      - Safety net if future code forgets explicit context
@@ -59,8 +61,22 @@ from loguru import logger
 # ==============================================================================
 
 import multiprocessing
+import os
 import sys
 
+# Limit glibc malloc arena count to prevent RSS growth from concurrent requests.
+# Default: glibc creates up to 8×CPU_cores arenas. Each concurrent thread/connection
+# can trigger a new arena, and freed memory stays mapped in those arenas as RSS forever.
+# With MALLOC_ARENA_MAX=2, at most 2 arenas are used; freed pages return to the OS faster.
+# Must be set before worker threads start; env var is read lazily by glibc on first arena creation.
+if 'MALLOC_ARENA_MAX' not in os.environ:
+    os.environ['MALLOC_ARENA_MAX'] = '2'
+    try:
+        import ctypes as _ctypes
+        _ctypes.CDLL('libc.so.6').mallopt(-8, 2)  # M_ARENA_MAX = -8
+    except Exception:
+        pass
+
 # Set spawn as global default (safety net - all our code uses explicit contexts anyway)
 # Skip in tests to avoid breaking pytest-flask's LiveServer fixture (uses unpicklable local functions)
 if 'pytest' not in sys.modules:
@@ -82,15 +98,26 @@ def get_version():
 def sigshutdown_handler(_signo, _stack_frame):
     name = signal.Signals(_signo).name
     logger.critical(f'Shutdown: Got Signal - {name} ({_signo}), Fast shutdown initiated')
-    
+
     # Set exit flag immediately to stop all loops
     app.config.exit.set()
     datastore.stop_thread = True
-    
+
+    # Log memory consumption before shutting down workers (cross-platform)
+    try:
+        import psutil
+        process = psutil.Process()
+        mem_info = process.memory_info()
+        rss_mb = mem_info.rss / 1024 / 1024
+        vms_mb = mem_info.vms / 1024 / 1024
+        logger.info(f"Memory consumption before worker shutdown: RSS={rss_mb:,.2f} MB, VMS={vms_mb:,.2f} MB")
+    except Exception as e:
+        logger.warning(f"Could not retrieve memory stats: {str(e)}")
+
     # Shutdown workers and queues immediately
     try:
-        from changedetectionio import worker_handler
-        worker_handler.shutdown_workers()
+        from changedetectionio import worker_pool
+        worker_pool.shutdown_workers()
     except Exception as e:
         logger.error(f"Error shutting down workers: {str(e)}")
     
@@ -99,9 +126,9 @@ def sigshutdown_handler(_signo, _stack_frame):
         from changedetectionio.flask_app import update_q, notification_q
         update_q.close()
         notification_q.close()
-        logger.debug("Janus queues closed successfully")
+        logger.debug("Queues closed successfully")
     except Exception as e:
-        logger.critical(f"CRITICAL: Failed to close janus queues: {e}")
+        logger.critical(f"CRITICAL: Failed to close queues: {e}")
     
     # Shutdown socketio server fast
     from changedetectionio.flask_app import socketio_server
@@ -111,31 +138,80 @@ def sigshutdown_handler(_signo, _stack_frame):
         except Exception as e:
             logger.error(f"Error shutting down Socket.IO server: {str(e)}")
     
-    # Save data quickly
-    try:
-        datastore.sync_to_json()
-        logger.success('Fast sync to disk complete.')
-    except Exception as e:
-        logger.error(f"Error syncing to disk: {str(e)}")
-    
+    # With immediate persistence, all data is already saved
+    logger.success('All data already persisted (immediate commits enabled).')
+
     sys.exit()
 
+def print_help():
+    """Print help text for command line options"""
+    print('Usage: changedetection.py [options]')
+    print('')
+    print('Standard options:')
+    print('  -s                SSL enable')
+    print('  -h HOST           Listen host (default: 0.0.0.0)')
+    print('  -p PORT           Listen port (default: 5000)')
+    print('  -d PATH           Datastore path')
+    print('  -l LEVEL          Log level (TRACE, DEBUG, INFO, SUCCESS, WARNING, ERROR, CRITICAL)')
+    print('  -c                Cleanup unused snapshots')
+    print('  -C                Create datastore directory if it doesn\'t exist')
+    print('  -P true/false     Set all watches paused (true) or active (false)')
+    print('')
+    print('Add URLs on startup:')
+    print('  -u URL            Add URL to watch (can be used multiple times)')
+    print('  -u0 \'JSON\'        Set options for first -u URL (e.g. \'{"processor":"text_json_diff"}\')')
+    print('  -u1 \'JSON\'        Set options for second -u URL (0-indexed)')
+    print('  -u2 \'JSON\'        Set options for third -u URL, etc.')
+    print('                    Available options: processor, fetch_backend, headers, method, etc.')
+    print('                    See model/Watch.py for all available options')
+    print('')
+    print('Recheck on startup:')
+    print('  -r all            Queue all watches for recheck on startup')
+    print('  -r UUID,...       Queue specific watches (comma-separated UUIDs)')
+    print('  -r all N          Queue all watches, wait for completion, repeat N times')
+    print('  -r UUID,... N     Queue specific watches, wait for completion, repeat N times')
+    print('')
+    print('Batch mode:')
+    print('  -b                Run in batch mode (process queue then exit)')
+    print('                    Useful for CI/CD, cron jobs, or one-time checks')
+    print('                    NOTE: Batch mode checks if Flask is running and aborts if port is in use')
+    print('                    Use -p PORT to specify a different port if needed')
+    print('')
+
 def main():
     global datastore
     global app
 
+    # Early help/version check before any initialization
+    if '--help' in sys.argv or '-help' in sys.argv:
+        print_help()
+        sys.exit(0)
+
+    if '--version' in sys.argv or '-v' in sys.argv:
+        print(f'changedetection.io {__version__}')
+        sys.exit(0)
+
+    # Import heavy modules after help/version checks to keep startup fast for those flags
+    from changedetectionio import store
+    from changedetectionio.flask_app import changedetection_app
+
     datastore_path = None
-    do_cleanup = False
-    # Optional URL to watch since start
-    default_url = None
     # Set a default logger level
     logger_level = 'DEBUG'
     include_default_watches = True
+    all_paused = None  # None means don't change, True/False to set
 
     host = os.environ.get("LISTEN_HOST", "0.0.0.0").strip()
     port = int(os.environ.get('PORT', 5000))
     ssl_mode = False
 
+    # Lists for multiple URLs and their options
+    urls_to_add = []
+    url_options = {}  # Key: index (0-based), Value: dict of options
+    recheck_watches = None  # None, 'all', or list of UUIDs
+    recheck_repeat_count = 1  # Number of times to repeat recheck cycle
+    batch_mode = False  # Run once then exit when queue is empty
+
     # On Windows, create and use a default path.
     if os.name == 'nt':
         datastore_path = os.path.expandvars(r'%APPDATA%\changedetection.io')
@@ -144,10 +220,68 @@ def main():
         # Must be absolute so that send_from_directory doesnt try to make it relative to backend/
         datastore_path = os.path.join(os.getcwd(), "../datastore")
 
+    # Pre-process arguments to extract -u, -u<N>, and -r options before getopt
+    # This allows unlimited -u0, -u1, -u2, ... options without predefining them
+    cleaned_argv = ['changedetection.py']  # Start with program name
+    i = 1
+    while i < len(sys.argv):
+        arg = sys.argv[i]
+
+        # Handle -u (add URL)
+        if arg == '-u' and i + 1 < len(sys.argv):
+            urls_to_add.append(sys.argv[i + 1])
+            i += 2
+            continue
+
+        # Handle -u<N> (set options for URL at index N)
+        if arg.startswith('-u') and len(arg) > 2 and arg[2:].isdigit():
+            idx = int(arg[2:])
+            if i + 1 < len(sys.argv):
+                try:
+                    import json
+                    url_options[idx] = json.loads(sys.argv[i + 1])
+                except json.JSONDecodeError as e:
+                    print(f'Error: Invalid JSON for {arg}: {sys.argv[i + 1]}')
+                    print(f'JSON decode error: {e}')
+                    sys.exit(2)
+                i += 2
+                continue
+
+        # Handle -r (recheck watches)
+        if arg == '-r' and i + 1 < len(sys.argv):
+            recheck_arg = sys.argv[i + 1]
+            if recheck_arg.lower() == 'all':
+                recheck_watches = 'all'
+            else:
+                # Parse comma-separated list of UUIDs
+                recheck_watches = [uuid.strip() for uuid in recheck_arg.split(',') if uuid.strip()]
+
+            # Check for optional repeat count as third argument
+            if i + 2 < len(sys.argv) and sys.argv[i + 2].isdigit():
+                recheck_repeat_count = int(sys.argv[i + 2])
+                if recheck_repeat_count < 1:
+                    print(f'Error: Repeat count must be at least 1, got {recheck_repeat_count}')
+                    sys.exit(2)
+                i += 3
+            else:
+                i += 2
+            continue
+
+        # Handle -b (batch mode - run once and exit)
+        if arg == '-b':
+            batch_mode = True
+            i += 1
+            continue
+
+        # Keep other arguments for getopt
+        cleaned_argv.append(arg)
+        i += 1
+
     try:
-        opts, args = getopt.getopt(sys.argv[1:], "6Ccsd:h:p:l:u:", "port")
-    except getopt.GetoptError:
-        print('backend.py -s SSL enable -h [host] -p [port] -d [datastore path] -u [default URL to watch] -l [debug level - TRACE, DEBUG(default), INFO, SUCCESS, WARNING, ERROR, CRITICAL]')
+        opts, args = getopt.getopt(cleaned_argv[1:], "6Csd:h:p:l:P:", "port")
+    except getopt.GetoptError as e:
+        print_help()
+        print(f'Error: {e}')
         sys.exit(2)
 
     create_datastore_dir = False
@@ -172,14 +306,6 @@ def main():
         if opt == '-d':
             datastore_path = arg
 
-        if opt == '-u':
-            default_url = arg
-            include_default_watches = False
-
-        # Cleanup (remove text files that arent in the index)
-        if opt == '-c':
-            do_cleanup = True
-
         # Create the datadir if it doesnt exist
         if opt == '-C':
             create_datastore_dir = True
@@ -187,6 +313,18 @@ def main():
         if opt == '-l':
             logger_level = int(arg) if arg.isdigit() else arg.upper()
 
+        if opt == '-P':
+            try:
+                all_paused = bool(strtobool(arg))
+            except ValueError:
+                print(f'Error: Invalid value for -P option: {arg}')
+                print('Expected: true, false, yes, no, 1, or 0')
+                sys.exit(2)
+
+    # If URLs are provided, don't include default watches
+    if urls_to_add:
+        include_default_watches = False
+
 
     logger.success(f"changedetection.io version {get_version()} starting.")
     # Launch using SocketIO run method for proper integration (if enabled)
@@ -223,11 +361,16 @@ def main():
     logging.getLogger('pyppeteer.connection.Connection').setLevel(logging.WARNING)
 
     # isnt there some @thingy to attach to each route to tell it, that this route needs a datastore
-    app_config = {'datastore_path': datastore_path}
+    app_config = {
+        'datastore_path': datastore_path,
+        'batch_mode': batch_mode,
+        'recheck_watches': recheck_watches,
+        'recheck_repeat_count': recheck_repeat_count
+    }
 
     if not os.path.isdir(app_config['datastore_path']):
         if create_datastore_dir:
-            os.mkdir(app_config['datastore_path'])
+            os.makedirs(app_config['datastore_path'], exist_ok=True)
         else:
             logger.critical(
                 f"ERROR: Directory path for the datastore '{app_config['datastore_path']}'"
@@ -242,17 +385,219 @@ def main():
         # Dont' start if the JSON DB looks corrupt
         logger.critical(f"ERROR: JSON DB or Proxy List JSON at '{app_config['datastore_path']}' appears to be corrupt, aborting.")
         logger.critical(str(e))
-        return
+        sys.exit(1)
+
+    # Testing mode: Exit cleanly after datastore initialization (for CI/CD upgrade tests)
+    if os.environ.get('TESTING_SHUTDOWN_AFTER_DATASTORE_LOAD'):
+        logger.success(f"TESTING MODE: Datastore loaded successfully from {app_config['datastore_path']}")
+        logger.success(f"TESTING MODE: Schema version: {datastore.data['settings']['application'].get('schema_version', 'unknown')}")
+        logger.success(f"TESTING MODE: Loaded {len(datastore.data['watching'])} watches")
+        logger.success("TESTING MODE: Exiting cleanly (TESTING_SHUTDOWN_AFTER_DATASTORE_LOAD is set)")
+        sys.exit(0)
+
+    # Apply all_paused setting if specified via CLI
+    if all_paused is not None:
+        datastore.data['settings']['application']['all_paused'] = all_paused
+        logger.info(f"Setting all watches paused: {all_paused}")
 
     # Inject datastore into plugins that need access to settings
     from changedetectionio.pluggy_interface import inject_datastore_into_plugins
     inject_datastore_into_plugins(datastore)
 
-    if default_url:
-        datastore.add_watch(url = default_url)
+    # Step 1: Add URLs with their options (if provided via -u flags)
+    added_watch_uuids = []
+    if urls_to_add:
+        logger.info(f"Adding {len(urls_to_add)} URL(s) from command line")
+        for idx, url in enumerate(urls_to_add):
+            extras = url_options.get(idx, {})
+            if extras:
+                logger.debug(f"Adding watch {idx}: {url} with options: {extras}")
+            else:
+                logger.debug(f"Adding watch {idx}: {url}")
+
+            new_uuid = datastore.add_watch(url=url, extras=extras)
+            if new_uuid:
+                added_watch_uuids.append(new_uuid)
+                logger.success(f"Added watch: {url} (UUID: {new_uuid})")
+            else:
+                logger.error(f"Failed to add watch: {url}")
 
     app = changedetection_app(app_config, datastore)
 
+    # Step 2: Queue newly added watches (if -u was provided in batch mode)
+    # This must happen AFTER app initialization so update_q is available
+    if batch_mode and added_watch_uuids:
+        from changedetectionio.flask_app import update_q
+        from changedetectionio import queuedWatchMetaData, worker_pool
+
+        logger.info(f"Batch mode: Queuing {len(added_watch_uuids)} newly added watches")
+        for watch_uuid in added_watch_uuids:
+            try:
+                worker_pool.queue_item_async_safe(
+                    update_q,
+                    queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid})
+                )
+                logger.debug(f"Queued newly added watch: {watch_uuid}")
+            except Exception as e:
+                logger.error(f"Failed to queue watch {watch_uuid}: {e}")
+
+    # Step 3: Queue watches for recheck (if -r was provided)
+    # This must happen AFTER app initialization so update_q is available
+    if recheck_watches is not None:
+        from changedetectionio.flask_app import update_q
+        from changedetectionio import queuedWatchMetaData, worker_pool
+
+        watches_to_queue = []
+        if recheck_watches == 'all':
+            # Queue all watches, excluding those already queued in batch mode
+            all_watches = list(datastore.data['watching'].keys())
+            if batch_mode and added_watch_uuids:
+                # Exclude newly added watches that were already queued in batch mode
+                watches_to_queue = [uuid for uuid in all_watches if uuid not in added_watch_uuids]
+                logger.info(f"Queuing {len(watches_to_queue)} existing watches for recheck ({len(added_watch_uuids)} newly added watches already queued)")
+            else:
+                watches_to_queue = all_watches
+                logger.info(f"Queuing all {len(watches_to_queue)} watches for recheck")
+        else:
+            # Queue specific UUIDs
+            watches_to_queue = recheck_watches
+            logger.info(f"Queuing {len(watches_to_queue)} specific watches for recheck")
+
+        queued_count = 0
+        for watch_uuid in watches_to_queue:
+            if watch_uuid in datastore.data['watching']:
+                try:
+                    worker_pool.queue_item_async_safe(
+                        update_q,
+                        queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid})
+                    )
+                    queued_count += 1
+                    logger.debug(f"Queued watch for recheck: {watch_uuid}")
+                except Exception as e:
+                    logger.error(f"Failed to queue watch {watch_uuid}: {e}")
+            else:
+                logger.warning(f"Watch UUID not found in datastore: {watch_uuid}")
+
+        logger.success(f"Successfully queued {queued_count} watches for recheck")
+
+    # Step 4: Setup batch mode monitor (if -b was provided)
+    if batch_mode:
+        from changedetectionio.flask_app import update_q
+
+        # Safety check: Ensure Flask app is not already running on this port
+        # Batch mode should never run alongside the web server
+        import socket
+        test_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+
+        try:
+            # Try to bind to the configured host:port (no SO_REUSEADDR - strict check)
+            test_socket.bind((host, port))
+            test_socket.close()
+            logger.debug(f"Batch mode: Port {port} is available (Flask app not running)")
+        except OSError as e:
+            test_socket.close()
+            # errno 98 = EADDRINUSE (Linux)
+            # errno 48 = EADDRINUSE (macOS)
+            # errno 10048 = WSAEADDRINUSE (Windows)
+            if e.errno in (48, 98, 10048) or "Address already in use" in str(e) or "already in use" in str(e).lower():
+                logger.critical(f"ERROR: Batch mode cannot run - port {port} is already in use")
+                logger.critical(f"The Flask web server appears to be running on {host}:{port}")
+                logger.critical(f"Batch mode is designed for standalone operation (CI/CD, cron jobs, etc.)")
+                logger.critical(f"Please either stop the Flask web server, or use a different port with -p PORT")
+                sys.exit(1)
+            else:
+                # Some other socket error - log but continue (might be network configuration issue)
+                logger.warning(f"Port availability check failed with unexpected error: {e}")
+                logger.warning(f"Continuing with batch mode anyway - be aware of potential conflicts")
+
+        def queue_watches_for_recheck(datastore, iteration):
+            """Helper function to queue watches for recheck"""
+            watches_to_queue = []
+            if recheck_watches == 'all':
+                all_watches = list(datastore.data['watching'].keys())
+                if batch_mode and added_watch_uuids and iteration == 1:
+                    # Only exclude newly added watches on first iteration
+                    watches_to_queue = [uuid for uuid in all_watches if uuid not in added_watch_uuids]
+                else:
+                    watches_to_queue = all_watches
+                logger.info(f"Batch mode (iteration {iteration}): Queuing all {len(watches_to_queue)} watches")
+            elif recheck_watches:
+                watches_to_queue = recheck_watches
+                logger.info(f"Batch mode (iteration {iteration}): Queuing {len(watches_to_queue)} specific watches")
+
+            queued_count = 0
+            for watch_uuid in watches_to_queue:
+                if watch_uuid in datastore.data['watching']:
+                    try:
+                        worker_pool.queue_item_async_safe(
+                            update_q,
+                            queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid})
+                        )
+                        queued_count += 1
+                    except Exception as e:
+                        logger.error(f"Failed to queue watch {watch_uuid}: {e}")
+                else:
+                    logger.warning(f"Watch UUID not found in datastore: {watch_uuid}")
+            logger.success(f"Batch mode (iteration {iteration}): Successfully queued {queued_count} watches")
+            return queued_count
+
+        def batch_mode_monitor():
+            """Monitor queue and workers, shutdown or repeat when work is complete"""
+            import time
+
+            # Track iterations if repeat mode is enabled
+            current_iteration = 1
+            total_iterations = recheck_repeat_count if recheck_watches and recheck_repeat_count > 1 else 1
+
+            if total_iterations > 1:
+                logger.info(f"Batch mode: Will repeat recheck {total_iterations} times")
+            else:
+                logger.info("Batch mode: Waiting for all queued items to complete...")
+
+            # Wait a bit for workers to start processing
+            time.sleep(3)
+
+            try:
+                while current_iteration <= total_iterations:
+                    logger.info(f"Batch mode: Waiting for iteration {current_iteration}/{total_iterations} to complete...")
+
+                    # Use the shared wait_for_all_checks function
+                    completed = worker_pool.wait_for_all_checks(update_q, timeout=300)
+
+                    if not completed:
+                        logger.warning(f"Batch mode: Iteration {current_iteration} timed out after 300 seconds")
+
+                    logger.success(f"Batch mode: Iteration {current_iteration}/{total_iterations} completed")
+
+                    # Check if we need to repeat
+                    if current_iteration < total_iterations:
+                        logger.info(f"Batch mode: Starting iteration {current_iteration + 1}...")
+                        current_iteration += 1
+
+                        # Re-queue watches for next iteration
+                        queue_watches_for_recheck(datastore, current_iteration)
+
+                        # Brief pause before continuing
+                        time.sleep(2)
+                    else:
+                        # All iterations complete
+                        logger.success(f"Batch mode: All {total_iterations} iterations completed, initiating shutdown")
+                        # Trigger shutdown
+                        import os, signal
+                        os.kill(os.getpid(), signal.SIGTERM)
+                        return
+
+            except Exception as e:
+                logger.error(f"Batch mode monitor error: {e}")
+                logger.error(f"Initiating emergency shutdown")
+                import os, signal
+                os.kill(os.getpid(), signal.SIGTERM)
+
+        # Start monitor in background thread
+        monitor_thread = threading.Thread(target=batch_mode_monitor, daemon=True, name="BatchModeMonitor")
+        monitor_thread.start()
+        logger.info("Batch mode enabled: Will exit after all queued items are processed")
+
     # Get the SocketIO instance from the Flask app (created in flask_app.py)
     from changedetectionio.flask_app import socketio_server
     global socketio
@@ -274,19 +619,17 @@ def main():
     else:
         logger.info("SIGUSR1 handler only registered on Linux, skipped.")
 
-    # Go into cleanup mode
-    if do_cleanup:
-        datastore.remove_unused_snapshots()
-
     app.config['datastore_path'] = datastore_path
 
 
     @app.context_processor
     def inject_template_globals():
-        return dict(right_sticky="v{}".format(datastore.data['version_tag']),
+        return dict(right_sticky="v"+__version__,
                     new_version_available=app.config['NEW_VERSION_AVAILABLE'],
                     has_password=datastore.data['settings']['application']['password'] != False,
-                    socket_io_enabled=datastore.data['settings']['application']['ui'].get('socket_io_enabled', True)
+                    socket_io_enabled=datastore.data['settings']['application'].get('ui', {}).get('socket_io_enabled', True),
+                    all_paused=datastore.data['settings']['application'].get('all_paused', False),
+                    all_muted=datastore.data['settings']['application'].get('all_muted', False)
                     )
 
     # Monitored websites will not receive a Referer header when a user clicks on an outgoing link.
@@ -308,23 +651,43 @@ def main():
     if os.getenv('USE_X_SETTINGS'):
         logger.info("USE_X_SETTINGS is ENABLED")
         from werkzeug.middleware.proxy_fix import ProxyFix
-        app.wsgi_app = ProxyFix(app.wsgi_app, x_prefix=1, x_host=1)
+        app.wsgi_app = ProxyFix(
+            app.wsgi_app,
+            x_for=1,      # X-Forwarded-For (client IP)
+            x_proto=1,    # X-Forwarded-Proto (http/https)
+            x_host=1,     # X-Forwarded-Host (original host)
+            x_port=1,     # X-Forwarded-Port (original port)
+            x_prefix=1    # X-Forwarded-Prefix (URL prefix)
+        )
 
 
-    # SocketIO instance is already initialized in flask_app.py
-    if socketio_server:
-        if ssl_mode:
-            logger.success(f"SSL mode enabled, attempting to start with '{ssl_cert_file}' and '{ssl_privkey_file}' in {os.getcwd()}")
-            socketio.run(app, host=host, port=int(port), debug=False,
-                         ssl_context=(ssl_cert_file, ssl_privkey_file), allow_unsafe_werkzeug=True)
-        else:
-            socketio.run(app, host=host, port=int(port), debug=False, allow_unsafe_werkzeug=True)
+    # In batch mode, skip starting the HTTP server - just keep workers running
+    if batch_mode:
+        logger.info("Batch mode: Skipping HTTP server startup, workers will process queue")
+        logger.info("Batch mode: Main thread will wait for shutdown signal")
+        # Keep main thread alive until batch monitor triggers shutdown
+        try:
+            while True:
+                time.sleep(1)
+        except KeyboardInterrupt:
+            logger.info("Batch mode: Keyboard interrupt received")
+            pass
     else:
-        # Run Flask app without Socket.IO if disabled
-        logger.info("Starting Flask app without Socket.IO server")
-        if ssl_mode:
-            logger.success(f"SSL mode enabled, attempting to start with '{ssl_cert_file}' and '{ssl_privkey_file}' in {os.getcwd()}")
-            app.run(host=host, port=int(port), debug=False,
-                    ssl_context=(ssl_cert_file, ssl_privkey_file))
+        # Normal mode: Start HTTP server
+        # SocketIO instance is already initialized in flask_app.py
+        if socketio_server:
+            if ssl_mode:
+                logger.success(f"SSL mode enabled, attempting to start with '{ssl_cert_file}' and '{ssl_privkey_file}' in {os.getcwd()}")
+                socketio.run(app, host=host, port=int(port), debug=False,
+                             ssl_context=(ssl_cert_file, ssl_privkey_file), allow_unsafe_werkzeug=True)
+            else:
+                socketio.run(app, host=host, port=int(port), debug=False, allow_unsafe_werkzeug=True)
         else:
-            app.run(host=host, port=int(port), debug=False)
+            # Run Flask app without Socket.IO if disabled
+            logger.info("Starting Flask app without Socket.IO server")
+            if ssl_mode:
+                logger.success(f"SSL mode enabled, attempting to start with '{ssl_cert_file}' and '{ssl_privkey_file}' in {os.getcwd()}")
+                app.run(host=host, port=int(port), debug=False,
+                        ssl_context=(ssl_cert_file, ssl_privkey_file))
+            else:
+                app.run(host=host, port=int(port), debug=False)

changedetectionio/api/Import.py

@@ -4,6 +4,10 @@ from flask import request
 from functools import wraps
 from . import auth, validate_openapi_request
 from ..validate_url import is_safe_valid_url
+import json
+
+# Number of URLs above which import switches to background processing
+IMPORT_SWITCH_TO_BACKGROUND_THRESHOLD = 20
 
 
 def default_content_type(content_type='text/plain'):
@@ -19,6 +23,76 @@ def default_content_type(content_type='text/plain'):
     return decorator
 
 
+def convert_query_param_to_type(value, schema_property):
+    """
+    Convert a query parameter string to the appropriate type based on schema definition.
+
+    Args:
+        value: String value from query parameter
+        schema_property: Schema property definition with 'type' or 'anyOf' field
+
+    Returns:
+        Converted value in the appropriate type
+
+    Supports both OpenAPI 3.1 formats:
+    - type: [string, 'null']  (array format)
+    - anyOf: [{type: string}, {type: null}]  (anyOf format)
+    """
+    prop_type = schema_property.get('type')
+
+    # Handle OpenAPI 3.1 type arrays: type: [string, 'null']
+    if isinstance(prop_type, list):
+        # Use the first non-null type from the array
+        for t in prop_type:
+            if t != 'null':
+                prop_type = t
+                break
+        else:
+            prop_type = None
+
+    # Handle anyOf schemas (older format)
+    elif 'anyOf' in schema_property:
+        # Use the first non-null type from anyOf
+        for option in schema_property['anyOf']:
+            if option.get('type') and option.get('type') != 'null':
+                prop_type = option.get('type')
+                break
+        else:
+            prop_type = None
+
+    # Handle array type (e.g., notification_urls)
+    if prop_type == 'array':
+        # Support both comma-separated and JSON array format
+        if value.startswith('['):
+            try:
+                return json.loads(value)
+            except json.JSONDecodeError:
+                return [v.strip() for v in value.split(',')]
+        return [v.strip() for v in value.split(',')]
+
+    # Handle object type (e.g., time_between_check, headers)
+    elif prop_type == 'object':
+        try:
+            return json.loads(value)
+        except json.JSONDecodeError:
+            raise ValueError(f"Invalid JSON object for field: {value}")
+
+    # Handle boolean type
+    elif prop_type == 'boolean':
+        return strtobool(value)
+
+    # Handle integer type
+    elif prop_type == 'integer':
+        return int(value)
+
+    # Handle number type (float)
+    elif prop_type == 'number':
+        return float(value)
+
+    # Default: return as string
+    return value
+
+
 class Import(Resource):
     def __init__(self, **kwargs):
         # datastore is a black box dependency
@@ -28,40 +102,128 @@ class Import(Resource):
     @default_content_type('text/plain') #3547 #3542
     @validate_openapi_request('importWatches')
     def post(self):
-        """Import a list of watched URLs."""
+        """Import a list of watched URLs with optional watch configuration."""
+        from . import get_watch_schema_properties
+        # Special parameters that are NOT watch configuration
+        special_params = {'tag', 'tag_uuids', 'dedupe', 'proxy'}
 
         extras = {}
 
+        # Handle special 'proxy' parameter
         if request.args.get('proxy'):
             plist = self.datastore.proxy_list
             if not request.args.get('proxy') in plist:
-                return "Invalid proxy choice, currently supported proxies are '{}'".format(', '.join(plist)), 400
+                proxy_list_str = ', '.join(plist) if plist else 'none configured'
+                return f"Invalid proxy choice, currently supported proxies are '{proxy_list_str}'", 400
             else:
                 extras['proxy'] = request.args.get('proxy')
 
+        # Handle special 'dedupe' parameter
         dedupe = strtobool(request.args.get('dedupe', 'true'))
 
+        # Handle special 'tag' and 'tag_uuids' parameters
         tags = request.args.get('tag')
         tag_uuids = request.args.get('tag_uuids')
 
         if tag_uuids:
             tag_uuids = tag_uuids.split(',')
 
+        # Extract ALL other query parameters as watch configuration
+        # Get schema from OpenAPI spec (replaces old schema_create_watch)
+        schema_properties = get_watch_schema_properties()
+        for param_name, param_value in request.args.items():
+            # Skip special parameters
+            if param_name in special_params:
+                continue
+
+            # Skip if not in schema (unknown parameter)
+            if param_name not in schema_properties:
+                return f"Unknown watch configuration parameter: {param_name}", 400
+
+            # Convert to appropriate type based on schema
+            try:
+                converted_value = convert_query_param_to_type(param_value, schema_properties[param_name])
+                extras[param_name] = converted_value
+            except (ValueError, json.JSONDecodeError) as e:
+                return f"Invalid value for parameter '{param_name}': {str(e)}", 400
+
+        # Validate processor if provided
+        if 'processor' in extras:
+            from changedetectionio.processors import available_processors
+            available = [p[0] for p in available_processors()]
+            if extras['processor'] not in available:
+                return f"Invalid processor '{extras['processor']}'. Available processors: {', '.join(available)}", 400
+
+        # Validate fetch_backend if provided
+        if 'fetch_backend' in extras:
+            from changedetectionio.content_fetchers import available_fetchers
+            available = [f[0] for f in available_fetchers()]
+            # Also allow 'system' and extra_browser_* patterns
+            is_valid = (
+                extras['fetch_backend'] == 'system' or
+                extras['fetch_backend'] in available or
+                extras['fetch_backend'].startswith('extra_browser_')
+            )
+            if not is_valid:
+                return f"Invalid fetch_backend '{extras['fetch_backend']}'. Available: system, {', '.join(available)}", 400
+
+        # Validate notification_urls if provided
+        if 'notification_urls' in extras:
+            from wtforms import ValidationError
+            from changedetectionio.api.Notifications import validate_notification_urls
+            try:
+                validate_notification_urls(extras['notification_urls'])
+            except ValidationError as e:
+                return f"Invalid notification_urls: {str(e)}", 400
+
         urls = request.get_data().decode('utf8').splitlines()
-        added = []
+        # Clean and validate URLs upfront
+        urls_to_import = []
         for url in urls:
             url = url.strip()
             if not len(url):
                 continue
 
-            # If hosts that only contain alphanumerics are allowed ("localhost" for example)
+            # Validate URL
             if not is_safe_valid_url(url):
                 return f"Invalid or unsupported URL - {url}", 400
 
+            # Check for duplicates if dedupe is enabled
             if dedupe and self.datastore.url_exists(url):
                 continue
 
-            new_uuid = self.datastore.add_watch(url=url, extras=extras, tag=tags, tag_uuids=tag_uuids)
-            added.append(new_uuid)
+            urls_to_import.append(url)
 
-        return added
+        # For small imports, process synchronously for immediate feedback
+        if len(urls_to_import) < IMPORT_SWITCH_TO_BACKGROUND_THRESHOLD:
+            added = []
+            for url in urls_to_import:
+                new_uuid = self.datastore.add_watch(url=url, extras=extras, tag=tags, tag_uuids=tag_uuids)
+                added.append(new_uuid)
+            return added, 200
+
+        # For large imports (>= 20), process in background thread
+        else:
+            import threading
+            from loguru import logger
+
+            def import_watches_background():
+                """Background thread to import watches - discarded after completion."""
+                try:
+                    added_count = 0
+                    for url in urls_to_import:
+                        try:
+                            self.datastore.add_watch(url=url, extras=extras, tag=tags, tag_uuids=tag_uuids)
+                            added_count += 1
+                        except Exception as e:
+                            logger.error(f"Error importing URL {url}: {e}")
+
+                    logger.info(f"Background import complete: {added_count} watches created")
+                except Exception as e:
+                    logger.error(f"Error in background import: {e}")
+
+            # Start background thread and return immediately
+            thread = threading.Thread(target=import_watches_background, daemon=True, name="ImportWatches-Background")
+            thread.start()
+
+            return {'status': f'Importing {len(urls_to_import)} URLs in background', 'count': len(urls_to_import)}, 202

changedetectionio/api/Notifications.py

@@ -1,8 +1,6 @@
-from flask_expects_json import expects_json
 from flask_restful import Resource, abort
 from flask import request
 from . import auth, validate_openapi_request
-from . import schema_create_notification_urls, schema_delete_notification_urls
 
 class Notifications(Resource):
     def __init__(self, **kwargs):
@@ -22,7 +20,6 @@ class Notifications(Resource):
     
     @auth.check_token
     @validate_openapi_request('addNotifications')
-    @expects_json(schema_create_notification_urls)
     def post(self):
         """Create Notification URLs."""
 
@@ -50,7 +47,6 @@ class Notifications(Resource):
     
     @auth.check_token
     @validate_openapi_request('replaceNotifications')
-    @expects_json(schema_create_notification_urls)
     def put(self):
         """Replace Notification URLs."""
         json_data = request.get_json()
@@ -67,13 +63,12 @@ class Notifications(Resource):
 
         clean_urls = [url.strip() for url in notification_urls if isinstance(url, str)]
         self.datastore.data['settings']['application']['notification_urls'] = clean_urls
-        self.datastore.needs_write = True
+        self.datastore.commit()
 
         return {'notification_urls': clean_urls}, 200
         
     @auth.check_token
     @validate_openapi_request('deleteNotifications')
-    @expects_json(schema_delete_notification_urls)
     def delete(self):
         """Delete Notification URLs."""
 
@@ -95,7 +90,7 @@ class Notifications(Resource):
             abort(400, message="No matching notification URLs found.")
 
         self.datastore.data['settings']['application']['notification_urls'] = notification_urls
-        self.datastore.needs_write = True
+        self.datastore.commit()
 
         return 'OK', 204
     

changedetectionio/api/Spec.py

@@ -0,0 +1,21 @@
+import functools
+from flask import make_response
+from flask_restful import Resource
+
+
+@functools.cache
+def _get_spec_yaml():
+    """Build and cache the merged spec as a YAML string (only serialized once per process)."""
+    import yaml
+    from changedetectionio.api import build_merged_spec_dict
+    return yaml.dump(build_merged_spec_dict(), default_flow_style=False, allow_unicode=True)
+
+
+class Spec(Resource):
+    def get(self):
+        """Return the merged OpenAPI spec including all registered processor extensions."""
+        return make_response(
+            _get_spec_yaml(),
+            200,
+            {'Content-Type': 'application/yaml'}
+        )

changedetectionio/api/Tags.py

@@ -1,13 +1,13 @@
 from changedetectionio import queuedWatchMetaData
-from changedetectionio import worker_handler
-from flask_expects_json import expects_json
+from changedetectionio import worker_pool
 from flask_restful import abort, Resource
+from loguru import logger
 
+import threading
 from flask import request
 from . import auth
 
-# Import schemas from __init__.py
-from . import schema_tag, schema_create_tag, schema_update_tag, validate_openapi_request
+from . import validate_openapi_request
 
 
 class Tag(Resource):
@@ -17,38 +17,75 @@ class Tag(Resource):
         self.update_q = kwargs['update_q']
 
     # Get information about a single tag
-    # curl http://localhost:5000/api/v1/tag/<string:uuid>
+    # curl http://localhost:5000/api/v1/tag/<uuid_str:uuid>
     @auth.check_token
     @validate_openapi_request('getTag')
     def get(self, uuid):
         """Get data for a single tag/group, toggle notification muting, or recheck all."""
-        from copy import deepcopy
-        tag = deepcopy(self.datastore.data['settings']['application']['tags'].get(uuid))
+        tag = self.datastore.data['settings']['application']['tags'].get(uuid)
         if not tag:
             abort(404, message=f'No tag exists with the UUID of {uuid}')
 
         if request.args.get('recheck'):
-            # Recheck all, including muted
-            # Get most overdue first
-            i=0
+            # Recheck all watches with this tag, including muted
+            # First collect watches to queue
+            watches_to_queue = []
             for k in sorted(self.datastore.data['watching'].items(), key=lambda item: item[1].get('last_checked', 0)):
                 watch_uuid = k[0]
                 watch = k[1]
-                if not watch['paused'] and tag['uuid'] not in watch['tags']:
-                    continue
-                worker_handler.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
-                i+=1
+                if not watch['paused'] and tag['uuid'] in watch['tags']:
+                    watches_to_queue.append(watch_uuid)
 
-            return f"OK, {i} watches queued", 200
+            # If less than 20 watches, queue synchronously for immediate feedback
+            if len(watches_to_queue) < 20:
+                for watch_uuid in watches_to_queue:
+                    worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
+                return {'status': f'OK, queued {len(watches_to_queue)} watches for rechecking'}, 200
+            else:
+                # 20+ watches - queue in background thread to avoid blocking API response
+                def queue_watches_background():
+                    """Background thread to queue watches - discarded after completion."""
+                    try:
+                        for watch_uuid in watches_to_queue:
+                            worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
+                        logger.info(f"Background queueing complete for tag {tag['uuid']}: {len(watches_to_queue)} watches queued")
+                    except Exception as e:
+                        logger.error(f"Error in background queueing for tag {tag['uuid']}: {e}")
+
+                # Start background thread and return immediately
+                thread = threading.Thread(target=queue_watches_background, daemon=True, name=f"QueueTag-{tag['uuid'][:8]}")
+                thread.start()
+
+                return {'status': f'OK, queueing {len(watches_to_queue)} watches in background'}, 202
 
         if request.args.get('muted', '') == 'muted':
-            self.datastore.data['settings']['application']['tags'][uuid]['notification_muted'] = True
+            tag['notification_muted'] = True
+            tag.commit()
             return "OK", 200
         elif request.args.get('muted', '') == 'unmuted':
-            self.datastore.data['settings']['application']['tags'][uuid]['notification_muted'] = False
+            tag['notification_muted'] = False
+            tag.commit()
             return "OK", 200
 
-        return tag
+        # Filter out Watch-specific runtime fields that don't apply to Tags (yet)
+        # TODO: Future enhancement - aggregate these values from all Watches that have this tag:
+        #   - check_count: sum of all watches' check_count
+        #   - last_checked: most recent last_checked from all watches
+        #   - last_changed: most recent last_changed from all watches
+        #   - consecutive_filter_failures: count of watches with failures
+        #   - etc.
+        # These come from watch_base inheritance but currently have no meaningful value for Tags
+        watch_only_fields = {
+            'browser_steps_last_error_step', 'check_count', 'consecutive_filter_failures',
+            'content-type', 'fetch_time', 'last_changed', 'last_checked', 'last_error',
+            'last_notification_error', 'last_viewed', 'notification_alert_count',
+            'page_title', 'previous_md5', 'remote_server_reply'
+        }
+
+        # Create clean tag dict without Watch-specific fields
+        clean_tag = {k: v for k, v in tag.items() if k not in watch_only_fields}
+
+        return clean_tag
 
     @auth.check_token
     @validate_openapi_request('deleteTag')
@@ -59,38 +96,84 @@ class Tag(Resource):
 
         # Delete the tag, and any tag reference
         del self.datastore.data['settings']['application']['tags'][uuid]
-        
+
         # Remove tag from all watches
         for watch_uuid, watch in self.datastore.data['watching'].items():
             if watch.get('tags') and uuid in watch['tags']:
                 watch['tags'].remove(uuid)
+                watch.commit()
 
         return 'OK', 204
 
     @auth.check_token
     @validate_openapi_request('updateTag')
-    @expects_json(schema_update_tag)
     def put(self, uuid):
         """Update tag information."""
         tag = self.datastore.data['settings']['application']['tags'].get(uuid)
         if not tag:
             abort(404, message='No tag exists with the UUID of {}'.format(uuid))
 
-        tag.update(request.json)
-        self.datastore.needs_write_urgent = True
+        # Make a mutable copy of request.json for modification
+        json_data = dict(request.json)
+
+        # Validate notification_urls if provided
+        if 'notification_urls' in json_data:
+            from wtforms import ValidationError
+            from changedetectionio.api.Notifications import validate_notification_urls
+            try:
+                notification_urls = json_data.get('notification_urls', [])
+                validate_notification_urls(notification_urls)
+            except ValidationError as e:
+                return str(e), 400
+
+        # Filter out readOnly fields (extracted from OpenAPI spec Tag schema)
+        # These are system-managed fields that should never be user-settable
+        from . import get_readonly_tag_fields
+        readonly_fields = get_readonly_tag_fields()
+
+        # Tag model inherits from watch_base but has no @property attributes of its own
+        # So we only need to filter readOnly fields
+        for field in readonly_fields:
+            json_data.pop(field, None)
+
+        # Validate remaining fields - reject truly unknown fields
+        # Get valid fields from Tag schema
+        from . import get_tag_schema_properties
+        valid_fields = set(get_tag_schema_properties().keys())
+
+        # Check for unknown fields
+        unknown_fields = set(json_data.keys()) - valid_fields
+        if unknown_fields:
+            return f"Unknown field(s): {', '.join(sorted(unknown_fields))}", 400
+
+        tag.update(json_data)
+        tag.commit()
+
+        # Clear checksums for all watches using this tag to force reprocessing
+        # Tag changes affect inherited configuration
+        cleared_count = self.datastore.clear_checksums_for_tag(uuid)
+        logger.info(f"Tag {uuid} updated via API, cleared {cleared_count} watch checksums")
 
         return "OK", 200
 
 
     @auth.check_token
     @validate_openapi_request('createTag')
-    # Only cares for {'title': 'xxxx'}
     def post(self):
         """Create a single tag/group."""
 
         json_data = request.get_json()
         title = json_data.get("title",'').strip()
 
+        # Validate that only valid fields are provided
+        # Get valid fields from Tag schema
+        from . import get_tag_schema_properties
+        valid_fields = set(get_tag_schema_properties().keys())
+
+        # Check for unknown fields
+        unknown_fields = set(json_data.keys()) - valid_fields
+        if unknown_fields:
+            return f"Unknown field(s): {', '.join(sorted(unknown_fields))}", 400
 
         new_uuid = self.datastore.add_tag(title=title)
         if new_uuid:

changedetectionio/api/Watch.py

@@ -1,18 +1,18 @@
 import os
+import threading
 
 from changedetectionio.validate_url import is_safe_valid_url
+from changedetectionio.favicon_utils import get_favicon_mime_type
 
 from . import auth
 from changedetectionio import queuedWatchMetaData, strtobool
-from changedetectionio import worker_handler
+from changedetectionio import worker_pool
 from flask import request, make_response, send_from_directory
-from flask_expects_json import expects_json
 from flask_restful import abort, Resource
 from loguru import logger
 import copy
 
-# Import schemas from __init__.py
-from . import schema, schema_create_watch, schema_update_watch, validate_openapi_request
+from . import validate_openapi_request, get_readonly_watch_fields
 from ..notification import valid_notification_formats
 from ..notification.handler import newline_re
 
@@ -57,50 +57,53 @@ class Watch(Resource):
         self.update_q = kwargs['update_q']
 
     # Get information about a single watch, excluding the history list (can be large)
-    # curl http://localhost:5000/api/v1/watch/<string:uuid>
+    # curl http://localhost:5000/api/v1/watch/<uuid_str:uuid>
     # @todo - version2 - ?muted and ?paused should be able to be called together, return the watch struct not "OK"
     # ?recheck=true
     @auth.check_token
     @validate_openapi_request('getWatch')
     def get(self, uuid):
         """Get information about a single watch, recheck, pause, or mute."""
-        import time
-        from copy import deepcopy
-        watch = None
-        for _ in range(20):
-            try:
-                watch = deepcopy(self.datastore.data['watching'].get(uuid))
-                break
-            except RuntimeError:
-                # Incase dict changed, try again
-                time.sleep(0.01)
-
-        if not watch:
+        # Get watch reference first (for pause/mute operations)
+        watch_obj = self.datastore.data['watching'].get(uuid)
+        if not watch_obj:
             abort(404, message='No watch exists with the UUID of {}'.format(uuid))
 
+        # Create a dict copy for JSON response (with lock for thread safety)
+        # This is much faster than deepcopy and doesn't copy the datastore reference
+        # WARNING: dict() is a SHALLOW copy - nested dicts are shared with original!
+        # Only safe because we only ADD scalar properties (line 97-101), never modify nested dicts
+        # If you need to modify nested dicts, use: from copy import deepcopy; watch = deepcopy(dict(watch_obj))
+        with self.datastore.lock:
+            watch = dict(watch_obj)
+
         if request.args.get('recheck'):
-            worker_handler.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+            worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
             return "OK", 200
         if request.args.get('paused', '') == 'paused':
-            self.datastore.data['watching'].get(uuid).pause()
+            watch_obj.pause()
+            watch_obj.commit()
             return "OK", 200
         elif request.args.get('paused', '') == 'unpaused':
-            self.datastore.data['watching'].get(uuid).unpause()
+            watch_obj.unpause()
+            watch_obj.commit()
             return "OK", 200
         if request.args.get('muted', '') == 'muted':
-            self.datastore.data['watching'].get(uuid).mute()
+            watch_obj.mute()
+            watch_obj.commit()
             return "OK", 200
         elif request.args.get('muted', '') == 'unmuted':
-            self.datastore.data['watching'].get(uuid).unmute()
+            watch_obj.unmute()
+            watch_obj.commit()
             return "OK", 200
 
         # Return without history, get that via another API call
         # Properties are not returned as a JSON, so add the required props manually
-        watch['history_n'] = watch.history_n
+        watch['history_n'] = watch_obj.history_n
         # attr .last_changed will check for the last written text snapshot on change
-        watch['last_changed'] = watch.last_changed
-        watch['viewed'] = watch.viewed
-        watch['link'] = watch.link,
+        watch['last_changed'] = watch_obj.last_changed
+        watch['viewed'] = watch_obj.viewed
+        watch['link'] = watch_obj.link,
 
         return watch
 
@@ -116,7 +119,6 @@ class Watch(Resource):
 
     @auth.check_token
     @validate_openapi_request('updateWatch')
-    @expects_json(schema_update_watch)
     def put(self, uuid):
         """Update watch information."""
         watch = self.datastore.data['watching'].get(uuid)
@@ -125,72 +127,86 @@ class Watch(Resource):
 
         if request.json.get('proxy'):
             plist = self.datastore.proxy_list
-            if not request.json.get('proxy') in plist:
-                return "Invalid proxy choice, currently supported proxies are '{}'".format(', '.join(plist)), 400
+            if not plist or request.json.get('proxy') not in plist:
+                proxy_list_str = ', '.join(plist) if plist else 'none configured'
+                return f"Invalid proxy choice, currently supported proxies are '{proxy_list_str}'", 400
 
         # Validate time_between_check when not using defaults
         validation_error = validate_time_between_check_required(request.json)
         if validation_error:
             return validation_error, 400
 
-        # XSS etc protection
-        if request.json.get('url') and not is_safe_valid_url(request.json.get('url')):
-            return "Invalid URL", 400
+        # Validate notification_urls if provided
+        if 'notification_urls' in request.json:
+            from wtforms import ValidationError
+            from changedetectionio.api.Notifications import validate_notification_urls
+            try:
+                notification_urls = request.json.get('notification_urls', [])
+                validate_notification_urls(notification_urls)
+            except ValidationError as e:
+                return str(e), 400
+
+        # XSS etc protection - validate URL if it's being updated
+        if 'url' in request.json:
+            new_url = request.json.get('url')
+
+            # URL must be a non-empty string
+            if new_url is None:
+                return "URL cannot be null", 400
+
+            if not isinstance(new_url, str):
+                return "URL must be a string", 400
+
+            if not new_url.strip():
+                return "URL cannot be empty or whitespace only", 400
+
+            if not is_safe_valid_url(new_url.strip()):
+                return "Invalid or unsupported URL format. URL must use http://, https://, or ftp:// protocol", 400
 
         # Handle processor-config-* fields separately (save to JSON, not datastore)
         from changedetectionio import processors
-        processor_config_data = {}
-        regular_data = {}
 
-        for key, value in request.json.items():
-            if key.startswith('processor_config_'):
-                config_key = key.replace('processor_config_', '')
-                if value:  # Only save non-empty values
-                    processor_config_data[config_key] = value
-            else:
-                regular_data[key] = value
+        # Make a mutable copy of request.json for modification
+        json_data = dict(request.json)
+
+        # Extract and remove processor config fields from json_data
+        processor_config_data = processors.extract_processor_config_from_form_data(json_data)
+
+        # Filter out readOnly fields (extracted from OpenAPI spec Watch schema)
+        # These are system-managed fields that should never be user-settable
+        readonly_fields = get_readonly_watch_fields()
+
+        # Also filter out @property attributes (computed/derived values from the model)
+        # These are not stored and should be ignored in PUT requests
+        from changedetectionio.model.Watch import model as WatchModel
+        property_fields = WatchModel.get_property_names()
+
+        # Combine both sets of fields to ignore
+        fields_to_ignore = readonly_fields | property_fields
+
+        # Remove all ignored fields from update data
+        for field in fields_to_ignore:
+            json_data.pop(field, None)
+
+        # Validate remaining fields - reject truly unknown fields
+        # Get valid fields from WatchBase schema
+        from . import get_watch_schema_properties
+        valid_fields = set(get_watch_schema_properties().keys())
+
+        # Also allow last_viewed (explicitly defined in UpdateWatch schema)
+        valid_fields.add('last_viewed')
+
+        # Check for unknown fields
+        unknown_fields = set(json_data.keys()) - valid_fields
+        if unknown_fields:
+            return f"Unknown field(s): {', '.join(sorted(unknown_fields))}", 400
 
         # Update watch with regular (non-processor-config) fields
-        watch.update(regular_data)
+        watch.update(json_data)
+        watch.commit()
 
-        # Save processor config to JSON file if any config data exists
-        if processor_config_data:
-            try:
-                processor_name = request.json.get('processor', watch.get('processor'))
-                if processor_name:
-                    # Create a processor instance to access config methods
-                    from changedetectionio.processors import difference_detection_processor
-                    processor_instance = difference_detection_processor(self.datastore, uuid)
-                    # Use processor name as filename so each processor keeps its own config
-                    config_filename = f'{processor_name}.json'
-                    processor_instance.update_extra_watch_config(config_filename, processor_config_data)
-                    logger.debug(f"API: Saved processor config to {config_filename}: {processor_config_data}")
-
-                    # Call optional edit_hook if processor has one
-                    try:
-                        import importlib
-                        edit_hook_module_name = f'changedetectionio.processors.{processor_name}.edit_hook'
-
-                        try:
-                            edit_hook = importlib.import_module(edit_hook_module_name)
-                            logger.debug(f"API: Found edit_hook module for {processor_name}")
-
-                            if hasattr(edit_hook, 'on_config_save'):
-                                logger.info(f"API: Calling edit_hook.on_config_save for {processor_name}")
-                                # Call hook and get updated config
-                                updated_config = edit_hook.on_config_save(watch, processor_config_data, self.datastore)
-                                # Save updated config back to file
-                                processor_instance.update_extra_watch_config(config_filename, updated_config)
-                                logger.info(f"API: Edit hook updated config: {updated_config}")
-                            else:
-                                logger.debug(f"API: Edit hook module found but no on_config_save function")
-                        except ModuleNotFoundError:
-                            logger.debug(f"API: No edit_hook module for processor {processor_name} (this is normal)")
-                    except Exception as hook_error:
-                        logger.error(f"API: Edit hook error (non-fatal): {hook_error}", exc_info=True)
-
-            except Exception as e:
-                logger.error(f"API: Failed to save processor config: {e}")
+        # Save processor config to JSON file
+        processors.save_processor_config(self.datastore, uuid, processor_config_data)
 
         return "OK", 200
 
@@ -201,7 +217,7 @@ class WatchHistory(Resource):
         self.datastore = kwargs['datastore']
 
     # Get a list of available history for a watch by UUID
-    # curl http://localhost:5000/api/v1/watch/<string:uuid>/history
+    # curl http://localhost:5000/api/v1/watch/<uuid_str:uuid>/history
     @auth.check_token
     @validate_openapi_request('getWatchHistory')
     def get(self, uuid):
@@ -231,6 +247,10 @@ class WatchSingleHistory(Resource):
         if timestamp == 'latest':
             timestamp = list(watch.history.keys())[-1]
 
+        # Validate that the timestamp exists in history
+        if timestamp not in watch.history:
+            abort(404, message=f"No history snapshot found for timestamp '{timestamp}'")
+
         if request.args.get('html'):
             content = watch.get_fetched_html(timestamp)
             if content:
@@ -379,18 +399,11 @@ class WatchFavicon(Resource):
 
         favicon_filename = watch.get_favicon_filename()
         if favicon_filename:
-            try:
-                import magic
-                mime = magic.from_file(
-                    os.path.join(watch.watch_data_dir, favicon_filename),
-                    mime=True
-                )
-            except ImportError:
-                # Fallback, no python-magic
-                import mimetypes
-                mime, encoding = mimetypes.guess_type(favicon_filename)
+            # Use cached MIME type detection
+            filepath = os.path.join(watch.data_dir, favicon_filename)
+            mime = get_favicon_mime_type(filepath)
 
-            response = make_response(send_from_directory(watch.watch_data_dir, favicon_filename))
+            response = make_response(send_from_directory(watch.data_dir, favicon_filename))
             response.headers['Content-type'] = mime
             response.headers['Cache-Control'] = 'max-age=300, must-revalidate'  # Cache for 5 minutes, then revalidate
             return response
@@ -406,7 +419,6 @@ class CreateWatch(Resource):
 
     @auth.check_token
     @validate_openapi_request('createWatch')
-    @expects_json(schema_create_watch)
     def post(self):
         """Create a single watch."""
 
@@ -418,16 +430,33 @@ class CreateWatch(Resource):
 
         if json_data.get('proxy'):
             plist = self.datastore.proxy_list
-            if not json_data.get('proxy') in plist:
-                return "Invalid proxy choice, currently supported proxies are '{}'".format(', '.join(plist)), 400
+            if not plist or json_data.get('proxy') not in plist:
+                proxy_list_str = ', '.join(plist) if plist else 'none configured'
+                return f"Invalid proxy choice, currently supported proxies are '{proxy_list_str}'", 400
 
         # Validate time_between_check when not using defaults
         validation_error = validate_time_between_check_required(json_data)
         if validation_error:
             return validation_error, 400
 
+        # Validate notification_urls if provided
+        if 'notification_urls' in json_data:
+            from wtforms import ValidationError
+            from changedetectionio.api.Notifications import validate_notification_urls
+            try:
+                notification_urls = json_data.get('notification_urls', [])
+                validate_notification_urls(notification_urls)
+            except ValidationError as e:
+                return str(e), 400
+
+        # Handle processor-config-* fields separately (save to JSON, not watch)
+        from changedetectionio import processors
+
         extras = copy.deepcopy(json_data)
 
+        # Extract and remove processor config fields from extras
+        processor_config_data = processors.extract_processor_config_from_form_data(extras)
+
         # Because we renamed 'tag' to 'tags' but don't want to change the API (can do this in v2 of the API)
         tags = None
         if extras.get('tag'):
@@ -437,10 +466,25 @@ class CreateWatch(Resource):
         del extras['url']
 
         new_uuid = self.datastore.add_watch(url=url, extras=extras, tag=tags)
+
+        # Save processor config to separate JSON file
+        if new_uuid and processor_config_data:
+            processors.save_processor_config(self.datastore, new_uuid, processor_config_data)
         if new_uuid:
-            worker_handler.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': new_uuid}))
+# Dont queue because the scheduler will check that it hasnt been checked before anyway
+#            worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': new_uuid}))
             return {'uuid': new_uuid}, 201
         else:
+            # Check if it was a limit issue
+            page_watch_limit = os.getenv('PAGE_WATCH_LIMIT')
+            if page_watch_limit:
+                try:
+                    page_watch_limit = int(page_watch_limit)
+                    current_watch_count = len(self.datastore.data['watching'])
+                    if current_watch_count >= page_watch_limit:
+                        return f"Watch limit reached ({current_watch_count}/{page_watch_limit} watches). Cannot add more watches.", 429
+                except ValueError:
+                    pass
             return "Invalid or unsupported URL", 400
 
     @auth.check_token
@@ -462,14 +506,65 @@ class CreateWatch(Resource):
                 'last_error': watch['last_error'],
                 'link': watch.link,
                 'page_title': watch['page_title'],
+                'tags': [*tags],  # Unpack dict keys to list (can't use list() since variable named 'list')
                 'title': watch['title'],
                 'url': watch['url'],
                 'viewed': watch.viewed
             }
 
         if request.args.get('recheck_all'):
-            for uuid in self.datastore.data['watching'].keys():
-                worker_handler.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
-            return {'status': "OK"}, 200
+            # Collect all watches to queue
+            watches_to_queue = self.datastore.data['watching'].keys()
+
+            # If less than 20 watches, queue synchronously for immediate feedback
+            if len(watches_to_queue) < 20:
+                # Get already queued/running UUIDs once (efficient)
+                queued_uuids = set(self.update_q.get_queued_uuids())
+                running_uuids = set(worker_pool.get_running_uuids())
+
+                # Filter out watches that are already queued or running
+                watches_to_queue_filtered = [
+                    uuid for uuid in watches_to_queue
+                    if uuid not in queued_uuids and uuid not in running_uuids
+                ]
+
+                # Queue only the filtered watches
+                for uuid in watches_to_queue_filtered:
+                    worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+
+                # Provide feedback about skipped watches
+                skipped_count = len(watches_to_queue) - len(watches_to_queue_filtered)
+                if skipped_count > 0:
+                    return {'status': f'OK, queued {len(watches_to_queue_filtered)} watches for rechecking ({skipped_count} already queued or running)'}, 200
+                else:
+                    return {'status': f'OK, queued {len(watches_to_queue_filtered)} watches for rechecking'}, 200
+            else:
+                # 20+ watches - queue in background thread to avoid blocking API response
+                # Capture queued/running state before background thread
+                queued_uuids = set(self.update_q.get_queued_uuids())
+                running_uuids = set(worker_pool.get_running_uuids())
+
+                def queue_all_watches_background():
+                    """Background thread to queue all watches - discarded after completion."""
+                    try:
+                        queued_count = 0
+                        skipped_count = 0
+                        for uuid in watches_to_queue:
+                            # Check if already queued or running (state captured at start)
+                            if uuid not in queued_uuids and uuid not in running_uuids:
+                                worker_pool.queue_item_async_safe(self.update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                                queued_count += 1
+                            else:
+                                skipped_count += 1
+
+                        logger.info(f"Background queueing complete: {queued_count} watches queued, {skipped_count} skipped (already queued/running)")
+                    except Exception as e:
+                        logger.error(f"Error in background queueing all watches: {e}")
+
+                # Start background thread and return immediately
+                thread = threading.Thread(target=queue_all_watches_background, daemon=True, name="QueueAllWatches-Background")
+                thread.start()
+
+                return {'status': f'OK, queueing {len(watches_to_queue)} watches in background'}, 202
 
         return list, 200

changedetectionio/api/__init__.py

@@ -1,50 +1,137 @@
-import copy
 import functools
 from flask import request, abort
 from loguru import logger
-from . import api_schema
-from ..model import watch_base
-
-# Build a JSON Schema atleast partially based on our Watch model
-watch_base_config = watch_base()
-schema = api_schema.build_watch_json_schema(watch_base_config)
-
-schema_create_watch = copy.deepcopy(schema)
-schema_create_watch['required'] = ['url']
-del schema_create_watch['properties']['last_viewed']
-
-schema_update_watch = copy.deepcopy(schema)
-schema_update_watch['additionalProperties'] = False
-
-# Tag schema is also based on watch_base since Tag inherits from it
-schema_tag = copy.deepcopy(schema)
-schema_create_tag = copy.deepcopy(schema_tag)
-schema_create_tag['required'] = ['title']
-schema_update_tag = copy.deepcopy(schema_tag)
-schema_update_tag['additionalProperties'] = False
-
-schema_notification_urls = copy.deepcopy(schema)
-schema_create_notification_urls = copy.deepcopy(schema_notification_urls)
-schema_create_notification_urls['required'] = ['notification_urls']
-schema_delete_notification_urls = copy.deepcopy(schema_notification_urls)
-schema_delete_notification_urls['required'] = ['notification_urls']
 
 @functools.cache
-def get_openapi_spec():
-    """Lazy load OpenAPI spec and dependencies only when validation is needed."""
+def build_merged_spec_dict():
+    """
+    Load the base OpenAPI spec and merge in any per-processor api.yaml extensions.
+
+    Each processor can provide an api.yaml file alongside its __init__.py that defines
+    additional schemas (e.g., processor_config_restock_diff). These are merged into
+    WatchBase.properties so the spec accurately reflects what the API accepts.
+
+    Plugin processors (via pluggy) are also supported - they just need an api.yaml
+    next to their processor module.
+
+    Returns the merged dict (cached - do not mutate the returned value).
+    """
     import os
-    import yaml  # Lazy import - only loaded when API validation is actually used
-    from openapi_core import OpenAPI  # Lazy import - saves ~10.7 MB on startup
+    import yaml
 
     spec_path = os.path.join(os.path.dirname(__file__), '../../docs/api-spec.yaml')
     if not os.path.exists(spec_path):
-        # Possibly for pip3 packages
         spec_path = os.path.join(os.path.dirname(__file__), '../docs/api-spec.yaml')
 
     with open(spec_path, 'r', encoding='utf-8') as f:
         spec_dict = yaml.safe_load(f)
-    _openapi_spec = OpenAPI.from_dict(spec_dict)
-    return _openapi_spec
+
+    try:
+        from changedetectionio.processors import find_processors, get_parent_module
+        for module, proc_name in find_processors():
+            parent = get_parent_module(module)
+            if not parent or not hasattr(parent, '__file__'):
+                continue
+            api_yaml_path = os.path.join(os.path.dirname(parent.__file__), 'api.yaml')
+            if not os.path.exists(api_yaml_path):
+                continue
+            with open(api_yaml_path, 'r', encoding='utf-8') as f:
+                proc_spec = yaml.safe_load(f)
+            # Merge schemas
+            proc_schemas = proc_spec.get('components', {}).get('schemas', {})
+            spec_dict['components']['schemas'].update(proc_schemas)
+            # Inject processor_config_{name} into WatchBase if the schema is defined
+            schema_key = f'processor_config_{proc_name}'
+            if schema_key in proc_schemas:
+                spec_dict['components']['schemas']['WatchBase']['properties'][schema_key] = {
+                    '$ref': f'#/components/schemas/{schema_key}'
+                }
+            # Append x-code-samples from processor paths into existing path operations
+            for path, path_item in proc_spec.get('paths', {}).items():
+                if path not in spec_dict.get('paths', {}):
+                    continue
+                for method, operation in path_item.items():
+                    if method not in spec_dict['paths'][path]:
+                        continue
+                    if 'x-code-samples' in operation:
+                        existing = spec_dict['paths'][path][method].get('x-code-samples', [])
+                        spec_dict['paths'][path][method]['x-code-samples'] = existing + operation['x-code-samples']
+    except Exception as e:
+        logger.warning(f"Failed to merge processor API specs: {e}")
+
+    return spec_dict
+
+
+@functools.cache
+def get_openapi_spec():
+    """Lazy load OpenAPI spec and dependencies only when validation is needed."""
+    from openapi_core import OpenAPI  # Lazy import - saves ~10.7 MB on startup
+    return OpenAPI.from_dict(build_merged_spec_dict())
+
+@functools.cache
+def get_openapi_schema_dict():
+    """
+    Get the raw OpenAPI spec dictionary for schema access.
+
+    Used by Import endpoint to validate and convert query parameters.
+    Returns the merged YAML dict (not the OpenAPI object).
+    """
+    return build_merged_spec_dict()
+
+@functools.cache
+def _resolve_schema_properties(schema_name):
+    """
+    Generic helper to resolve schema properties, including allOf inheritance.
+
+    Args:
+        schema_name: Name of the schema (e.g., 'WatchBase', 'Watch', 'Tag')
+
+    Returns:
+        dict: All properties including inherited ones from $ref schemas
+    """
+    spec_dict = get_openapi_schema_dict()
+    schema = spec_dict['components']['schemas'].get(schema_name, {})
+
+    properties = {}
+
+    # Handle allOf (schema inheritance)
+    if 'allOf' in schema:
+        for item in schema['allOf']:
+            # Resolve $ref to parent schema
+            if '$ref' in item:
+                ref_path = item['$ref'].split('/')[-1]
+                ref_schema = spec_dict['components']['schemas'].get(ref_path, {})
+                properties.update(ref_schema.get('properties', {}))
+            # Add schema-specific properties
+            if 'properties' in item:
+                properties.update(item['properties'])
+    else:
+        # Direct properties (no inheritance)
+        properties = schema.get('properties', {})
+
+    return properties
+
+
+@functools.cache
+def get_watch_schema_properties():
+    """
+    Extract watch schema properties from OpenAPI spec for Import endpoint.
+
+    Returns WatchBase properties (all writable Watch fields).
+    """
+    return _resolve_schema_properties('WatchBase')
+
+# Import readonly field utilities from shared module (avoids circular dependencies with model layer)
+from changedetectionio.model.schema_utils import get_readonly_watch_fields, get_readonly_tag_fields
+
+@functools.cache
+def get_tag_schema_properties():
+    """
+    Extract Tag schema properties from OpenAPI spec.
+
+    Returns WatchBase properties + Tag-specific properties (overrides_watch).
+    """
+    return _resolve_schema_properties('Tag')
 
 def validate_openapi_request(operation_id):
     """Decorator to validate incoming requests against OpenAPI spec."""
@@ -57,6 +144,7 @@ def validate_openapi_request(operation_id):
                 if request.method.upper() != 'GET':
                     # Lazy import - only loaded when actually validating a request
                     from openapi_core.contrib.flask import FlaskOpenAPIRequest
+                    from openapi_core.templating.paths.exceptions import ServerNotFound, PathNotFound, PathError
 
                     spec = get_openapi_spec()
                     openapi_request = FlaskOpenAPIRequest(request)
@@ -64,8 +152,29 @@ def validate_openapi_request(operation_id):
                     if result.errors:
                         error_details = []
                         for error in result.errors:
-                            error_details.append(str(error))
-                        raise BadRequest(f"OpenAPI validation failed: {error_details}")
+                            # Skip path/server validation errors for reverse proxy compatibility
+                            # Flask routing already validates that endpoints exist (returns 404 if not).
+                            # OpenAPI validation here is primarily for request body schema validation.
+                            # When behind nginx/reverse proxy, URLs may have path prefixes that don't
+                            # match the OpenAPI server definitions, causing false positives.
+                            if isinstance(error, PathError):
+                                logger.debug(f"API Call - Skipping path/server validation (delegated to Flask): {error}")
+                                continue
+
+                            error_str = str(error)
+                            # Extract detailed schema errors from __cause__
+                            if hasattr(error, '__cause__') and hasattr(error.__cause__, 'schema_errors'):
+                                for schema_error in error.__cause__.schema_errors:
+                                    field = '.'.join(str(p) for p in schema_error.path) if schema_error.path else 'body'
+                                    msg = schema_error.message if hasattr(schema_error, 'message') else str(schema_error)
+                                    error_details.append(f"{field}: {msg}")
+                            else:
+                                error_details.append(error_str)
+
+                        # Only raise if we have actual validation errors (not path/server issues)
+                        if error_details:
+                            logger.error(f"API Call - Validation failed: {'; '.join(error_details)}")
+                            raise BadRequest(f"Validation failed: {'; '.join(error_details)}")
             except BadRequest:
                 # Re-raise BadRequest exceptions (validation failures)
                 raise
@@ -82,5 +191,6 @@ from .Watch import Watch, WatchHistory, WatchSingleHistory, WatchHistoryDiff, Cr
 from .Tags import Tags, Tag
 from .Import import Import
 from .SystemInfo import SystemInfo
+from .Spec import Spec
 from .Notifications import Notifications
 

changedetectionio/api/api_schema.py

@@ -1,162 +0,0 @@
-# Responsible for building the storage dict into a set of rules ("JSON Schema") acceptable via the API
-# Probably other ways to solve this when the backend switches to some ORM
-from changedetectionio.notification import valid_notification_formats
-
-
-def build_time_between_check_json_schema():
-    # Setup time between check schema
-    schema_properties_time_between_check = {
-        "type": "object",
-        "additionalProperties": False,
-        "properties": {}
-    }
-    for p in ['weeks', 'days', 'hours', 'minutes', 'seconds']:
-        schema_properties_time_between_check['properties'][p] = {
-            "anyOf": [
-                {
-                    "type": "integer"
-                },
-                {
-                    "type": "null"
-                }
-            ]
-        }
-
-    return schema_properties_time_between_check
-
-def build_watch_json_schema(d):
-    # Base JSON schema
-    schema = {
-        'type': 'object',
-        'properties': {},
-    }
-
-    for k, v in d.items():
-        # @todo 'integer' is not covered here because its almost always for internal usage
-
-        if isinstance(v, type(None)):
-            schema['properties'][k] = {
-                "anyOf": [
-                    {"type": "null"},
-                ]
-            }
-        elif isinstance(v, list):
-            schema['properties'][k] = {
-                "anyOf": [
-                    {"type": "array",
-                     # Always is an array of strings, like text or regex or something
-                     "items": {
-                         "type": "string",
-                         "maxLength": 5000
-                     }
-                     },
-                ]
-            }
-        elif isinstance(v, bool):
-            schema['properties'][k] = {
-                "anyOf": [
-                    {"type": "boolean"},
-                ]
-            }
-        elif isinstance(v, str):
-            schema['properties'][k] = {
-                "anyOf": [
-                    {"type": "string",
-                     "maxLength": 5000},
-                ]
-            }
-
-    # Can also be a string (or None by default above)
-    for v in ['body',
-              'notification_body',
-              'notification_format',
-              'notification_title',
-              'proxy',
-              'tag',
-              'title',
-              'webdriver_js_execute_code'
-              ]:
-        schema['properties'][v]['anyOf'].append({'type': 'string', "maxLength": 5000})
-
-    for v in ['last_viewed']:
-        schema['properties'][v] = {
-            "type": "integer",
-            "description": "Unix timestamp in seconds of the last time the watch was viewed.",
-            "minimum": 0
-        }
-
-    # None or Boolean
-    schema['properties']['track_ldjson_price_data']['anyOf'].append({'type': 'boolean'})
-
-    schema['properties']['method'] = {"type": "string",
-                                      "enum": ["GET", "POST", "DELETE", "PUT"]
-                                      }
-
-    schema['properties']['fetch_backend']['anyOf'].append({"type": "string",
-                                                           "enum": ["html_requests", "html_webdriver"]
-                                                           })
-
-    schema['properties']['processor'] = {"anyOf": [
-        {"type": "string", "enum": ["restock_diff", "text_json_diff"]},
-        {"type": "null"}
-    ]}
-
-    # All headers must be key/value type dict
-    schema['properties']['headers'] = {
-        "type": "object",
-        "patternProperties": {
-            # Should always be a string:string type value
-            ".*": {"type": "string"},
-        }
-    }
-
-    schema['properties']['notification_format'] = {'type': 'string',
-                                                   'enum': list(valid_notification_formats.keys())
-                                                   }
-
-    # Stuff that shouldn't be available but is just state-storage
-    for v in ['previous_md5', 'last_error', 'has_ldjson_price_data', 'previous_md5_before_filters', 'uuid']:
-        del schema['properties'][v]
-
-    schema['properties']['webdriver_delay']['anyOf'].append({'type': 'integer'})
-
-    schema['properties']['time_between_check'] = build_time_between_check_json_schema()
-
-    schema['properties']['time_between_check_use_default'] = {
-        "type": "boolean",
-        "default": True,
-        "description": "Whether to use global settings for time between checks - defaults to true if not set"
-    }
-
-    schema['properties']['browser_steps'] = {
-        "anyOf": [
-            {
-                "type": "array",
-                "items": {
-                    "type": "object",
-                    "properties": {
-                        "operation": {
-                            "type": ["string", "null"],
-                            "maxLength": 5000  # Allows null and any string up to 5000 chars (including "")
-                        },
-                        "selector": {
-                            "type": ["string", "null"],
-                            "maxLength": 5000
-                        },
-                        "optional_value": {
-                            "type": ["string", "null"],
-                            "maxLength": 5000
-                        }
-                    },
-                    "required": ["operation", "selector", "optional_value"],
-                    "additionalProperties": False  # No extra keys allowed
-                }
-            },
-            {"type": "null"},  # Allows null for `browser_steps`
-            {"type": "array", "maxItems": 0}  # Allows empty array []
-        ]
-    }
-
-    # headers ?
-    return schema
-

changedetectionio/blueprint/backups/__init__.py

@@ -13,7 +13,7 @@ from loguru import logger
 BACKUP_FILENAME_FORMAT = "changedetection-backup-{}.zip"
 
 
-def create_backup(datastore_path, watches: dict):
+def create_backup(datastore_path, watches: dict, tags: dict = None):
     logger.debug("Creating backup...")
     import zipfile
     from pathlib import Path
@@ -27,15 +27,31 @@ def create_backup(datastore_path, watches: dict):
                          compression=zipfile.ZIP_DEFLATED,
                          compresslevel=8) as zipObj:
 
-        # Add the index
-        zipObj.write(os.path.join(datastore_path, "url-watches.json"), arcname="url-watches.json")
+        # Add the settings file (supports both formats)
+        # New format: changedetection.json
+        changedetection_json = os.path.join(datastore_path, "changedetection.json")
+        if os.path.isfile(changedetection_json):
+            zipObj.write(changedetection_json, arcname="changedetection.json")
+            logger.debug("Added changedetection.json to backup")
 
-        # Add the flask app secret
-        zipObj.write(os.path.join(datastore_path, "secret.txt"), arcname="secret.txt")
+        # Legacy format: url-watches.json (for backward compatibility)
+        url_watches_json = os.path.join(datastore_path, "url-watches.json")
+        if os.path.isfile(url_watches_json):
+            zipObj.write(url_watches_json, arcname="url-watches.json")
+            logger.debug("Added url-watches.json to backup")
+
+        # Add tag data directories (each tag has its own {uuid}/tag.json)
+        for uuid, tag in (tags or {}).items():
+            for f in Path(tag.data_dir).glob('*'):
+                zipObj.write(f,
+                             arcname=os.path.join(f.parts[-2], f.parts[-1]),
+                             compress_type=zipfile.ZIP_DEFLATED,
+                             compresslevel=8)
+            logger.debug(f"Added tag '{tag.get('title')}' ({uuid}) to backup")
 
         # Add any data in the watch data directory.
         for uuid, w in watches.items():
-            for f in Path(w.watch_data_dir).glob('*'):
+            for f in Path(w.data_dir).glob('*'):
                 zipObj.write(f,
                              # Use the full path to access the file, but make the file 'relative' in the Zip.
                              arcname=os.path.join(f.parts[-2], f.parts[-1]),
@@ -76,7 +92,10 @@ def create_backup(datastore_path, watches: dict):
 
 
 def construct_blueprint(datastore: ChangeDetectionStore):
+    from .restore import construct_restore_blueprint
+
     backups_blueprint = Blueprint('backups', __name__, template_folder="templates")
+    backups_blueprint.register_blueprint(construct_restore_blueprint(datastore))
     backup_threads = []
 
     @login_optionally_required
@@ -84,20 +103,25 @@ def construct_blueprint(datastore: ChangeDetectionStore):
     def request_backup():
         if any(thread.is_alive() for thread in backup_threads):
             flash(gettext("A backup is already running, check back in a few minutes"), "error")
-            return redirect(url_for('backups.index'))
+            return redirect(url_for('backups.create'))
 
         if len(find_backups()) > int(os.getenv("MAX_NUMBER_BACKUPS", 100)):
             flash(gettext("Maximum number of backups reached, please remove some"), "error")
-            return redirect(url_for('backups.index'))
+            return redirect(url_for('backups.create'))
 
-        # Be sure we're written fresh
-        datastore.sync_to_json()
-        zip_thread = threading.Thread(target=create_backup, args=(datastore.datastore_path, datastore.data.get("watching")))
+        # With immediate persistence, all data is already saved
+        zip_thread = threading.Thread(
+            target=create_backup,
+            args=(datastore.datastore_path, datastore.data.get("watching")),
+            kwargs={'tags': datastore.data['settings']['application'].get('tags', {})},
+            daemon=True,
+            name="BackupCreator"
+        )
         zip_thread.start()
         backup_threads.append(zip_thread)
         flash(gettext("Backup building in background, check back in a few minutes."))
 
-        return redirect(url_for('backups.index'))
+        return redirect(url_for('backups.create'))
 
     def find_backups():
         backup_filepath = os.path.join(datastore.datastore_path, BACKUP_FILENAME_FORMAT.format("*"))
@@ -122,31 +146,34 @@ def construct_blueprint(datastore: ChangeDetectionStore):
     def download_backup(filename):
         import re
         filename = filename.strip()
-        backup_filename_regex = BACKUP_FILENAME_FORMAT.format("\d+")
-
-        full_path = os.path.join(os.path.abspath(datastore.datastore_path), filename)
-        if not full_path.startswith(os.path.abspath(datastore.datastore_path)):
-            abort(404)
+        backup_filename_regex = BACKUP_FILENAME_FORMAT.format(r"\d+")
 
+        # Resolve 'latest' before any validation so checks run against the real filename.
         if filename == 'latest':
             backups = find_backups()
+            if not backups:
+                abort(404)
             filename = backups[0]['filename']
 
         if not re.match(r"^" + backup_filename_regex + "$", filename):
             abort(400)  # Bad Request if the filename doesn't match the pattern
 
+        full_path = os.path.join(os.path.abspath(datastore.datastore_path), filename)
+        if not full_path.startswith(os.path.abspath(datastore.datastore_path) + os.sep):
+            abort(404)
+
         logger.debug(f"Backup download request for '{full_path}'")
         return send_from_directory(os.path.abspath(datastore.datastore_path), filename, as_attachment=True)
 
     @login_optionally_required
-    @backups_blueprint.route("", methods=['GET'])
-    def index():
+    @backups_blueprint.route("/", methods=['GET'])
+    @backups_blueprint.route("/create", methods=['GET'])
+    def create():
         backups = find_backups()
-        output = render_template("overview.html",
+        output = render_template("backup_create.html",
                                  available_backups=backups,
                                  backup_running=any(thread.is_alive() for thread in backup_threads)
                                  )
-
         return output
 
     @login_optionally_required
@@ -160,6 +187,6 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
         flash(gettext("Backups were deleted."))
 
-        return redirect(url_for('backups.index'))
+        return redirect(url_for('backups.create'))
 
     return backups_blueprint

changedetectionio/blueprint/backups/restore.py

@@ -0,0 +1,248 @@
+import io
+import json
+import os
+import re
+import shutil
+import tempfile
+import threading
+import zipfile
+
+from flask import Blueprint, render_template, flash, url_for, redirect, request
+from flask_babel import gettext, lazy_gettext as _l
+from wtforms import Form, BooleanField, SubmitField
+from flask_wtf.file import FileField, FileAllowed
+from loguru import logger
+
+from changedetectionio.flask_app import login_optionally_required
+
+# Maximum size of the uploaded zip file. Override via env var MAX_RESTORE_UPLOAD_MB.
+_MAX_UPLOAD_BYTES = int(os.getenv("MAX_RESTORE_UPLOAD_MB", 256)) * 1024 * 1024
+# Maximum total uncompressed size of all entries (zip-bomb guard). Override via MAX_RESTORE_DECOMPRESSED_MB.
+_MAX_DECOMPRESSED_BYTES = int(os.getenv("MAX_RESTORE_DECOMPRESSED_MB", 1024)) * 1024 * 1024
+# Only top-level directories whose name is a valid UUID are treated as watch/tag entries.
+_UUID_RE = re.compile(
+    r'^[0-9a-f]{8}-[0-9a-f]{4}-4[0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$',
+    re.IGNORECASE,
+)
+
+
+class RestoreForm(Form):
+    zip_file = FileField(_l('Backup zip file'), validators=[
+        FileAllowed(['zip'], _l('Must be a .zip backup file!'))
+    ])
+    include_groups = BooleanField(_l('Include groups'), default=True)
+    include_groups_replace_existing = BooleanField(_l('Replace existing groups of the same UUID'), default=True)
+    include_watches = BooleanField(_l('Include watches'), default=True)
+    include_watches_replace_existing = BooleanField(_l('Replace existing watches of the same UUID'), default=True)
+    submit = SubmitField(_l('Restore backup'))
+
+
+def import_from_zip(zip_stream, datastore, include_groups, include_groups_replace, include_watches, include_watches_replace):
+    """
+    Extract and import watches and groups from a backup zip stream.
+
+    Mirrors the store's _load_watches / _load_tags loading pattern:
+      - UUID dirs with tag.json  → Tag.model + tag_obj.commit()
+      - UUID dirs with watch.json → rehydrate_entity + watch_obj.commit()
+
+    Returns a dict with counts: restored_groups, skipped_groups, restored_watches, skipped_watches.
+    Raises zipfile.BadZipFile if the stream is not a valid zip.
+    """
+    from changedetectionio.model import Tag
+
+    restored_groups = 0
+    skipped_groups = 0
+    restored_watches = 0
+    skipped_watches = 0
+
+    current_tags = datastore.data['settings']['application'].get('tags', {})
+    current_watches = datastore.data['watching']
+
+    with tempfile.TemporaryDirectory() as tmpdir:
+        logger.debug(f"Restore: extracting zip to {tmpdir}")
+        with zipfile.ZipFile(zip_stream, 'r') as zf:
+            total_uncompressed = sum(m.file_size for m in zf.infolist())
+            if total_uncompressed > _MAX_DECOMPRESSED_BYTES:
+                raise ValueError(
+                    f"Backup archive decompressed size ({total_uncompressed // (1024 * 1024)} MB) "
+                    f"exceeds the {_MAX_DECOMPRESSED_BYTES // (1024 * 1024)} MB limit"
+                )
+            resolved_dest = os.path.realpath(tmpdir)
+            for member in zf.infolist():
+                member_dest = os.path.realpath(os.path.join(resolved_dest, member.filename))
+                if not member_dest.startswith(resolved_dest + os.sep) and member_dest != resolved_dest:
+                    raise ValueError(f"Zip Slip path traversal detected in backup archive: {member.filename!r}")
+                zf.extract(member, tmpdir)
+        logger.debug("Restore: zip extracted, scanning UUID directories")
+
+        for entry in os.scandir(tmpdir):
+            if not entry.is_dir():
+                continue
+
+            uuid = entry.name
+            if not _UUID_RE.match(uuid):
+                logger.warning(f"Restore: skipping non-UUID directory {uuid!r}")
+                continue
+            tag_json_path = os.path.join(entry.path, 'tag.json')
+            watch_json_path = os.path.join(entry.path, 'watch.json')
+
+            # --- Tags (groups) ---
+            if include_groups and os.path.exists(tag_json_path):
+                if uuid in current_tags and not include_groups_replace:
+                    logger.debug(f"Restore: skipping existing group {uuid} (replace not requested)")
+                    skipped_groups += 1
+                    continue
+
+                try:
+                    with open(tag_json_path, 'r', encoding='utf-8') as f:
+                        tag_data = json.load(f)
+                except (json.JSONDecodeError, IOError) as e:
+                    logger.error(f"Restore: failed to read tag.json for {uuid}: {e}")
+                    continue
+
+                title = tag_data.get('title', uuid)
+                logger.debug(f"Restore: importing group '{title}' ({uuid})")
+
+                # Mirror _load_tags: set uuid and force processor
+                tag_data['uuid'] = uuid
+                tag_data['processor'] = 'restock_diff'
+
+                # Copy the UUID directory so data_dir exists for commit()
+                dst_dir = os.path.join(datastore.datastore_path, uuid)
+                if os.path.exists(dst_dir):
+                    shutil.rmtree(dst_dir)
+                shutil.copytree(entry.path, dst_dir)
+
+                tag_obj = Tag.model(
+                    datastore_path=datastore.datastore_path,
+                    __datastore=datastore.data,
+                    default=tag_data
+                )
+                current_tags[uuid] = tag_obj
+                tag_obj.commit()
+                restored_groups += 1
+                logger.success(f"Restore: group '{title}' ({uuid}) restored")
+
+            # --- Watches ---
+            elif include_watches and os.path.exists(watch_json_path):
+                if uuid in current_watches and not include_watches_replace:
+                    logger.debug(f"Restore: skipping existing watch {uuid} (replace not requested)")
+                    skipped_watches += 1
+                    continue
+
+                try:
+                    with open(watch_json_path, 'r', encoding='utf-8') as f:
+                        watch_data = json.load(f)
+                except (json.JSONDecodeError, IOError) as e:
+                    logger.error(f"Restore: failed to read watch.json for {uuid}: {e}")
+                    continue
+
+                url = watch_data.get('url', uuid)
+                logger.debug(f"Restore: importing watch '{url}' ({uuid})")
+
+                # Copy UUID directory first so data_dir and history files exist
+                dst_dir = os.path.join(datastore.datastore_path, uuid)
+                if os.path.exists(dst_dir):
+                    shutil.rmtree(dst_dir)
+                shutil.copytree(entry.path, dst_dir)
+
+                # Mirror _load_watches / rehydrate_entity
+                watch_data['uuid'] = uuid
+                watch_obj = datastore.rehydrate_entity(uuid, watch_data)
+                current_watches[uuid] = watch_obj
+                watch_obj.commit()
+                restored_watches += 1
+                logger.success(f"Restore: watch '{url}' ({uuid}) restored")
+
+        logger.debug(f"Restore: scan complete - groups {restored_groups} restored / {skipped_groups} skipped, "
+                     f"watches {restored_watches} restored / {skipped_watches} skipped")
+
+    # Persist changedetection.json (includes the updated tags dict)
+    logger.debug("Restore: committing datastore settings")
+    datastore.commit()
+
+    return {
+        'restored_groups': restored_groups,
+        'skipped_groups': skipped_groups,
+        'restored_watches': restored_watches,
+        'skipped_watches': skipped_watches,
+    }
+
+
+
+def construct_restore_blueprint(datastore):
+    restore_blueprint = Blueprint('restore', __name__, template_folder="templates")
+    restore_threads = []
+
+    @login_optionally_required
+    @restore_blueprint.route("/restore", methods=['GET'])
+    def restore():
+        form = RestoreForm()
+        return render_template("backup_restore.html",
+                               form=form,
+                               restore_running=any(t.is_alive() for t in restore_threads),
+                               max_upload_mb=_MAX_UPLOAD_BYTES // (1024 * 1024),
+                               max_decompressed_mb=_MAX_DECOMPRESSED_BYTES // (1024 * 1024))
+
+    @login_optionally_required
+    @restore_blueprint.route("/restore/start", methods=['POST'])
+    def backups_restore_start():
+        if any(t.is_alive() for t in restore_threads):
+            flash(gettext("A restore is already running, check back in a few minutes"), "error")
+            return redirect(url_for('backups.restore.restore'))
+
+        zip_file = request.files.get('zip_file')
+        if not zip_file or not zip_file.filename:
+            flash(gettext("No file uploaded"), "error")
+            return redirect(url_for('backups.restore.restore'))
+
+        if not zip_file.filename.lower().endswith('.zip'):
+            flash(gettext("File must be a .zip backup file"), "error")
+            return redirect(url_for('backups.restore.restore'))
+
+        # Reject oversized uploads before reading the stream into memory.
+        content_length = request.content_length
+        if content_length and content_length > _MAX_UPLOAD_BYTES:
+            flash(gettext("Backup file is too large (max %(mb)s MB)", mb=_MAX_UPLOAD_BYTES // (1024 * 1024)), "error")
+            return redirect(url_for('backups.restore.restore'))
+
+        # Read into memory now — the request stream is gone once we return.
+        # Read one byte beyond the limit so we can detect truncated-but-still-oversized streams.
+        try:
+            raw = zip_file.read(_MAX_UPLOAD_BYTES + 1)
+            if len(raw) > _MAX_UPLOAD_BYTES:
+                flash(gettext("Backup file is too large (max %(mb)s MB)", mb=_MAX_UPLOAD_BYTES // (1024 * 1024)), "error")
+                return redirect(url_for('backups.restore.restore'))
+            zip_bytes = io.BytesIO(raw)
+            with zipfile.ZipFile(zip_bytes):  # quick validity check before spawning
+                pass
+            zip_bytes.seek(0)
+        except zipfile.BadZipFile:
+            flash(gettext("Invalid or corrupted zip file"), "error")
+            return redirect(url_for('backups.restore.restore'))
+
+        include_groups = request.form.get('include_groups') == 'y'
+        include_groups_replace = request.form.get('include_groups_replace_existing') == 'y'
+        include_watches = request.form.get('include_watches') == 'y'
+        include_watches_replace = request.form.get('include_watches_replace_existing') == 'y'
+
+        restore_thread = threading.Thread(
+            target=import_from_zip,
+            kwargs={
+                'zip_stream': zip_bytes,
+                'datastore': datastore,
+                'include_groups': include_groups,
+                'include_groups_replace': include_groups_replace,
+                'include_watches': include_watches,
+                'include_watches_replace': include_watches_replace,
+            },
+            daemon=True,
+            name="BackupRestore"
+        )
+        restore_thread.start()
+        restore_threads[:] = [t for t in restore_threads if t.is_alive()]
+        restore_threads.append(restore_thread)
+        flash(gettext("Restore started in background, check back in a few minutes."))
+        return redirect(url_for('backups.restore.restore'))
+
+    return restore_blueprint

changedetectionio/blueprint/backups/templates/backup_create.html

@@ -0,0 +1,49 @@
+{% extends 'base.html' %}
+{% block content %}
+    {% from '_helpers.html' import render_simple_field, render_field %}
+
+    <div class="edit-form">
+        <div class="tabs collapsable">
+            <ul>
+                <li class="tab active" id=""><a href="{{ url_for('backups.create') }}">{{ _('Create') }}</a></li>
+                <li class="tab"><a href="{{ url_for('backups.restore.restore') }}">{{ _('Restore') }}</a></li>
+            </ul>
+        </div>
+        <div class="box-wrap inner">
+            <div id="general">
+                {% if backup_running %}
+                    <p>
+                        <span class="spinner"></span>&nbsp;<strong>{{ _('A backup is running!') }}</strong>
+                    </p>
+                {% endif %}
+
+                <p>
+                    {{ _('Here you can download and request a new backup, when a backup is completed you will see it listed below.') }}
+                </p>
+                <br>
+                {% if available_backups %}
+                    <ul>
+                        {% for backup in available_backups %}
+                            <li>
+                                <a href="{{ url_for('backups.download_backup', filename=backup["filename"]) }}">{{ backup["filename"] }}</a> {{ backup["filesize"] }} {{ _('Mb') }}
+                            </li>
+                        {% endfor %}
+                    </ul>
+                {% else %}
+                    <p>
+                        <strong>{{ _('No backups found.') }}</strong>
+                    </p>
+                {% endif %}
+
+                <a class="pure-button pure-button-primary"
+                   href="{{ url_for('backups.request_backup') }}">{{ _('Create backup') }}</a>
+                {% if available_backups %}
+                    <a class="pure-button button-small button-error "
+                       href="{{ url_for('backups.remove_backups') }}">{{ _('Remove backups') }}</a>
+                {% endif %}
+
+            </div>
+
+        </div>
+    </div>
+{% endblock %}

changedetectionio/blueprint/backups/templates/backup_restore.html

@@ -0,0 +1,62 @@
+{% extends 'base.html' %}
+{% block content %}
+    {% from '_helpers.html' import render_field, render_checkbox_field %}
+
+    <div class="edit-form">
+        <div class="tabs collapsable">
+            <ul>
+                <li class="tab"><a href="{{ url_for('backups.create') }}">{{ _('Create') }}</a></li>
+                <li class="tab active"><a href="{{ url_for('backups.restore.restore') }}">{{ _('Restore') }}</a></li>
+            </ul>
+        </div>
+        <div class="box-wrap inner">
+            <div id="general">
+                {% if restore_running %}
+                    <p>
+                        <span class="spinner"></span>&nbsp;<strong>{{ _('A restore is running!') }}</strong>
+                    </p>
+                {% endif %}
+
+                <p>{{ _('Restore a backup. Must be a .zip backup file created on/after v0.53.1 (new database layout).') }}</p>
+                <p>{{ _('Note: This does not override the main application settings, only watches and groups.') }}</p>
+                <p class="pure-form-message">
+                    {{ _('Max upload size: %(upload)s MB &nbsp;·&nbsp; Max decompressed size: %(decomp)s MB',
+                         upload=max_upload_mb, decomp=max_decompressed_mb) }}
+                </p>
+
+                <form class="pure-form pure-form-stacked settings"
+                      action="{{ url_for('backups.restore.backups_restore_start') }}"
+                      method="POST"
+                      enctype="multipart/form-data">
+                    <input type="hidden" name="csrf_token" value="{{ csrf_token() }}">
+
+                    <div class="pure-control-group">
+                        {{ render_checkbox_field(form.include_groups) }}
+                        <span class="pure-form-message-inline">{{ _('Include all groups found in backup?') }}</span>
+                    </div>
+                    <div class="pure-control-group">
+                        {{ render_checkbox_field(form.include_groups_replace_existing) }}
+                        <span class="pure-form-message-inline">{{ _('Replace any existing groups of the same UUID?') }}</span>
+                    </div>
+
+                    <div class="pure-control-group">
+                        {{ render_checkbox_field(form.include_watches) }}
+                        <span class="pure-form-message-inline">{{ _('Include all watches found in backup?') }}</span>
+                    </div>
+                    <div class="pure-control-group">
+                        {{ render_checkbox_field(form.include_watches_replace_existing) }}
+                        <span class="pure-form-message-inline">{{ _('Replace any existing watches of the same UUID?') }}</span>
+                    </div>
+
+                    <div class="pure-control-group">
+                        {{ render_field(form.zip_file) }}
+                    </div>
+
+                    <div class="pure-controls">
+                        <button type="submit" class="pure-button pure-button-primary">{{ _('Restore backup') }}</button>
+                    </div>
+                </form>
+            </div>
+        </div>
+    </div>
+{% endblock %}

changedetectionio/blueprint/backups/templates/overview.html

@@ -1,36 +0,0 @@
-{% extends 'base.html' %}
-{% block content %}
-    {% from '_helpers.html' import render_simple_field, render_field %}
-    <div class="edit-form">
-        <div class="box-wrap inner">
-            <h4>{{ _('Backups') }}</h4>
-            {% if backup_running %}
-                <p>
-                    <strong>{{ _('A backup is running!') }}</strong>
-                </p>
-            {% endif %}
-            <p>
-                {{ _('Here you can download and request a new backup, when a backup is completed you will see it listed below.') }}
-            </p>
-            <br>
-                {% if available_backups %}
-                    <ul>
-                    {% for backup in available_backups %}
-                        <li><a href="{{ url_for('backups.download_backup', filename=backup["filename"]) }}">{{ backup["filename"] }}</a> {{  backup["filesize"] }} {{ _('Mb') }}</li>
-                    {% endfor %}
-                    </ul>
-                {% else %}
-                    <p>
-                    <strong>{{ _('No backups found.') }}</strong>
-                    </p>
-                {% endif %}
-
-            <a class="pure-button pure-button-primary" href="{{ url_for('backups.request_backup') }}">{{ _('Create backup') }}</a>
-            {% if available_backups %}
-                <a class="pure-button button-small button-error " href="{{ url_for('backups.remove_backups') }}">{{ _('Remove backups') }}</a>
-            {% endif %}
-        </div>
-    </div>
-
-
-{% endblock %}

changedetectionio/blueprint/browser_steps/__init__.py

@@ -21,37 +21,160 @@ from changedetectionio.flask_app import login_optionally_required
 from loguru import logger
 
 browsersteps_sessions = {}
+browsersteps_watch_to_session = {}  # Maps watch_uuid -> browsersteps_session_id
 io_interface_context = None
 import json
 import hashlib
 from flask import Response
 import asyncio
 import threading
+import time
 
-def run_async_in_browser_loop(coro):
-    """Run async coroutine using the existing async worker event loop"""
-    from changedetectionio import worker_handler
-    
-    # Use the existing async worker event loop instead of creating a new one
-    if worker_handler.USE_ASYNC_WORKERS and worker_handler.async_loop and not worker_handler.async_loop.is_closed():
-        logger.debug("Browser steps using existing async worker event loop")
-        future = asyncio.run_coroutine_threadsafe(coro, worker_handler.async_loop)
-        return future.result()
-    else:
-        # Fallback: create a new event loop (for sync workers or if async loop not available)
-        logger.debug("Browser steps creating temporary event loop")
-        loop = asyncio.new_event_loop()
-        asyncio.set_event_loop(loop)
+# Dedicated event loop for ALL browser steps sessions
+_browser_steps_loop = None
+_browser_steps_thread = None
+_browser_steps_loop_lock = threading.Lock()
+
+def _start_browser_steps_loop():
+    """Start a dedicated event loop for browser steps in its own thread"""
+    global _browser_steps_loop
+
+    # Create and set the event loop for this thread
+    loop = asyncio.new_event_loop()
+    asyncio.set_event_loop(loop)
+    _browser_steps_loop = loop
+
+    logger.debug("Browser steps event loop started")
+
+    try:
+        # Run the loop forever - handles all browsersteps sessions
+        loop.run_forever()
+    except Exception as e:
+        logger.error(f"Browser steps event loop error: {e}")
+    finally:
         try:
-            return loop.run_until_complete(coro)
+            # Cancel all remaining tasks
+            pending = asyncio.all_tasks(loop)
+            for task in pending:
+                task.cancel()
+
+            # Wait for tasks to finish cancellation
+            if pending:
+                loop.run_until_complete(asyncio.gather(*pending, return_exceptions=True))
+        except Exception as e:
+            logger.debug(f"Error during browser steps loop cleanup: {e}")
         finally:
             loop.close()
+            logger.debug("Browser steps event loop closed")
+
+def _ensure_browser_steps_loop():
+    """Ensure the browser steps event loop is running"""
+    global _browser_steps_loop, _browser_steps_thread
+
+    with _browser_steps_loop_lock:
+        if _browser_steps_thread is None or not _browser_steps_thread.is_alive():
+            logger.debug("Starting browser steps event loop thread")
+            _browser_steps_thread = threading.Thread(
+                target=_start_browser_steps_loop,
+                daemon=True,
+                name="BrowserStepsEventLoop"
+            )
+            _browser_steps_thread.start()
+
+            # Wait for the loop to be ready
+            timeout = 5.0
+            start_time = time.time()
+            while _browser_steps_loop is None:
+                if time.time() - start_time > timeout:
+                    raise RuntimeError("Browser steps event loop failed to start")
+                time.sleep(0.01)
+
+            logger.debug("Browser steps event loop thread started and ready")
+
+def run_async_in_browser_loop(coro):
+    """Run async coroutine using the dedicated browser steps event loop"""
+    _ensure_browser_steps_loop()
+
+    if _browser_steps_loop and not _browser_steps_loop.is_closed():
+        logger.debug("Browser steps using dedicated event loop")
+        future = asyncio.run_coroutine_threadsafe(coro, _browser_steps_loop)
+        return future.result()
+    else:
+        raise RuntimeError("Browser steps event loop is not available")
+
+def cleanup_expired_sessions():
+    """Remove expired browsersteps sessions and cleanup their resources"""
+    global browsersteps_sessions, browsersteps_watch_to_session
+
+    expired_session_ids = []
+
+    # Find expired sessions
+    for session_id, session_data in browsersteps_sessions.items():
+        browserstepper = session_data.get('browserstepper')
+        if browserstepper and browserstepper.has_expired:
+            expired_session_ids.append(session_id)
+
+    # Cleanup expired sessions
+    for session_id in expired_session_ids:
+        logger.debug(f"Cleaning up expired browsersteps session {session_id}")
+        session_data = browsersteps_sessions[session_id]
+
+        # Cleanup playwright resources asynchronously
+        browserstepper = session_data.get('browserstepper')
+        if browserstepper:
+            try:
+                run_async_in_browser_loop(browserstepper.cleanup())
+            except Exception as e:
+                logger.error(f"Error cleaning up session {session_id}: {e}")
+
+        # Remove from sessions dict
+        del browsersteps_sessions[session_id]
+
+        # Remove from watch mapping
+        for watch_uuid, mapped_session_id in list(browsersteps_watch_to_session.items()):
+            if mapped_session_id == session_id:
+                del browsersteps_watch_to_session[watch_uuid]
+                break
+
+    if expired_session_ids:
+        logger.info(f"Cleaned up {len(expired_session_ids)} expired browsersteps session(s)")
+
+def cleanup_session_for_watch(watch_uuid):
+    """Cleanup a specific browsersteps session for a watch UUID"""
+    global browsersteps_sessions, browsersteps_watch_to_session
+
+    session_id = browsersteps_watch_to_session.get(watch_uuid)
+    if not session_id:
+        logger.debug(f"No browsersteps session found for watch {watch_uuid}")
+        return
+
+    logger.debug(f"Cleaning up browsersteps session {session_id} for watch {watch_uuid}")
+
+    session_data = browsersteps_sessions.get(session_id)
+    if session_data:
+        browserstepper = session_data.get('browserstepper')
+        if browserstepper:
+            try:
+                run_async_in_browser_loop(browserstepper.cleanup())
+            except Exception as e:
+                logger.error(f"Error cleaning up session {session_id} for watch {watch_uuid}: {e}")
+
+        # Remove from sessions dict
+        del browsersteps_sessions[session_id]
+
+    # Remove from watch mapping
+    del browsersteps_watch_to_session[watch_uuid]
+
+    logger.debug(f"Cleaned up session for watch {watch_uuid}")
+
+    # Opportunistically cleanup any other expired sessions
+    cleanup_expired_sessions()
 
 def construct_blueprint(datastore: ChangeDetectionStore):
     browser_steps_blueprint = Blueprint('browser_steps', __name__, template_folder="templates")
 
     async def start_browsersteps_session(watch_uuid):
-        from . import browser_steps
+        from changedetectionio.browser_steps import browser_steps
         import time
         from playwright.async_api import async_playwright
 
@@ -115,7 +238,6 @@ def construct_blueprint(datastore: ChangeDetectionStore):
     @browser_steps_blueprint.route("/browsersteps_start_session", methods=['GET'])
     def browsersteps_start_session():
         # A new session was requested, return sessionID
-        import asyncio
         import uuid
         browsersteps_session_id = str(uuid.uuid4())
         watch_uuid = request.args.get('uuid')
@@ -123,6 +245,9 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         if not watch_uuid:
             return make_response('No Watch UUID specified', 500)
 
+        # Cleanup any existing session for this watch
+        cleanup_session_for_watch(watch_uuid)
+
         logger.debug("Starting connection with playwright")
         logger.debug("browser_steps.py connecting")
 
@@ -131,6 +256,10 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             browsersteps_sessions[browsersteps_session_id] = run_async_in_browser_loop(
                 start_browsersteps_session(watch_uuid)
             )
+
+            # Store the mapping of watch_uuid -> browsersteps_session_id
+            browsersteps_watch_to_session[watch_uuid] = browsersteps_session_id
+
         except Exception as e:
             if 'ECONNREFUSED' in str(e):
                 return make_response('Unable to start the Playwright Browser session, is sockpuppetbrowser running? Network configuration is OK?', 401)
@@ -155,8 +284,8 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         watch = datastore.data['watching'].get(uuid)
         filename = f"step_before-{step_n}.jpeg" if request.args.get('type', '') == 'before' else f"step_{step_n}.jpeg"
 
-        if step_n and watch and os.path.isfile(os.path.join(watch.watch_data_dir, filename)):
-            response = make_response(send_from_directory(directory=watch.watch_data_dir, path=filename))
+        if step_n and watch and os.path.isfile(os.path.join(watch.data_dir, filename)):
+            response = make_response(send_from_directory(directory=watch.data_dir, path=filename))
             response.headers['Content-type'] = 'image/jpeg'
             response.headers['Cache-Control'] = 'no-cache, no-store, must-revalidate'
             response.headers['Pragma'] = 'no-cache'
@@ -171,11 +300,10 @@ def construct_blueprint(datastore: ChangeDetectionStore):
     @browser_steps_blueprint.route("/browsersteps_update", methods=['POST'])
     def browsersteps_ui_update():
         import base64
-        import playwright._impl._errors
-        from changedetectionio.blueprint.browser_steps import browser_steps
 
-        remaining =0
+        remaining = 0
         uuid = request.args.get('uuid')
+        goto_website_url_first_step = request.args.get('goto_website_url_first_step')
 
         browsersteps_session_id = request.args.get('browsersteps_session_id')
 
@@ -186,33 +314,33 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             return make_response('No session exists under that ID', 500)
 
         is_last_step = False
-        # Actions - step/apply/etc, do the thing and return state
-        if request.method == 'POST':
-            # @todo - should always be an existing session
+
+        # @todo - should always be an existing session
+        if goto_website_url_first_step:
+            logger.debug("Going to site (requested automatically before stepping)..")
+            step_operation = "Goto site"
+            step_selector = None
+            step_optional_value = None
+        else:
             step_operation = request.form.get('operation')
             step_selector = request.form.get('selector')
             step_optional_value = request.form.get('optional_value')
             is_last_step = strtobool(request.form.get('is_last_step'))
 
-            try:
-                # Run the async call_action method in the dedicated browser steps event loop
-                run_async_in_browser_loop(
-                    browsersteps_sessions[browsersteps_session_id]['browserstepper'].call_action(
-                        action_name=step_operation,
-                        selector=step_selector,
-                        optional_value=step_optional_value
-                    )
+        try:
+            # Run the async call_action method in the dedicated browser steps event loop
+            run_async_in_browser_loop(
+                browsersteps_sessions[browsersteps_session_id]['browserstepper'].call_action(
+                    action_name=step_operation,
+                    selector=step_selector,
+                    optional_value=step_optional_value
                 )
+            )
 
-            except Exception as e:
-                logger.error(f"Exception when calling step operation {step_operation} {str(e)}")
-                # Try to find something of value to give back to the user
-                return make_response(str(e).splitlines()[0], 401)
-
-
-#        if not this_session.page:
-#            cleanup_playwright_session()
-#            return make_response('Browser session ran out of time :( Please reload this page.', 401)
+        except Exception as e:
+            logger.error(f"Exception when calling step operation {step_operation} {str(e)}")
+            # Try to find something of value to give back to the user
+            return make_response(str(e).splitlines()[0], 401)
 
         # Screenshots and other info only needed on requesting a step (POST)
         try:
@@ -220,7 +348,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             (screenshot, xpath_data) = run_async_in_browser_loop(
                 browsersteps_sessions[browsersteps_session_id]['browserstepper'].get_current_state()
             )
-                
+
             if is_last_step:
                 watch = datastore.data['watching'].get(uuid)
                 u = browsersteps_sessions[browsersteps_session_id]['browserstepper'].page.url

changedetectionio/blueprint/check_proxies/__init__.py

@@ -94,13 +94,13 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         return results
 
     @login_required
-    @check_proxies_blueprint.route("/<string:uuid>/status", methods=['GET'])
+    @check_proxies_blueprint.route("/<uuid_str:uuid>/status", methods=['GET'])
     def get_recheck_status(uuid):
         results = _recalc_check_status(uuid=uuid)
         return results
 
     @login_required
-    @check_proxies_blueprint.route("/<string:uuid>/start", methods=['GET'])
+    @check_proxies_blueprint.route("/<uuid_str:uuid>/start", methods=['GET'])
     def start_check(uuid):
 
         if not datastore.proxy_list:

changedetectionio/blueprint/imports/__init__.py

@@ -1,13 +1,8 @@
 from flask import Blueprint, request, redirect, url_for, flash, render_template
+from loguru import logger
+
 from changedetectionio.store import ChangeDetectionStore
 from changedetectionio.auth_decorator import login_optionally_required
-from changedetectionio import worker_handler
-from changedetectionio.blueprint.imports.importer import (
-    import_url_list, 
-    import_distill_io_json, 
-    import_xlsx_wachete, 
-    import_xlsx_custom
-)
 
 def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMetaData):
     import_blueprint = Blueprint('imports', __name__, template_folder="templates")
@@ -17,15 +12,27 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
     def import_page():
         remaining_urls = []
         from changedetectionio import forms
-
+#
         if request.method == 'POST':
+#            from changedetectionio import worker_pool
+
+            from changedetectionio.blueprint.imports.importer import (
+                import_url_list,
+                import_distill_io_json,
+                import_xlsx_wachete,
+                import_xlsx_custom
+            )
+
             # URL List import
             if request.values.get('urls') and len(request.values.get('urls').strip()):
                 # Import and push into the queue for immediate update check
+                from changedetectionio import processors
                 importer_handler = import_url_list()
-                importer_handler.run(data=request.values.get('urls'), flash=flash, datastore=datastore, processor=request.values.get('processor', 'text_json_diff'))
-                for uuid in importer_handler.new_uuids:
-                    worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                importer_handler.run(data=request.values.get('urls'), flash=flash, datastore=datastore, processor=request.values.get('processor', processors.get_default_processor()))
+                logger.debug(f"Imported {len(importer_handler.new_uuids)} new UUIDs")
+                # Dont' add to queue because scheduler can see that they haven't been checked and will add them to the queue
+#                for uuid in importer_handler.new_uuids:
+#                    worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
 
                 if len(importer_handler.remaining_data) == 0:
                     return redirect(url_for('watchlist.index'))
@@ -37,8 +44,10 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                 # Import and push into the queue for immediate update check
                 d_importer = import_distill_io_json()
                 d_importer.run(data=request.values.get('distill-io'), flash=flash, datastore=datastore)
-                for uuid in d_importer.new_uuids:
-                    worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                # Dont' add to queue because scheduler can see that they haven't been checked and will add them to the queue
+#                for uuid in importer_handler.new_uuids:
+#                    worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+
 
             # XLSX importer
             if request.files and request.files.get('xlsx_file'):
@@ -60,8 +69,10 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                     w_importer.import_profile = map
                     w_importer.run(data=file, flash=flash, datastore=datastore)
 
-                for uuid in w_importer.new_uuids:
-                    worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                # Dont' add to queue because scheduler can see that they haven't been checked and will add them to the queue
+#                for uuid in importer_handler.new_uuids:
+#                    worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+
 
         # Could be some remaining, or we could be on GET
         form = forms.importForm(formdata=request.form if request.method == 'POST' else None)

changedetectionio/blueprint/imports/importer.py

@@ -62,7 +62,7 @@ class import_url_list(Importer):
                 extras = None
                 if processor:
                     extras = {'processor': processor}
-                new_uuid = datastore.add_watch(url=url.strip(), tag=tags, write_to_disk_now=False, extras=extras)
+                new_uuid = datastore.add_watch(url=url.strip(), tag=tags, save_immediately=False, extras=extras)
 
                 if new_uuid:
                     # Straight into the queue.
@@ -129,7 +129,7 @@ class import_distill_io_json(Importer):
                 new_uuid = datastore.add_watch(url=d['uri'].strip(),
                                                tag=",".join(d.get('tags', [])),
                                                extras=extras,
-                                               write_to_disk_now=False)
+                                               save_immediately=False)
 
                 if new_uuid:
                     # Straight into the queue.
@@ -204,7 +204,7 @@ class import_xlsx_wachete(Importer):
                     new_uuid = datastore.add_watch(url=data['url'].strip(),
                                                    extras=extras,
                                                    tag=data.get('folder'),
-                                                   write_to_disk_now=False)
+                                                   save_immediately=False)
                     if new_uuid:
                         # Straight into the queue.
                         self.new_uuids.append(new_uuid)
@@ -287,7 +287,7 @@ class import_xlsx_custom(Importer):
                     new_uuid = datastore.add_watch(url=url,
                                                    extras=extras,
                                                    tag=tags,
-                                                   write_to_disk_now=False)
+                                                   save_immediately=False)
                     if new_uuid:
                         # Straight into the queue.
                         self.new_uuids.append(new_uuid)

changedetectionio/blueprint/imports/templates/import.html

@@ -16,6 +16,11 @@
         <form class="pure-form" action="{{url_for('imports.import_page')}}" method="POST" enctype="multipart/form-data">
             <input type="hidden" name="csrf_token" value="{{ csrf_token() }}">
             <div class="tab-pane-inner" id="url-list">
+
+                <p>
+                {{ _('Restoring changedetection.io backups is in the') }}<a href="{{ url_for('backups.restore.restore') }}"> {{ _('backups section') }}</a>.
+                <br>
+                </p>
                 <div class="pure-control-group">
                         {{ _('Enter one URL per line, and optionally add tags for each URL after a space, delineated by comma (,):') }}
                         <br>
@@ -37,9 +42,6 @@
             </div>
 
             <div class="tab-pane-inner" id="distill-io">
-
-
-
                     <div class="pure-control-group">
                         {{ _('Copy and Paste your Distill.io watch \'export\' file, this should be a JSON file.') }}<br>
                         {{ _('This is') }} <i>{{ _('experimental') }}</i>, {{ _('supported fields are') }} <code>name</code>, <code>uri</code>, <code>tags</code>, <code>config:selections</code>, {{ _('the rest (including') }} <code>schedule</code>) {{ _('are ignored.') }}
@@ -49,8 +51,6 @@
                         {{ _('Be sure to set your default fetcher to Chrome if required.') }}<br>
                         </p>
                     </div>
-
-
                     <textarea name="distill-io" class="pure-input-1-2" style="width: 100%;
                                 font-family:monospace;
                                 white-space: pre;
@@ -114,6 +114,7 @@
                 </div>
             </div>
             <button type="submit" class="pure-button pure-input-1-2 pure-button-primary">{{ _('Import') }}</button>
+
         </form>
 
     </div>

changedetectionio/blueprint/price_data_follower/__init__.py

@@ -4,7 +4,7 @@ from flask import Blueprint, flash, redirect, url_for
 from flask_login import login_required
 from changedetectionio.store import ChangeDetectionStore
 from changedetectionio import queuedWatchMetaData
-from changedetectionio import worker_handler
+from changedetectionio import worker_pool
 from queue import PriorityQueue
 
 PRICE_DATA_TRACK_ACCEPT = 'accepted'
@@ -15,18 +15,20 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q: PriorityQueue
     price_data_follower_blueprint = Blueprint('price_data_follower', __name__)
 
     @login_required
-    @price_data_follower_blueprint.route("/<string:uuid>/accept", methods=['GET'])
+    @price_data_follower_blueprint.route("/<uuid_str:uuid>/accept", methods=['GET'])
     def accept(uuid):
         datastore.data['watching'][uuid]['track_ldjson_price_data'] = PRICE_DATA_TRACK_ACCEPT
         datastore.data['watching'][uuid]['processor'] = 'restock_diff'
         datastore.data['watching'][uuid].clear_watch()
-        worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+        datastore.data['watching'][uuid].commit()
+        worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
         return redirect(url_for("watchlist.index"))
 
     @login_required
-    @price_data_follower_blueprint.route("/<string:uuid>/reject", methods=['GET'])
+    @price_data_follower_blueprint.route("/<uuid_str:uuid>/reject", methods=['GET'])
     def reject(uuid):
         datastore.data['watching'][uuid]['track_ldjson_price_data'] = PRICE_DATA_TRACK_REJECT
+        datastore.data['watching'][uuid].commit()
         return redirect(url_for("watchlist.index"))
 
 

changedetectionio/blueprint/rss/single_watch.py

@@ -9,11 +9,12 @@ def construct_single_watch_routes(rss_blueprint, datastore):
         datastore: The ChangeDetectionStore instance
     """
 
-    @rss_blueprint.route("/watch/<string:uuid>", methods=['GET'])
+    @rss_blueprint.route("/watch/<uuid_str:uuid>", methods=['GET'])
     def rss_single_watch(uuid):
         import time
 
-        from flask import make_response, request
+        from flask import make_response, request, Response
+        from flask_babel import lazy_gettext as _l
         from feedgen.feed import FeedGenerator
         from loguru import logger
 
@@ -37,15 +38,17 @@ def construct_single_watch_routes(rss_blueprint, datastore):
 
         rss_content_format = datastore.data['settings']['application'].get('rss_content_format')
 
+        if uuid == 'first':
+            uuid = list(datastore.data['watching'].keys()).pop()
         # Get the watch by UUID
         watch = datastore.data['watching'].get(uuid)
         if not watch:
-            return f"Watch with UUID {uuid} not found", 404
+            return Response(_l("Watch with UUID %(uuid)s not found", uuid=uuid), status=404, mimetype='text/plain')
 
         # Check if watch has at least 2 history snapshots
         dates = list(watch.history.keys())
         if len(dates) < 2:
-            return f"Watch {uuid} does not have enough history snapshots to show changes (need at least 2)", 400
+            return Response(_l("Watch %(uuid)s does not have enough history snapshots to show changes (need at least 2)", uuid=uuid), status=400, mimetype='text/plain')
 
         # Get the number of diffs to include (default: 5)
         rss_diff_length = datastore.data['settings']['application'].get('rss_diff_length', 5)

changedetectionio/blueprint/rss/tag.py

@@ -7,7 +7,7 @@ def construct_tag_routes(rss_blueprint, datastore):
         datastore: The ChangeDetectionStore instance
     """
 
-    @rss_blueprint.route("/tag/<string:tag_uuid>", methods=['GET'])
+    @rss_blueprint.route("/tag/<uuid_str:tag_uuid>", methods=['GET'])
     def rss_tag_feed(tag_uuid):
 
         from flask import make_response, request, url_for

changedetectionio/blueprint/settings/__init__.py

@@ -1,8 +1,9 @@
 import os
 from copy import deepcopy
-from datetime import datetime
+from datetime import datetime, timedelta
 from zoneinfo import ZoneInfo, available_timezones
 import secrets
+import time
 import flask_login
 from flask import Blueprint, render_template, request, redirect, url_for, flash
 from flask_babel import gettext
@@ -74,26 +75,37 @@ def construct_blueprint(datastore: ChangeDetectionStore):
                     del (app_update['password'])
 
                 datastore.data['settings']['application'].update(app_update)
-                
+
                 # Handle dynamic worker count adjustment
                 old_worker_count = datastore.data['settings']['requests'].get('workers', 1)
                 new_worker_count = form.data['requests'].get('workers', 1)
-                
+
                 datastore.data['settings']['requests'].update(form.data['requests'])
-                
+                datastore.commit()
+
+                # Clear all checksums to force reprocessing with new settings
+                # Global settings can affect watch behavior (filters, rendering, etc.)
+                datastore.clear_all_last_checksums()
+
                 # Adjust worker count if it changed
                 if new_worker_count != old_worker_count:
-                    from changedetectionio import worker_handler
+                    from changedetectionio import worker_pool
                     from changedetectionio.flask_app import update_q, notification_q, app, datastore as ds
-                    
-                    result = worker_handler.adjust_async_worker_count(
+
+                    # Check CPU core availability and warn if worker count is high
+                    cpu_count = os.cpu_count()
+                    if cpu_count and new_worker_count >= (cpu_count * 0.9):
+                        flash(gettext("Warning: Worker count ({}) is close to or exceeds available CPU cores ({})").format(
+                            new_worker_count, cpu_count), 'warning')
+
+                    result = worker_pool.adjust_async_worker_count(
                         new_count=new_worker_count,
                         update_q=update_q,
                         notification_q=notification_q,
                         app=app,
                         datastore=ds
                     )
-                    
+
                     if result['status'] == 'success':
                         flash(gettext("Worker count adjusted: {}").format(result['message']), 'notice')
                     elif result['status'] == 'not_supported':
@@ -103,13 +115,11 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
                 if not os.getenv("SALTED_PASS", False) and len(form.application.form.password.encrypted_password):
                     datastore.data['settings']['application']['password'] = form.application.form.password.encrypted_password
-                    datastore.needs_write_urgent = True
+                    datastore.commit()
                     flash(gettext("Password protection enabled."), 'notice')
                     flask_login.logout_user()
                     return redirect(url_for('watchlist.index'))
 
-                datastore.needs_write_urgent = True
-
                 # Also save plugin settings from the same form submission
                 plugin_tabs_list = get_plugin_settings_tabs()
                 for tab in plugin_tabs_list:
@@ -137,6 +147,9 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         active_plugins = get_active_plugins()
         python_version = f"{sys.version_info.major}.{sys.version_info.minor}.{sys.version_info.micro}"
 
+        # Calculate uptime in seconds
+        uptime_seconds = time.time() - datastore.start_time
+
         # Get plugin settings tabs and instantiate forms
         plugin_tabs = get_plugin_settings_tabs()
         plugin_forms = {}
@@ -155,6 +168,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
                                 active_plugins=active_plugins,
                                 api_key=datastore.data['settings']['application'].get('api_access_token'),
                                 python_version=python_version,
+                                uptime_seconds=uptime_seconds,
                                 available_timezones=sorted(available_timezones()),
                                 emailprefix=os.getenv('NOTIFICATION_MAIL_BUTTON_PREFIX', False),
                                 extra_notification_token_placeholder_info=datastore.get_unique_notification_token_placeholders_available(),
@@ -175,7 +189,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
     def settings_reset_api_key():
         secret = secrets.token_hex(16)
         datastore.data['settings']['application']['api_access_token'] = secret
-        datastore.needs_write_urgent = True
+        datastore.commit()
         flash(gettext("API Key was regenerated."))
         return redirect(url_for('settings.settings_page')+'#api')
         
@@ -187,4 +201,32 @@ def construct_blueprint(datastore: ChangeDetectionStore):
                                logs=notification_debug_log if len(notification_debug_log) else ["Notification logs are empty - no notifications sent yet."])
         return output
 
+    @settings_blueprint.route("/toggle-all-paused", methods=['GET'])
+    @login_optionally_required
+    def toggle_all_paused():
+        current_state = datastore.data['settings']['application'].get('all_paused', False)
+        datastore.data['settings']['application']['all_paused'] = not current_state
+        datastore.commit()
+
+        if datastore.data['settings']['application']['all_paused']:
+            flash(gettext("Automatic scheduling paused - checks will not be queued."), 'notice')
+        else:
+            flash(gettext("Automatic scheduling resumed - checks will be queued normally."), 'notice')
+
+        return redirect(url_for('watchlist.index'))
+
+    @settings_blueprint.route("/toggle-all-muted", methods=['GET'])
+    @login_optionally_required
+    def toggle_all_muted():
+        current_state = datastore.data['settings']['application'].get('all_muted', False)
+        datastore.data['settings']['application']['all_muted'] = not current_state
+        datastore.commit()
+
+        if datastore.data['settings']['application']['all_muted']:
+            flash(gettext("All notifications muted."), 'notice')
+        else:
+            flash(gettext("All notifications unmuted."), 'notice')
+
+        return redirect(url_for('watchlist.index'))
+
     return settings_blueprint

changedetectionio/blueprint/settings/templates/settings.html

@@ -25,6 +25,7 @@
             <li class="tab"><a href="#ui-options">{{ _('UI Options') }}</a></li>
             <li class="tab"><a href="#api">{{ _('API') }}</a></li>
             <li class="tab"><a href="#rss">{{ _('RSS') }}</a></li>
+            <li class="tab"><a href="{{ url_for('backups.create') }}">{{ _('Backups') }}</a></li>
             <li class="tab"><a href="#timedate">{{ _('Time & Date') }}</a></li>
             <li class="tab"><a href="#proxies">{{ _('CAPTCHA & Proxies') }}</a></li>
             {% if plugin_tabs %}
@@ -53,33 +54,50 @@
                     </div>
                     <div class="pure-control-group">
                         {{ render_field(form.application.form.filter_failure_notification_threshold_attempts, class="filter_failure_notification_threshold_attempts") }}
-                        <span class="pure-form-message-inline">After this many consecutive times that the CSS/xPath filter is missing, send a notification
+                        <span class="pure-form-message-inline">{{ _('After this many consecutive times that the CSS/xPath filter is missing, send a notification') }}
                             <br>
-                        Set to <strong>0</strong> to disable
+                        {{ _('Set to') }} <strong>0</strong> {{ _('to disable') }}
                         </span>
                     </div>
+                    <div class="pure-control-group">
+                        {{ render_field(form.application.form.history_snapshot_max_length, class="history_snapshot_max_length") }}
+                        <span class="pure-form-message-inline">{{ _('Limit collection of history snapshots for each watch to this number of history items.') }}
+                            <br>
+                        {{ _('Set to empty to disable / no limit') }}
+                        </span>
+                    </div>
+
                     <div class="pure-control-group">
                         {% if not hide_remove_pass %}
                             {% if current_user.is_authenticated %}
                                 {{ render_button(form.application.form.removepassword_button) }}
                             {% else %}
                             {{ render_field(form.application.form.password) }}
-                            <span class="pure-form-message-inline">Password protection for your changedetection.io application.</span>
+                            <span class="pure-form-message-inline">{{ _('Password protection for your changedetection.io application.') }}</span>
                             {% endif %}
                         {% else %}
-                            <span class="pure-form-message-inline">Password is locked.</span>
+                            <span class="pure-form-message-inline">{{ _('Password is locked.') }}</span>
                         {% endif %}
                     </div>
 
                     <div class="pure-control-group">
                         {{ render_checkbox_field(form.application.form.shared_diff_access, class="shared_diff_access") }}
-                        <span class="pure-form-message-inline">Allow access to the watch change history page when password is enabled (Good for sharing the diff page)
-                        </span>
+                        <span class="pure-form-message-inline">{{ _('Allow access to the watch change history page when password is enabled (Good for sharing the diff page)') }}</span>
                     </div>
                     <div class="pure-control-group">
                         {{ render_checkbox_field(form.application.form.empty_pages_are_a_change) }}
-                        <span class="pure-form-message-inline">When a request returns no content, or the HTML does not contain any text, is this considered a change?</span>
+                        <span class="pure-form-message-inline">{{ _('When a request returns no content, or the HTML does not contain any text, is this considered a change?') }}</span>
                     </div>
+                {% if form.requests.proxy %}
+                    <div>
+                    <br>
+                        <div class="inline-radio">
+                            {{ render_field(form.requests.form.proxy, class="fetch-backend-proxy") }}
+                            <span class="pure-form-message-inline">{{ _('Choose a default proxy for all watches') }}</span>
+                        </div>
+                    </div>
+                {% endif %}
+
                 </fieldset>
             </div>
 
@@ -90,8 +108,8 @@
                 <div class="pure-control-group" id="notification-base-url">
                     {{ render_field(form.application.form.base_url, class="m-d") }}
                     <span class="pure-form-message-inline">
-                        Base URL used for the <code>{{ '{{ base_url }}' }}</code> token in notification links.<br>
-                        Default value is the system environment variable '<code>BASE_URL</code>' - <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Configurable-BASE_URL-setting">read more here</a>.
+                        {{ _('Base URL used for the') }} <code>{{ '{{ base_url }}' }}</code> {{ _('token in notification links.') }}<br>
+                        {{ _('Default value is the system environment variable') }} '<code>BASE_URL</code>' - <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Configurable-BASE_URL-setting">{{ _('read more here') }}</a>.
                     </span>
                 </div>
             </div>
@@ -100,15 +118,15 @@
                 <div class="pure-control-group inline-radio">
                     {{ render_field(form.application.form.fetch_backend, class="fetch-backend") }}
                     <span class="pure-form-message-inline">
-                        <p>Use the <strong>Basic</strong> method (default) where your watched sites don't need Javascript to render.</p>
-                        <p>The <strong>Chrome/Javascript</strong> method requires a network connection to a running WebDriver+Chrome server, set by the ENV var 'WEBDRIVER_URL'. </p>
+                        <p>{{ _('Use the') }} <strong>{{ _('Basic') }}</strong> {{ _('method (default) where your watched sites don\'t need Javascript to render.') }}</p>
+                        <p>{{ _('The') }} <strong>{{ _('Chrome/Javascript') }}</strong> {{ _('method requires a network connection to a running WebDriver+Chrome server, set by the ENV var') }} 'WEBDRIVER_URL'. </p>
                     </span>
                 </div>
                 <fieldset class="pure-group" id="webdriver-override-options" data-visible-for="application-fetch_backend=html_webdriver">
                     <div class="pure-form-message-inline">
-                        <strong>If you're having trouble waiting for the page to be fully rendered (text missing etc), try increasing the 'wait' time here.</strong>
+                        <strong>{{ _('If you\'re having trouble waiting for the page to be fully rendered (text missing etc), try increasing the \'wait\' time here.') }}</strong>
                         <br>
-                        This will wait <i>n</i> seconds before extracting the text.
+                        {{ _('This will wait') }} <i>n</i> {{ _('seconds before extracting the text.') }}
                     </div>
                     <div class="pure-control-group">
                         {{ render_field(form.application.form.webdriver_delay) }}
@@ -117,27 +135,27 @@
                 <div class="pure-control-group">
                     {{ render_field(form.requests.form.workers) }}
                     {% set worker_info = get_worker_status_info() %}
-                    <span class="pure-form-message-inline">Number of concurrent workers to process watches. More workers = faster processing but higher memory usage.<br>
-                    Currently running: <strong>{{ worker_info.count }}</strong> operational {{ worker_info.type }} workers{% if worker_info.active_workers > 0 %} ({{ worker_info.active_workers }} actively processing){% endif %}.</span>
+                    <span class="pure-form-message-inline">{{ _('Number of concurrent workers to process watches. More workers = faster processing but higher memory usage.') }}<br>
+                    {{ _('Currently running:') }} <strong>{{ worker_info.count }}</strong> {{ _('operational') }} {{ worker_info.type }} {{ _('workers') }}{% if worker_info.active_workers > 0 %} ({{ worker_info.active_workers }} {{ _('actively processing') }}){% endif %}.</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_field(form.requests.form.jitter_seconds, class="jitter_seconds") }}
-                    <span class="pure-form-message-inline">Example - 3 seconds random jitter could trigger up to 3 seconds earlier or up to 3 seconds later</span>
+                    <span class="pure-form-message-inline">{{ _('Example - 3 seconds random jitter could trigger up to 3 seconds earlier or up to 3 seconds later') }}</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_field(form.requests.form.timeout) }}
-                    <span class="pure-form-message-inline">For regular plain requests (not chrome based), maximum number of seconds until timeout, 1-999.</span><br>
+                    <span class="pure-form-message-inline">{{ _('For regular plain requests (not chrome based), maximum number of seconds until timeout, 1-999.') }}</span><br>
                 </div>
                 <div class="pure-control-group inline-radio">
                     {{ render_field(form.requests.form.default_ua) }}
                     <span class="pure-form-message-inline">
-                        Applied to all requests.<br><br>
-                        Note: Simply changing the User-Agent often does not defeat anti-robot technologies, it's important to consider <a href="https://changedetection.io/tutorial/what-are-main-types-anti-robot-mechanisms">all of the ways that the browser is detected</a>.
+                        {{ _('Applied to all requests.') }}<br><br>
+                        {{ _('Note: Simply changing the User-Agent often does not defeat anti-robot technologies, it\'s important to consider') }} <a href="https://changedetection.io/tutorial/what-are-main-types-anti-robot-mechanisms">{{ _('all of the ways that the browser is detected') }}</a>.
                     </span>
                 </div>
                 <div class="pure-control-group">
                                         <br>
-                    Tip: <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Proxy-configuration#brightdata-proxy-support">Connect using Bright Data and Oxylabs Proxies, find out more here.</a>
+                    {{ _('Tip:') }} <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Proxy-configuration#brightdata-proxy-support">{{ _('Connect using Bright Data and Oxylabs Proxies, find out more here.') }}</a>
 
                 </div>
             </div>
@@ -146,15 +164,15 @@
 
                     <fieldset class="pure-group">
                     {{ render_checkbox_field(form.application.form.ignore_whitespace) }}
-                    <span class="pure-form-message-inline">Ignore whitespace, tabs and new-lines/line-feeds when considering if a change was detected.<br>
-                    <i>Note:</i> Changing this will change the status of your existing watches, possibly trigger alerts etc.
+                    <span class="pure-form-message-inline">{{ _('Ignore whitespace, tabs and new-lines/line-feeds when considering if a change was detected.') }}<br>
+                    <i>{{ _('Note:') }}</i> {{ _('Changing this will change the status of your existing watches, possibly trigger alerts etc.') }}
                     </span>
                     </fieldset>
                 <fieldset class="pure-group">
                     {{ render_checkbox_field(form.application.form.render_anchor_tag_content) }}
-                    <span class="pure-form-message-inline">Render anchor tag content, default disabled, when enabled renders links as <code>(link text)[https://somesite.com]</code>
+                    <span class="pure-form-message-inline">{{ _('Render anchor tag content, default disabled, when enabled renders links as') }} <code>(link text)[https://somesite.com]</code>
                         <br>
-                    <i>Note:</i> Changing this could affect the content of your existing watches, possibly trigger alerts etc.
+                    <i>{{ _('Note:') }}</i> {{ _('Changing this could affect the content of your existing watches, possibly trigger alerts etc.') }}
                     </span>
                     </fieldset>
                     <fieldset class="pure-group">
@@ -165,9 +183,9 @@ nav
 //*[contains(text(), 'Advertisement')]") }}
                       <span class="pure-form-message-inline">
                         <ul>
-                          <li> Remove HTML element(s) by CSS and XPath selectors before text conversion. </li>
-                          <li> Don't paste HTML here, use only CSS and XPath selectors </li>
-                          <li> Add multiple elements, CSS or XPath selectors per line to ignore multiple parts of the HTML. </li>
+                          <li> {{ _('Remove HTML element(s) by CSS and XPath selectors before text conversion.') }} </li>
+                          <li> {{ _('Don\'t paste HTML here, use only CSS and XPath selectors') }} </li>
+                          <li> {{ _('Add multiple elements, CSS or XPath selectors per line to ignore multiple parts of the HTML.') }} </li>
                         </ul>
                       </span>
                     </fieldset>
@@ -175,50 +193,50 @@ nav
                     {{ render_field(form.application.form.global_ignore_text, rows=5, placeholder="Some text to ignore in a line
 /some.regex\d{2}/ for case-INsensitive regex
                     ") }}
-                    <span class="pure-form-message-inline">Note: This is applied globally in addition to the per-watch rules.</span><br>
+                    <span class="pure-form-message-inline">{{ _('Note: This is applied globally in addition to the per-watch rules.') }}</span><br>
                     <span class="pure-form-message-inline">
                         <ul>
-                            <li>Matching text will be <strong>ignored</strong> in the text snapshot (you can still see it but it wont trigger a change)</li>
-                            <li>Note: This is applied globally in addition to the per-watch rules.</li>
-                            <li>Each line processed separately, any line matching will be ignored (removed before creating the checksum)</li>
-                            <li>Regular Expression support, wrap the entire line in forward slash <code>/regex/</code></li>
-                            <li>Changing this will affect the comparison checksum which may trigger an alert</li>
+                            <li>{{ _('Matching text will be') }} <strong>{{ _('ignored') }}</strong> {{ _('in the text snapshot (you can still see it but it wont trigger a change)') }}</li>
+                            <li>{{ _('Note: This is applied globally in addition to the per-watch rules.') }}</li>
+                            <li>{{ _('Each line processed separately, any line matching will be ignored (removed before creating the checksum)') }}</li>
+                            <li>{{ _('Regular Expression support, wrap the entire line in forward slash') }} <code>/regex/</code></li>
+                            <li>{{ _('Changing this will affect the comparison checksum which may trigger an alert') }}</li>
                         </ul>
                      </span>
                     </fieldset>
                     <fieldset class="pure-group">
                         {{ render_checkbox_field(form.application.form.strip_ignored_lines) }}
-                        <span class="pure-form-message-inline">Remove any text that appears in the "Ignore text" from the output (otherwise its just ignored for change-detection)<br>
-                        <i>Note:</i> Changing this will change the status of your existing watches, possibly trigger alerts etc.
+                        <span class="pure-form-message-inline">{{ _('Remove any text that appears in the "Ignore text" from the output (otherwise its just ignored for change-detection)') }}<br>
+                        <i>{{ _('Note:') }}</i> {{ _('Changing this will change the status of your existing watches, possibly trigger alerts etc.') }}
                         </span>
                     </fieldset>
            </div>
 
             <div class="tab-pane-inner" id="api">
-                <h4>API Access</h4>
-                <p>Drive your changedetection.io via API, More about <a href="https://changedetection.io/docs/api_v1/index.html">API access and examples here</a>.</p>
+                <h4>{{ _('API Access') }}</h4>
+                <p>{{ _('Drive your changedetection.io via API, More about') }} <a href="https://changedetection.io/docs/api_v1/index.html">{{ _('API access and examples here') }}</a>.</p>
 
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.api_access_token_enabled) }}
-                    <div class="pure-form-message-inline">Restrict API access limit by using <code>x-api-key</code> header - required for the Chrome Extension to work</div><br>
-                    <div class="pure-form-message-inline"><br>API Key <span id="api-key">{{api_key}}</span>
-                        <span style="display:none;" id="api-key-copy" >copy</span>
+                    <div class="pure-form-message-inline">{{ _('Restrict API access limit by using') }} <code>x-api-key</code> {{ _('header - required for the Chrome Extension to work') }}</div><br>
+                    <div class="pure-form-message-inline"><br>{{ _('API Key') }} <span id="api-key">{{api_key}}</span>
+                        <span style="display:none;" id="api-key-copy" >{{ _('copy') }}</span>
                     </div>
                 </div>
                 <div class="pure-control-group">
-                    <a href="{{url_for('settings.settings_reset_api_key')}}" class="pure-button button-small button-cancel">Regenerate API key</a>
+                    <a href="{{url_for('settings.settings_reset_api_key')}}" class="pure-button button-small button-cancel">{{ _('Regenerate API key') }}</a>
                 </div>
                 <div class="pure-control-group">
-                    <h4>Chrome Extension</h4>
-                    <p>Easily add any web-page to your changedetection.io installation from within Chrome.</p>
-                    <strong>Step 1</strong> Install the extension, <strong>Step 2</strong> Navigate to this page,
-                    <strong>Step 3</strong> Open the extension from the toolbar and click "<i>Sync API Access</i>"
+                    <h4>{{ _('Chrome Extension') }}</h4>
+                    <p>{{ _('Easily add any web-page to your changedetection.io installation from within Chrome.') }}</p>
+                    <strong>{{ _('Step 1') }}</strong> {{ _('Install the extension,') }} <strong>{{ _('Step 2') }}</strong> {{ _('Navigate to this page,') }}
+                    <strong>{{ _('Step 3') }}</strong> {{ _('Open the extension from the toolbar and click') }} "<i>{{ _('Sync API Access') }}</i>"
                     <p>
                         <a id="chrome-extension-link"
-                           title="Try our new Chrome Extension!"
+                           title="{{ _('Try our new Chrome Extension!') }}"
                            href="https://chromewebstore.google.com/detail/changedetectionio-website/kefcfmgmlhmankjmnbijimhofdjekbop">
-                            <img alt="Chrome store icon" src="{{ url_for('static_content', group='images', filename='google-chrome-icon.png') }}" >
-                            Chrome Webstore
+                            <img alt="{{ _('Chrome store icon') }}" src="{{ url_for('static_content', group='images', filename='google-chrome-icon.png') }}" >
+                            {{ _('Chrome Webstore') }}
                         </a>
                     </p>
                 </div>
@@ -229,20 +247,20 @@ nav
                 </div>
                 <div class="pure-control-group">
                     {{ render_field(form.application.form.rss_diff_length) }}
-                    <span class="pure-form-message-inline">Maximum number of history snapshots to include in the watch specific RSS feed.</span>
+                    <span class="pure-form-message-inline">{{ _('Maximum number of history snapshots to include in the watch specific RSS feed.') }}</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.rss_reader_mode) }}
-                    <span class="pure-form-message-inline">For watching other RSS feeds - When watching RSS/Atom feeds, convert them into clean text for better change detection.</span>
+                    <span class="pure-form-message-inline">{{ _('For watching other RSS feeds - When watching RSS/Atom feeds, convert them into clean text for better change detection.') }}</span>
                 </div>
                 <div class="pure-control-group grey-form-border">
                     <div class="pure-control-group">
                         {{ render_field(form.application.form.rss_content_format) }}
-                        <span class="pure-form-message-inline">Does your reader support HTML? Set it here</span>
+                        <span class="pure-form-message-inline">{{ _('Does your reader support HTML? Set it here') }}</span>
                     </div>
                     <div class="pure-control-group">
                         {{ render_field(form.application.form.rss_template_type) }}
-                        <span class="pure-form-message-inline">'System default' for the same template for all items, or re-use your "Notification Body" as the template.</span>
+                        <span class="pure-form-message-inline">{{ _('\'System default\' for the same template for all items, or re-use your "Notification Body" as the template.') }}</span>
                     </div>
                     <div>
                        {{ render_field(form.application.form.rss_template_override) }}
@@ -255,11 +273,11 @@ nav
             </div>
                 <div class="tab-pane-inner" id="timedate">
                 <div class="pure-control-group">
-                    Ensure the settings below are correct, they are used to manage the time schedule for checking your web page watches.
+                    {{ _('Ensure the settings below are correct, they are used to manage the time schedule for checking your web page watches.') }}
                 </div>
                 <div class="pure-control-group">
-                    <p><strong>UTC Time &amp; Date from Server:</strong> <span id="utc-time" >{{ utc_time }}</span></p>
-                    <p><strong>Local Time &amp; Date in Browser:</strong> <span class="local-time" data-utc="{{ utc_time }}"></span></p>
+                    <p><strong>{{ _('UTC Time & Date from Server:') }}</strong> <span id="utc-time" >{{ utc_time }}</span></p>
+                    <p><strong>{{ _('Local Time & Date in Browser:') }}</strong> <span class="local-time" data-utc="{{ utc_time }}"></span></p>
                     <div>
                        {{ render_field(form.application.form.scheduler_timezone_default) }}
                         <datalist id="timezones" style="display: none;">
@@ -271,22 +289,22 @@ nav
             <div class="tab-pane-inner" id="ui-options">
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.ui.form.open_diff_in_new_tab, class="open_diff_in_new_tab") }}
-                    <span class="pure-form-message-inline">Enable this setting to open the diff page in a new tab. If disabled, the diff page will open in the current tab.</span>
+                    <span class="pure-form-message-inline">{{ _('Enable this setting to open the diff page in a new tab. If disabled, the diff page will open in the current tab.') }}</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.ui.form.socket_io_enabled, class="socket_io_enabled") }}
-                    <span class="pure-form-message-inline">Realtime UI Updates Enabled - (Restart required if this is changed)</span>
+                    <span class="pure-form-message-inline">{{ _('Realtime UI Updates Enabled - (Restart required if this is changed)') }}</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.ui.form.favicons_enabled, class="") }}
-                    <span class="pure-form-message-inline">Enable or Disable Favicons next to the watch list</span>
+                    <span class="pure-form-message-inline">{{ _('Enable or Disable Favicons next to the watch list') }}</span>
                 </div>
                 <div class="pure-control-group">
                     {{ render_checkbox_field(form.application.form.ui.use_page_title_in_list) }}
                 </div>
                 <div class="pure-control-group">
                     {{ render_field(form.application.form.pager_size) }}
-                    <span class="pure-form-message-inline">Number of items per page in the watch overview list, 0 to disable.</span>
+                    <span class="pure-form-message-inline">{{ _('Number of items per page in the watch overview list, 0 to disable.') }}</span>
                 </div>
 
             </div>
@@ -334,21 +352,12 @@ nav
                     </div>
                 </div>
 
-                <p><strong>Tip</strong>: "Residential" and "Mobile" proxy type can be more successfull than "Data Center" for blocked websites.</p>
+                <p><strong>{{ _('Tip') }}</strong>: {{ _('"Residential" and "Mobile" proxy type can be more successfull than "Data Center" for blocked websites.') }}</p>
 
                 <div class="pure-control-group" id="extra-proxies-setting">
                 {{ render_fieldlist_with_inline_errors(form.requests.form.extra_proxies) }}
-                <span class="pure-form-message-inline">"Name" will be used for selecting the proxy in the Watch Edit settings</span><br>
-                <span class="pure-form-message-inline">SOCKS5 proxies with authentication are only supported with 'plain requests' fetcher, for other fetchers you should whitelist the IP access instead</span>
-                {% if form.requests.proxy %}
-                <div>
-                <br>
-                    <div class="inline-radio">
-                        {{ render_field(form.requests.form.proxy, class="fetch-backend-proxy") }}
-                        <span class="pure-form-message-inline">Choose a default proxy for all watches</span>
-                    </div>
-                </div>
-                {% endif %}
+                <span class="pure-form-message-inline">{{ _('"Name" will be used for selecting the proxy in the Watch Edit settings') }}</span><br>
+                <span class="pure-form-message-inline">{{ _('SOCKS5 proxies with authentication are only supported with \'plain requests\' fetcher, for other fetchers you should whitelist the IP access instead') }}</span>
                 </div>
                 <div class="pure-control-group" id="extra-browsers-setting">
                     <p>
@@ -385,6 +394,7 @@ nav
                 {% endfor %}
             {% endif %}
             <div class="tab-pane-inner" id="info">
+                <p><strong>{{ _('Uptime:') }}</strong> {{ uptime_seconds|format_duration }}</p>
                 <p><strong>{{ _('Python version:') }}</strong> {{ python_version }}</p>
                 <p><strong>{{ _('Plugins active:') }}</strong></p>
                 {% if active_plugins %}

changedetectionio/blueprint/tags/__init__.py

@@ -1,5 +1,7 @@
+import threading
 from flask import Blueprint, request, render_template, flash, url_for, redirect
 from flask_babel import gettext
+from loguru import logger
 
 from changedetectionio.store import ChangeDetectionStore
 from changedetectionio.flask_app import login_optionally_required
@@ -52,52 +54,94 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
         return redirect(url_for('tags.tags_overview_page'))
 
-    @tags_blueprint.route("/mute/<string:uuid>", methods=['GET'])
+    @tags_blueprint.route("/mute/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def mute(uuid):
-        if datastore.data['settings']['application']['tags'].get(uuid):
-            datastore.data['settings']['application']['tags'][uuid]['notification_muted'] = not datastore.data['settings']['application']['tags'][uuid]['notification_muted']
+        tag = datastore.data['settings']['application']['tags'].get(uuid)
+        if tag:
+            tag['notification_muted'] = not tag['notification_muted']
+            tag.commit()
         return redirect(url_for('tags.tags_overview_page'))
 
-    @tags_blueprint.route("/delete/<string:uuid>", methods=['GET'])
+    @tags_blueprint.route("/delete/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def delete(uuid):
-        removed = 0
-        # Delete the tag, and any tag reference
+        # Delete the tag from settings immediately
         if datastore.data['settings']['application']['tags'].get(uuid):
             del datastore.data['settings']['application']['tags'][uuid]
 
-        for watch_uuid, watch in datastore.data['watching'].items():
-            if watch.get('tags') and uuid in watch['tags']:
-                removed += 1
-                watch['tags'].remove(uuid)
+        # Remove tag from all watches in background thread to avoid blocking
+        def remove_tag_background(tag_uuid):
+            """Background thread to remove tag from watches - discarded after completion."""
+            removed_count = 0
+            try:
+                for watch_uuid, watch in datastore.data['watching'].items():
+                    if watch.get('tags') and tag_uuid in watch['tags']:
+                        watch['tags'].remove(tag_uuid)
+                        watch.commit()
+                        removed_count += 1
+                logger.info(f"Background: Tag {tag_uuid} removed from {removed_count} watches")
+            except Exception as e:
+                logger.error(f"Error removing tag from watches: {e}")
 
-        flash(gettext("Tag deleted and removed from {} watches").format(removed))
+        # Start daemon thread
+        threading.Thread(target=remove_tag_background, args=(uuid,), daemon=True).start()
+
+        flash(gettext("Tag deleted, removing from watches in background"))
         return redirect(url_for('tags.tags_overview_page'))
 
-    @tags_blueprint.route("/unlink/<string:uuid>", methods=['GET'])
+    @tags_blueprint.route("/unlink/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def unlink(uuid):
-        unlinked = 0
-        for watch_uuid, watch in datastore.data['watching'].items():
-            if watch.get('tags') and uuid in watch['tags']:
-                unlinked += 1
-                watch['tags'].remove(uuid)
+        # Unlink tag from all watches in background thread to avoid blocking
+        def unlink_tag_background(tag_uuid):
+            """Background thread to unlink tag from watches - discarded after completion."""
+            unlinked_count = 0
+            try:
+                for watch_uuid, watch in datastore.data['watching'].items():
+                    if watch.get('tags') and tag_uuid in watch['tags']:
+                        watch['tags'].remove(tag_uuid)
+                        watch.commit()
+                        unlinked_count += 1
+                logger.info(f"Background: Tag {tag_uuid} unlinked from {unlinked_count} watches")
+            except Exception as e:
+                logger.error(f"Error unlinking tag from watches: {e}")
 
-        flash(gettext("Tag unlinked removed from {} watches").format(unlinked))
+        # Start daemon thread
+        threading.Thread(target=unlink_tag_background, args=(uuid,), daemon=True).start()
+
+        flash(gettext("Unlinking tag from watches in background"))
         return redirect(url_for('tags.tags_overview_page'))
 
     @tags_blueprint.route("/delete_all", methods=['GET'])
     @login_optionally_required
     def delete_all():
-        for watch_uuid, watch in datastore.data['watching'].items():
-            watch['tags'] = []
-        datastore.data['settings']['application']['tags'] = {}
 
-        flash(gettext("All tags deleted"))
+        for tag_uuid in list(datastore.data['settings']['application']['tags'].keys()):
+# TagsDict 'del' handler will remove the dir
+            del datastore.data['settings']['application']['tags'][tag_uuid]
+
+
+        # Clear tags from all watches in background thread to avoid blocking
+        def clear_all_tags_background():
+            """Background thread to clear tags from all watches - discarded after completion."""
+            cleared_count = 0
+            try:
+                for watch_uuid, watch in datastore.data['watching'].items():
+                    watch['tags'] = []
+                    watch.commit()
+                    cleared_count += 1
+                logger.info(f"Background: Cleared tags from {cleared_count} watches")
+            except Exception as e:
+                logger.error(f"Error clearing tags from watches: {e}")
+
+        # Start daemon thread
+        threading.Thread(target=clear_all_tags_background, daemon=True).start()
+
+        flash(gettext("All tags deleted, clearing from watches in background"))
         return redirect(url_for('tags.tags_overview_page'))
 
-    @tags_blueprint.route("/edit/<string:uuid>", methods=['GET'])
+    @tags_blueprint.route("/edit/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def form_tag_edit(uuid):
         from changedetectionio.blueprint.tags.form import group_restock_settings_form
@@ -116,6 +160,21 @@ def construct_blueprint(datastore: ChangeDetectionStore):
                                        default_system_settings = datastore.data['settings'],
                                        )
 
+        # Bridge API-stored processor_config_* values into the form's FormField sub-forms.
+        # The API stores processor_config_restock_diff in the tag dict; find the matching
+        # FormField by checking which one's sub-fields cover the config keys.
+        from wtforms.fields.form import FormField as WTFormField
+        for key, value in default.items():
+            if not key.startswith('processor_config_') or not isinstance(value, dict):
+                continue
+            for form_field in form:
+                if isinstance(form_field, WTFormField) and all(k in form_field.form._fields for k in value):
+                    for sub_key, sub_value in value.items():
+                        sub_field = form_field.form._fields.get(sub_key)
+                        if sub_field is not None:
+                            sub_field.data = sub_value
+                    break
+
         template_args = {
             'data': default,
             'form': form,
@@ -159,17 +218,17 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         return output
 
 
-    @tags_blueprint.route("/edit/<string:uuid>", methods=['POST'])
+    @tags_blueprint.route("/edit/<uuid_str:uuid>", methods=['POST'])
     @login_optionally_required
     def form_tag_edit_submit(uuid):
         from changedetectionio.blueprint.tags.form import group_restock_settings_form
         if uuid == 'first':
             uuid = list(datastore.data['settings']['application']['tags'].keys()).pop()
 
-        default = datastore.data['settings']['application']['tags'].get(uuid)
+        tag = datastore.data['settings']['application']['tags'].get(uuid)
 
         form = group_restock_settings_form(formdata=request.form if request.method == 'POST' else None,
-                               data=default,
+                               data=tag,
                                extra_notification_tokens=datastore.get_unique_notification_tokens_available()
                                )
         # @todo subclass form so validation works
@@ -178,15 +237,18 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 #                flash(','.join(l), 'error')
 #           return redirect(url_for('tags.form_tag_edit_submit', uuid=uuid))
 
-        datastore.data['settings']['application']['tags'][uuid].update(form.data)
-        datastore.data['settings']['application']['tags'][uuid]['processor'] = 'restock_diff'
-        datastore.needs_write_urgent = True
+        tag.update(form.data)
+        tag['processor'] = 'restock_diff'
+        tag.commit()
+
+        # Clear checksums for all watches using this tag to force reprocessing
+        # Tag changes affect inherited configuration
+        cleared_count = datastore.clear_checksums_for_tag(uuid)
+        logger.info(f"Tag {uuid} updated, cleared {cleared_count} watch checksums")
+
         flash(gettext("Updated"))
 
         return redirect(url_for('tags.tags_overview_page'))
 
 
-    @tags_blueprint.route("/delete/<string:uuid>", methods=['GET'])
-    def form_tag_delete(uuid):
-        return redirect(url_for('tags.tags_overview_page'))
     return tags_blueprint

changedetectionio/blueprint/tags/templates/groups-overview.html

@@ -50,7 +50,8 @@
                 <td>{{ "{:,}".format(tag_count[uuid]) if uuid in tag_count else 0 }}</td>
                 <td class="title-col inline"> <a href="{{url_for('watchlist.index', tag=uuid) }}">{{ tag.title }}</a></td>
                 <td>
-                    <a class="pure-button pure-button-primary" href="{{ url_for('tags.form_tag_edit', uuid=uuid) }}">{{ _('Edit') }}</a>&nbsp;
+                    <a class="pure-button pure-button-primary" href="{{ url_for('tags.form_tag_edit', uuid=uuid) }}">{{ _('Edit') }}</a>
+                    <a href="{{ url_for('ui.form_watch_checknow', tag=uuid) }}" class="pure-button pure-button-primary" >{{ _('Recheck') }}</a>
                     <a class="pure-button button-error"
                        href="{{ url_for('tags.delete', uuid=uuid) }}"
                        data-requires-confirm

changedetectionio/blueprint/ui/__init__.py

@@ -1,5 +1,6 @@
 import time
-from flask import Blueprint, request, redirect, url_for, flash, render_template, session
+import threading
+from flask import Blueprint, request, redirect, url_for, flash, render_template, session, current_app
 from flask_babel import gettext
 from loguru import logger
 
@@ -9,7 +10,7 @@ from changedetectionio.blueprint.ui.notification import construct_blueprint as c
 from changedetectionio.blueprint.ui.views import construct_blueprint as construct_views_blueprint
 from changedetectionio.blueprint.ui import diff, preview
 
-def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWatchMetaData, watch_check_update, extra_data=None, emit_flash=True):
+def _handle_operations(op, uuids, datastore, worker_pool, update_q, queuedWatchMetaData, watch_check_update, extra_data=None, emit_flash=True):
     from flask import request, flash
 
     if op == 'delete':
@@ -23,6 +24,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 datastore.data['watching'][uuid]['paused'] = True
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches paused").format(len(uuids)))
 
@@ -30,6 +32,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 datastore.data['watching'][uuid.strip()]['paused'] = False
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches unpaused").format(len(uuids)))
 
@@ -44,6 +47,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 datastore.data['watching'][uuid]['notification_muted'] = True
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches muted").format(len(uuids)))
 
@@ -51,6 +55,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 datastore.data['watching'][uuid]['notification_muted'] = False
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches un-muted").format(len(uuids)))
 
@@ -58,7 +63,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 # Recheck and require a full reprocessing
-                worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
         if emit_flash:
             flash(gettext("{} watches queued for rechecking").format(len(uuids)))
 
@@ -66,6 +71,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             if datastore.data['watching'].get(uuid):
                 datastore.data['watching'][uuid]["last_error"] = False
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches errors cleared").format(len(uuids)))
 
@@ -86,6 +92,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
                 datastore.data['watching'][uuid]['notification_body'] = None
                 datastore.data['watching'][uuid]['notification_urls'] = []
                 datastore.data['watching'][uuid]['notification_format'] = USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH
+                datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches set to use default notification settings").format(len(uuids)))
 
@@ -101,6 +108,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
                             datastore.data['watching'][uuid]['tags'] = []
 
                         datastore.data['watching'][uuid]['tags'].append(tag_uuid)
+                        datastore.data['watching'][uuid].commit()
         if emit_flash:
             flash(gettext("{} watches were tagged").format(len(uuids)))
 
@@ -108,7 +116,7 @@ def _handle_operations(op, uuids, datastore, worker_handler, update_q, queuedWat
         for uuid in uuids:
             watch_check_update.send(watch_uuid=uuid)
 
-def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handler, queuedWatchMetaData, watch_check_update):
+def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_pool, queuedWatchMetaData, watch_check_update):
     ui_blueprint = Blueprint('ui', __name__, template_folder="templates")
     
     # Register the edit blueprint
@@ -133,7 +141,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
     # Import the login decorator
     from changedetectionio.auth_decorator import login_optionally_required
 
-    @ui_blueprint.route("/clear_history/<string:uuid>", methods=['GET'])
+    @ui_blueprint.route("/clear_history/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def clear_watch_history(uuid):
         try:
@@ -148,12 +156,27 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
     @login_optionally_required
     def clear_all_history():
         if request.method == 'POST':
-            confirmtext = request.form.get('confirmtext')
+            confirmtext = request.form.get('confirmtext', '')
 
-            if confirmtext == 'clear':
-                for uuid in datastore.data['watching'].keys():
-                    datastore.clear_watch_history(uuid)
-                flash(gettext("Cleared snapshot history for all watches"))
+            if confirmtext.strip().lower() == gettext('clear').strip().lower():
+                # Run in background thread to avoid blocking
+                def clear_history_background():
+                    # Capture UUIDs first to avoid race conditions
+                    watch_uuids = list(datastore.data['watching'].keys())
+                    logger.info(f"Background: Clearing history for {len(watch_uuids)} watches")
+
+                    for uuid in watch_uuids:
+                        try:
+                            datastore.clear_watch_history(uuid)
+                        except Exception as e:
+                            logger.error(f"Error clearing history for watch {uuid}: {e}")
+
+                    logger.info("Background: Completed clearing history")
+
+                # Start daemon thread
+                threading.Thread(target=clear_history_background, daemon=True).start()
+
+                flash(gettext("History clearing started in background"))
             else:
                 flash(gettext('Incorrect confirmation text.'), 'error')
 
@@ -169,17 +192,37 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
         # Save the current newest history as the most recently viewed
         with_errors = request.args.get('with_errors') == "1"
         tag_limit = request.args.get('tag')
-        logger.debug(f"Limiting to tag {tag_limit}")
         now = int(time.time())
-        for watch_uuid, watch in datastore.data['watching'].items():
-            if with_errors and not watch.get('last_error'):
-                continue
 
-            if tag_limit and ( not watch.get('tags') or tag_limit not in watch['tags'] ):
-                logger.debug(f"Skipping watch {watch_uuid}")
-                continue
+        # Mark watches as viewed - use background thread only for large watch counts
+        def mark_viewed_impl():
+            """Mark watches as viewed - can run synchronously or in background thread."""
+            marked_count = 0
+            try:
+                for watch_uuid, watch in datastore.data['watching'].items():
+                    if with_errors and not watch.get('last_error'):
+                        continue
 
-            datastore.set_last_viewed(watch_uuid, now)
+                    if tag_limit and (not watch.get('tags') or tag_limit not in watch['tags']):
+                        continue
+
+                    datastore.set_last_viewed(watch_uuid, now)
+                    marked_count += 1
+
+                logger.info(f"Marking complete: {marked_count} watches marked as viewed")
+            except Exception as e:
+                logger.error(f"Error marking as viewed: {e}")
+
+        # For small watch counts (< 10), run synchronously to avoid race conditions in tests
+        # For larger counts, use background thread to avoid blocking the UI
+        watch_count = len(datastore.data['watching'])
+        if watch_count < 10:
+            # Run synchronously for small watch counts
+            mark_viewed_impl()
+        else:
+            # Start background thread for large watch counts
+            thread = threading.Thread(target=mark_viewed_impl, daemon=True)
+            thread.start()
 
         return redirect(url_for('watchlist.index', tag=tag_limit))
 
@@ -187,14 +230,14 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
     @login_optionally_required
     def form_delete():
         uuid = request.args.get('uuid')
+        # More for testing, possible to return the first/only
+        if uuid == 'first':
+            uuid = list(datastore.data['watching'].keys()).pop()
 
         if uuid != 'all' and not uuid in datastore.data['watching'].keys():
             flash(gettext('The watch by UUID {} does not exist.').format(uuid), 'error')
             return redirect(url_for('watchlist.index'))
 
-        # More for testing, possible to return the first/only
-        if uuid == 'first':
-            uuid = list(datastore.data['watching'].keys()).pop()
         datastore.delete(uuid)
         flash(gettext('Deleted.'))
 
@@ -204,14 +247,14 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
     @login_optionally_required
     def form_clone():
         uuid = request.args.get('uuid')
-        # More for testing, possible to return the first/only
+
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
 
         new_uuid = datastore.clone(uuid)
 
         if not datastore.data['watching'].get(uuid).get('paused'):
-            worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=5, item={'uuid': new_uuid}))
+            worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=5, item={'uuid': new_uuid}))
 
         flash(gettext('Cloned, you are editing the new watch.'))
 
@@ -225,38 +268,81 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
         uuid = request.args.get('uuid')
         with_errors = request.args.get('with_errors') == "1"
 
-        i = 0
-
-        running_uuids = worker_handler.get_running_uuids()
-
         if uuid:
-            if uuid not in running_uuids:
-                worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
-                i += 1
-
+            # Single watch - check if already queued or running
+            if worker_pool.is_watch_running(uuid) or uuid in update_q.get_queued_uuids():
+                flash(gettext("Watch is already queued or being checked."))
+            else:
+                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                flash(gettext("Queued 1 watch for rechecking."))
         else:
-            # Recheck all, including muted
-            # Get most overdue first
+            # Multiple watches - first count how many need to be queued
+            watches_to_queue = []
             for k in sorted(datastore.data['watching'].items(), key=lambda item: item[1].get('last_checked', 0)):
                 watch_uuid = k[0]
                 watch = k[1]
-                if not watch['paused']:
-                    if watch_uuid not in running_uuids:
-                        if with_errors and not watch.get('last_error'):
-                            continue
+                if not watch['paused'] and watch_uuid:
+                    if with_errors and not watch.get('last_error'):
+                        continue
+                    if tag != None and tag not in watch['tags']:
+                        continue
+                    watches_to_queue.append(watch_uuid)
 
-                        if tag != None and tag not in watch['tags']:
-                            continue
+            # If less than 20 watches, queue synchronously for immediate feedback
+            if len(watches_to_queue) < 20:
+                # Get already queued/running UUIDs once (efficient)
+                queued_uuids = set(update_q.get_queued_uuids())
+                running_uuids = set(worker_pool.get_running_uuids())
 
-                        worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
-                        i += 1
+                # Filter out watches that are already queued or running
+                watches_to_queue_filtered = []
+                for watch_uuid in watches_to_queue:
+                    if watch_uuid not in queued_uuids and watch_uuid not in running_uuids:
+                        watches_to_queue_filtered.append(watch_uuid)
 
-        if i == 1:
-            flash(gettext("Queued 1 watch for rechecking."))
-        if i > 1:
-            flash(gettext("Queued {} watches for rechecking.").format(i))
-        if i == 0:
-            flash(gettext("No watches available to recheck."))
+                # Queue only the filtered watches
+                for watch_uuid in watches_to_queue_filtered:
+                    worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
+
+                # Provide feedback about skipped watches
+                skipped_count = len(watches_to_queue) - len(watches_to_queue_filtered)
+                if skipped_count > 0:
+                    flash(gettext("Queued {} watches for rechecking ({} already queued or running).").format(
+                        len(watches_to_queue_filtered), skipped_count))
+                else:
+                    if len(watches_to_queue_filtered) == 1:
+                        flash(gettext("Queued 1 watch for rechecking."))
+                    else:
+                        flash(gettext("Queued {} watches for rechecking.").format(len(watches_to_queue_filtered)))
+            else:
+                # 20+ watches - queue in background thread to avoid blocking HTTP response
+                # Capture queued/running state before background thread
+                queued_uuids = set(update_q.get_queued_uuids())
+                running_uuids = set(worker_pool.get_running_uuids())
+
+                def queue_watches_background():
+                    """Background thread to queue watches - discarded after completion."""
+                    try:
+                        queued_count = 0
+                        skipped_count = 0
+                        for watch_uuid in watches_to_queue:
+                            # Check if already queued or running (state captured at start)
+                            if watch_uuid not in queued_uuids and watch_uuid not in running_uuids:
+                                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': watch_uuid}))
+                                queued_count += 1
+                            else:
+                                skipped_count += 1
+
+                        logger.info(f"Background queueing complete: {queued_count} watches queued, {skipped_count} skipped (already queued/running)")
+                    except Exception as e:
+                        logger.error(f"Error in background queueing: {e}")
+
+                # Start background thread and return immediately
+                thread = threading.Thread(target=queue_watches_background, daemon=True, name="QueueWatches-Background")
+                thread.start()
+
+                # Return immediately with approximate message
+                flash(gettext("Queueing watches for rechecking in background..."))
 
         return redirect(url_for('watchlist.index', **({'tag': tag} if tag else {})))
 
@@ -271,7 +357,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
             extra_data=extra_data,
             queuedWatchMetaData=queuedWatchMetaData,
             uuids=uuids,
-            worker_handler=worker_handler,
+            worker_pool=worker_pool,
             update_q=update_q,
             watch_check_update=watch_check_update,
             op=op,
@@ -280,7 +366,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
         return redirect(url_for('watchlist.index'))
 
 
-    @ui_blueprint.route("/share-url/<string:uuid>", methods=['GET'])
+    @ui_blueprint.route("/share-url/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def form_share_put_watch(uuid):
         """Given a watch UUID, upload the info and return a share-link
@@ -289,9 +375,6 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
         import json
         from copy import deepcopy
 
-        # more for testing
-        if uuid == 'first':
-            uuid = list(datastore.data['watching'].keys()).pop()
 
         # copy it to memory as trim off what we dont need (history)
         watch = deepcopy(datastore.data['watching'].get(uuid))
@@ -331,4 +414,25 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_handle
 
         return redirect(url_for('watchlist.index'))
 
+    @ui_blueprint.route("/language/auto-detect", methods=['GET'])
+    def delete_locale_language_session_var_if_it_exists():
+        """Clear the session locale preference to auto-detect from browser Accept-Language header"""
+        if 'locale' in session:
+            session.pop('locale', None)
+            # Refresh Flask-Babel to clear cached locale
+            from flask_babel import refresh
+            refresh()
+            flash(gettext("Language set to auto-detect from browser"))
+
+        # Check if there's a redirect parameter to return to the same page
+        redirect_url = request.args.get('redirect')
+
+        # If redirect is provided and safe, use it
+        from changedetectionio.is_safe_url import is_safe_url
+        if redirect_url and is_safe_url(redirect_url, current_app):
+            return redirect(redirect_url)
+
+        # Otherwise redirect to watchlist
+        return redirect(url_for('watchlist.index'))
+
     return ui_blueprint

changedetectionio/blueprint/ui/diff.py

@@ -66,7 +66,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
         return Markup(result)
 
-    @diff_blueprint.route("/diff/<string:uuid>", methods=['GET'])
+    @diff_blueprint.route("/diff/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def diff_history_page(uuid):
         """
@@ -83,7 +83,6 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         If a processor doesn't have a difference module, falls back to text_json_diff.
         """
 
-        # More for testing, possible to return the first/only
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
 
@@ -101,23 +100,21 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's difference module
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.difference')
+        # Try to get the processor's difference module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'difference')
 
-            # Call the processor's render() function
-            if hasattr(processor_module, 'render'):
-                return processor_module.render(
-                    watch=watch,
-                    datastore=datastore,
-                    request=request,
-                    url_for=url_for,
-                    render_template=render_template,
-                    flash=flash,
-                    redirect=redirect
-                )
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.warning(f"Processor {processor_name} does not have a difference module, falling back to text_json_diff: {e}")
+        # Call the processor's render() function
+        if processor_module and hasattr(processor_module, 'render'):
+            return processor_module.render(
+                watch=watch,
+                datastore=datastore,
+                request=request,
+                url_for=url_for,
+                render_template=render_template,
+                flash=flash,
+                redirect=redirect
+            )
 
         # Fallback: if processor doesn't have difference module, use text_json_diff as default
         from changedetectionio.processors.text_json_diff.difference import render as default_render
@@ -131,7 +128,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             redirect=redirect
         )
 
-    @diff_blueprint.route("/diff/<string:uuid>/extract", methods=['GET'])
+    @diff_blueprint.route("/diff/<uuid_str:uuid>/extract", methods=['GET'])
     @login_optionally_required
     def diff_history_page_extract_GET(uuid):
         """
@@ -144,10 +141,10 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         Each processor implements processors/{type}/extract.py::render_form()
         If a processor doesn't have an extract module, falls back to text_json_diff.
         """
-        # More for testing, possible to return the first/only
+
+
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
-
         try:
             watch = datastore.data['watching'][uuid]
         except KeyError:
@@ -157,23 +154,21 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's extract module
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.extract')
+        # Try to get the processor's extract module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'extract')
 
-            # Call the processor's render_form() function
-            if hasattr(processor_module, 'render_form'):
-                return processor_module.render_form(
-                    watch=watch,
-                    datastore=datastore,
-                    request=request,
-                    url_for=url_for,
-                    render_template=render_template,
-                    flash=flash,
-                    redirect=redirect
-                )
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.warning(f"Processor {processor_name} does not have an extract module, falling back to base extractor: {e}")
+        # Call the processor's render_form() function
+        if processor_module and hasattr(processor_module, 'render_form'):
+            return processor_module.render_form(
+                watch=watch,
+                datastore=datastore,
+                request=request,
+                url_for=url_for,
+                render_template=render_template,
+                flash=flash,
+                redirect=redirect
+            )
 
         # Fallback: if processor doesn't have extract module, use base processors.extract as default
         from changedetectionio.processors.extract import render_form as default_render_form
@@ -187,7 +182,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             redirect=redirect
         )
 
-    @diff_blueprint.route("/diff/<string:uuid>/extract", methods=['POST'])
+    @diff_blueprint.route("/diff/<uuid_str:uuid>/extract", methods=['POST'])
     @login_optionally_required
     def diff_history_page_extract_POST(uuid):
         """
@@ -200,7 +195,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         Each processor implements processors/{type}/extract.py::process_extraction()
         If a processor doesn't have an extract module, falls back to text_json_diff.
         """
-        # More for testing, possible to return the first/only
+
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
 
@@ -213,24 +208,22 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's extract module
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.extract')
+        # Try to get the processor's extract module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'extract')
 
-            # Call the processor's process_extraction() function
-            if hasattr(processor_module, 'process_extraction'):
-                return processor_module.process_extraction(
-                    watch=watch,
-                    datastore=datastore,
-                    request=request,
-                    url_for=url_for,
-                    make_response=make_response,
-                    send_from_directory=send_from_directory,
-                    flash=flash,
-                    redirect=redirect
-                )
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.warning(f"Processor {processor_name} does not have an extract module, falling back to base extractor: {e}")
+        # Call the processor's process_extraction() function
+        if processor_module and hasattr(processor_module, 'process_extraction'):
+            return processor_module.process_extraction(
+                watch=watch,
+                datastore=datastore,
+                request=request,
+                url_for=url_for,
+                make_response=make_response,
+                send_from_directory=send_from_directory,
+                flash=flash,
+                redirect=redirect
+            )
 
         # Fallback: if processor doesn't have extract module, use base processors.extract as default
         from changedetectionio.processors.extract import process_extraction as default_process_extraction
@@ -245,7 +238,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             redirect=redirect
         )
 
-    @diff_blueprint.route("/diff/<string:uuid>/processor-asset/<string:asset_name>", methods=['GET'])
+    @diff_blueprint.route("/diff/<uuid_str:uuid>/processor-asset/<string:asset_name>", methods=['GET'])
     @login_optionally_required
     def processor_asset(uuid, asset_name):
         """
@@ -267,7 +260,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         - /diff/{uuid}/processor-asset/after
         - /diff/{uuid}/processor-asset/rendered_diff
         """
-        # More for testing, possible to return the first/only
+
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
 
@@ -280,38 +273,33 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's difference module
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.difference')
+        # Try to get the processor's difference module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'difference')
 
-            # Call the processor's get_asset() function
-            if hasattr(processor_module, 'get_asset'):
-                result = processor_module.get_asset(
-                    asset_name=asset_name,
-                    watch=watch,
-                    datastore=datastore,
-                    request=request
-                )
+        # Call the processor's get_asset() function
+        if processor_module and hasattr(processor_module, 'get_asset'):
+            result = processor_module.get_asset(
+                asset_name=asset_name,
+                watch=watch,
+                datastore=datastore,
+                request=request
+            )
 
-                if result is None:
-                    from flask import abort
-                    abort(404, description=f"Asset '{asset_name}' not found")
-
-                binary_data, content_type, cache_control = result
-
-                response = make_response(binary_data)
-                response.headers['Content-Type'] = content_type
-                if cache_control:
-                    response.headers['Cache-Control'] = cache_control
-                return response
-            else:
-                logger.warning(f"Processor {processor_name} does not implement get_asset()")
+            if result is None:
                 from flask import abort
-                abort(404, description=f"Processor '{processor_name}' does not support assets")
+                abort(404, description=f"Asset '{asset_name}' not found")
 
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.warning(f"Processor {processor_name} does not have a difference module: {e}")
+            binary_data, content_type, cache_control = result
+
+            response = make_response(binary_data)
+            response.headers['Content-Type'] = content_type
+            if cache_control:
+                response.headers['Cache-Control'] = cache_control
+            return response
+        else:
+            logger.warning(f"Processor {processor_name} does not implement get_asset()")
             from flask import abort
-            abort(404, description=f"Processor '{processor_name}' not found")
+            abort(404, description=f"Processor '{processor_name}' does not support assets")
 
     return diff_blueprint

changedetectionio/blueprint/ui/edit.py

@@ -9,7 +9,7 @@ from jinja2 import Environment, FileSystemLoader
 from changedetectionio.store import ChangeDetectionStore
 from changedetectionio.auth_decorator import login_optionally_required
 from changedetectionio.time_handler import is_within_schedule
-from changedetectionio import worker_handler
+from changedetectionio import worker_pool
 
 def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMetaData):
     edit_blueprint = Blueprint('ui_edit', __name__, template_folder="../ui/templates")
@@ -20,24 +20,23 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             if tag_uuid in watch.get('tags', []) and (tag.get('include_filters') or tag.get('subtractive_selectors')):
                 return True
 
-    @edit_blueprint.route("/edit/<string:uuid>", methods=['GET', 'POST'])
+    @edit_blueprint.route("/edit/<uuid_str:uuid>", methods=['GET', 'POST'])
     @login_optionally_required
     # https://stackoverflow.com/questions/42984453/wtforms-populate-form-with-data-if-data-exists
     # https://wtforms.readthedocs.io/en/3.0.x/forms/#wtforms.form.Form.populate_obj ?
     def edit_page(uuid):
         from changedetectionio import forms
-        from changedetectionio.blueprint.browser_steps.browser_steps import browser_step_ui_config
+        from changedetectionio.browser_steps.browser_steps import browser_step_ui_config
         from changedetectionio import processors
         import importlib
 
+        if uuid == 'first':
+            uuid = list(datastore.data['watching'].keys()).pop()
         # More for testing, possible to return the first/only
         if not datastore.data['watching'].keys():
             flash(gettext("No watches to edit"), "error")
             return redirect(url_for('watchlist.index'))
 
-        if uuid == 'first':
-            uuid = list(datastore.data['watching'].keys()).pop()
-
         if not uuid in datastore.data['watching']:
             flash(gettext("No watch with the UUID {} found.").format(uuid), "error")
             return redirect(url_for('watchlist.index'))
@@ -52,7 +51,13 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                     redirect(url_for('ui_edit.edit_page', uuid=uuid))
 
         # be sure we update with a copy instead of accidently editing the live object by reference
-        default = deepcopy(datastore.data['watching'][uuid])
+        default = None
+        while not default:
+            try:
+                default = deepcopy(datastore.data['watching'][uuid])
+            except RuntimeError as e:
+                # Dictionary changed
+                continue
 
         # Defaults for proxy choice
         if datastore.proxy_list is not None:  # When enabled
@@ -66,8 +71,13 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
         processor_name = datastore.data['watching'][uuid].get('processor', '')
         processor_classes = next((tpl for tpl in processors.find_processors() if tpl[1] == processor_name), None)
         if not processor_classes:
-            flash(gettext("Cannot load the edit form for processor/plugin '{}', plugin missing?").format(processor_classes[1]), 'error')
-            return redirect(url_for('watchlist.index'))
+            flash(gettext("Could not load '{}' processor, processor plugin might be missing. Please select a different processor.").format(processor_name), 'error')
+            # Fall back to default processor so user can still edit and change processor
+            processor_classes = next((tpl for tpl in processors.find_processors() if tpl[1] == 'text_json_diff'), None)
+            if not processor_classes:
+                # If even text_json_diff is missing, something is very wrong
+                flash(gettext("Could not load '{}' processor, processor plugin might be missing.").format(processor_name), 'error')
+                return redirect(url_for('watchlist.index'))
 
         parent_module = processors.get_parent_module(processor_classes[0])
 
@@ -107,12 +117,25 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                 processor_config = processor_instance.get_extra_watch_config(config_filename)
 
                 if processor_config:
+                    from wtforms.fields.form import FormField
                     # Populate processor-config-* fields from JSON
                     for config_key, config_value in processor_config.items():
-                        field_name = f'processor_config_{config_key}'
-                        if hasattr(form, field_name):
-                            getattr(form, field_name).data = config_value
-                            logger.debug(f"Loaded processor config from {config_filename}: {field_name} = {config_value}")
+                        if not isinstance(config_value, dict):
+                            continue
+                        # Try exact API-named field first (e.g., processor_config_restock_diff)
+                        target_field = getattr(form, f'processor_config_{config_key}', None)
+                        # Fallback: find any FormField sub-form whose fields cover config_value keys
+                        if target_field is None:
+                            for form_field in form:
+                                if isinstance(form_field, FormField) and all(k in form_field.form._fields for k in config_value):
+                                    target_field = form_field
+                                    break
+                        if target_field is not None:
+                            for sub_key, sub_value in config_value.items():
+                                sub_field = target_field.form._fields.get(sub_key)
+                                if sub_field is not None:
+                                    sub_field.data = sub_value
+                                    logger.debug(f"Loaded processor config from {config_filename}: {sub_key} = {sub_value}")
             except Exception as e:
                 logger.warning(f"Failed to load processor config: {e}")
 
@@ -144,58 +167,10 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
             extra_update_obj['time_between_check'] = form.time_between_check.data
 
-             # Handle processor-config-* fields separately (save to JSON, not datastore)
-            processor_config_data = {}
-            fields_to_remove = []
-            for field_name, field_value in form.data.items():
-                if field_name.startswith('processor_config_'):
-                    config_key = field_name.replace('processor_config_', '')
-                    if field_value:  # Only save non-empty values
-                        processor_config_data[config_key] = field_value
-                    fields_to_remove.append(field_name)
-
-            # Save processor config to JSON file if any config data exists
-            if processor_config_data:
-                try:
-                    processor_name = form.data.get('processor')
-                    # Create a processor instance to access config methods
-                    processor_instance = processors.difference_detection_processor(datastore, uuid)
-                    # Use processor name as filename so each processor keeps its own config
-                    config_filename = f'{processor_name}.json'
-                    processor_instance.update_extra_watch_config(config_filename, processor_config_data)
-                    logger.debug(f"Saved processor config to {config_filename}: {processor_config_data}")
-
-                    # Call optional edit_hook if processor has one
-                    try:
-                        # Try to import the edit_hook module from the processor package
-                        import importlib
-                        edit_hook_module_name = f'changedetectionio.processors.{processor_name}.edit_hook'
-
-                        try:
-                            edit_hook = importlib.import_module(edit_hook_module_name)
-                            logger.debug(f"Found edit_hook module for {processor_name}")
-
-                            if hasattr(edit_hook, 'on_config_save'):
-                                logger.info(f"Calling edit_hook.on_config_save for {processor_name}")
-                                watch_obj = datastore.data['watching'][uuid]
-                                # Call hook and get updated config
-                                updated_config = edit_hook.on_config_save(watch_obj, processor_config_data, datastore)
-                                # Save updated config back to file
-                                processor_instance.update_extra_watch_config(config_filename, updated_config)
-                                logger.info(f"Edit hook updated config: {updated_config}")
-                            else:
-                                logger.debug(f"Edit hook module found but no on_config_save function")
-                        except ModuleNotFoundError:
-                            logger.debug(f"No edit_hook module for processor {processor_name} (this is normal)")
-                    except Exception as hook_error:
-                        logger.error(f"Edit hook error (non-fatal): {hook_error}", exc_info=True)
-
-                except Exception as e:
-                    logger.error(f"Failed to save processor config: {e}")
-
-            # Remove processor-config-* fields from form.data before updating datastore
-            for field_name in fields_to_remove:
-                form.data.pop(field_name, None)
+            # Handle processor-config-* fields separately (save to JSON, not datastore)
+            # IMPORTANT: These must NOT be saved to url-watches.json, only to the processor-specific JSON file
+            processor_config_data = processors.extract_processor_config_from_form_data(form.data)
+            processors.save_processor_config(datastore, uuid, processor_config_data)
 
             # Ignore text
             form_ignore_text = form.ignore_text.data
@@ -235,12 +210,19 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
             # Recast it if need be to right data Watch handler
             watch_class = processors.get_custom_watch_obj_for_processor(form.data.get('processor'))
-            datastore.data['watching'][uuid] = watch_class(datastore_path=datastore.datastore_path, default=datastore.data['watching'][uuid])
+            datastore.data['watching'][uuid] = watch_class(datastore_path=datastore.datastore_path, __datastore=datastore.data, default=datastore.data['watching'][uuid])
+
+            # Save the watch immediately
+            datastore.data['watching'][uuid].commit()
+
             flash(gettext("Updated watch - unpaused!") if request.args.get('unpause_on_save') else gettext("Updated watch."))
 
-            # Re #286 - We wait for syncing new data to disk in another thread every 60 seconds
-            # But in the case something is added we should save straight away
-            datastore.needs_write_urgent = True
+            # Cleanup any browsersteps session for this watch
+            try:
+                from changedetectionio.blueprint.browser_steps import cleanup_session_for_watch
+                cleanup_session_for_watch(uuid)
+            except Exception as e:
+                logger.debug(f"Error cleaning up browsersteps session: {e}")
 
             # Do not queue on edit if its not within the time range
 
@@ -270,7 +252,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             #############################
             if not datastore.data['watching'][uuid].get('paused') and is_in_schedule:
                 # Queue the watch for immediate recheck, with a higher priority
-                worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
 
             # Diff page [edit] link should go back to diff page
             if request.args.get("next") and request.args.get("next") == 'diff':
@@ -298,10 +280,17 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
             # Get fetcher capabilities instead of hardcoded logic
             capabilities = get_fetcher_capabilities(watch, datastore)
+
+            # Add processor capabilities from module
+            capabilities['supports_visual_selector'] = getattr(parent_module, 'supports_visual_selector', False)
+            capabilities['supports_text_filters_and_triggers'] = getattr(parent_module, 'supports_text_filters_and_triggers', False)
+            capabilities['supports_text_filters_and_triggers_elements'] = getattr(parent_module, 'supports_text_filters_and_triggers_elements', False)
+            capabilities['supports_request_type'] = getattr(parent_module, 'supports_request_type', False)
+
             app_rss_token = datastore.data['settings']['application'].get('rss_access_token'),
 
             c = [f"processor-{watch.get('processor')}"]
-            if worker_handler.is_watch_running(uuid):
+            if worker_pool.is_watch_running(uuid):
                 c.append('checking-now')
 
             template_args = {
@@ -325,8 +314,6 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                     'url': url_for('rss.rss_single_watch', uuid=watch['uuid'], token=app_rss_token)
                 },
                 'settings_application': datastore.data['settings']['application'],
-                'system_has_playwright_configured': os.getenv('PLAYWRIGHT_DRIVER_URL'),
-                'system_has_webdriver_configured': os.getenv('WEBDRIVER_URL'),
                 'ui_edit_stats_extras': collect_ui_edit_stats_extras(watch),
                 'visual_selector_data_ready': datastore.visualselector_data_is_ready(watch_uuid=uuid),
                 'timezone_default_config': datastore.data['settings']['application'].get('scheduler_timezone_default'),
@@ -353,17 +340,19 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
         return output
 
-    @edit_blueprint.route("/edit/<string:uuid>/get-html", methods=['GET'])
+    @edit_blueprint.route("/edit/<uuid_str:uuid>/get-html", methods=['GET'])
     @login_optionally_required
     def watch_get_latest_html(uuid):
         from io import BytesIO
         from flask import send_file
         import brotli
 
+        if uuid == 'first':
+            uuid = list(datastore.data['watching'].keys()).pop()
         watch = datastore.data['watching'].get(uuid)
-        if watch and watch.history.keys() and os.path.isdir(watch.watch_data_dir):
+        if watch and watch.history.keys() and os.path.isdir(watch.data_dir):
             latest_filename = list(watch.history.keys())[-1]
-            html_fname = os.path.join(watch.watch_data_dir, f"{latest_filename}.html.br")
+            html_fname = os.path.join(watch.data_dir, f"{latest_filename}.html.br")
             with open(html_fname, 'rb') as f:
                 if html_fname.endswith('.br'):
                     # Read and decompress the Brotli file
@@ -378,12 +367,65 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
         # Return a 500 error
         abort(500)
 
+    @edit_blueprint.route("/edit/<uuid_str:uuid>/get-data-package", methods=['GET'])
+    @login_optionally_required
+    def watch_get_data_package(uuid):
+        """Download all data for a single watch as a zip file"""
+        from io import BytesIO
+        from flask import send_file
+        import zipfile
+        from pathlib import Path
+        import datetime
+
+        watch = datastore.data['watching'].get(uuid)
+        if not watch:
+            abort(404)
+
+        # Create zip in memory
+        memory_file = BytesIO()
+
+        with zipfile.ZipFile(memory_file, 'w',
+                           compression=zipfile.ZIP_DEFLATED,
+                           compresslevel=8) as zipObj:
+
+            # Add the watch's JSON file if it exists
+            watch_json_path = os.path.join(watch.data_dir, 'watch.json')
+            if os.path.isfile(watch_json_path):
+                zipObj.write(watch_json_path,
+                           arcname=os.path.join(uuid, 'watch.json'),
+                           compress_type=zipfile.ZIP_DEFLATED,
+                           compresslevel=8)
+
+            # Add all files in the watch data directory
+            if os.path.isdir(watch.data_dir):
+                for f in Path(watch.data_dir).glob('*'):
+                    if f.is_file() and f.name != 'watch.json':  # Skip watch.json since we already added it
+                        zipObj.write(f,
+                                   arcname=os.path.join(uuid, f.name),
+                                   compress_type=zipfile.ZIP_DEFLATED,
+                                   compresslevel=8)
+
+        # Seek to beginning of file
+        memory_file.seek(0)
+
+        # Generate filename with timestamp
+        timestamp = datetime.datetime.now().strftime("%Y%m%d%H%M%S")
+        filename = f"watch-data-{uuid[:8]}-{timestamp}.zip"
+
+        return send_file(memory_file,
+                        as_attachment=True,
+                        download_name=filename,
+                        mimetype='application/zip')
+
     # Ajax callback
-    @edit_blueprint.route("/edit/<string:uuid>/preview-rendered", methods=['POST'])
+    @edit_blueprint.route("/edit/<uuid_str:uuid>/preview-rendered", methods=['POST'])
     @login_optionally_required
     def watch_get_preview_rendered(uuid):
         '''For when viewing the "preview" of the rendered text from inside of Edit'''
         from flask import jsonify
+
+        if uuid == 'first':
+            uuid = list(datastore.data['watching'].keys()).pop()
         from changedetectionio.processors.text_json_diff import prepare_filter_prevew
         result = prepare_filter_prevew(watch_uuid=uuid, form_data=request.form, datastore=datastore)
         return jsonify(result)
@@ -407,6 +449,9 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                     s = re.sub(r'[0-9]+', r'\\d+', s)
                     datastore.data["watching"][uuid]['ignore_text'].append('/' + s + '/')
 
+            # Save the updated ignore_text
+            datastore.data["watching"][uuid].commit()
+
         return f"<a href={url_for('ui.ui_preview.preview_page', uuid=uuid)}>Click to preview</a>"
     
     return edit_blueprint

changedetectionio/blueprint/ui/preview.py

@@ -10,7 +10,7 @@ from changedetectionio import html_tools
 def construct_blueprint(datastore: ChangeDetectionStore):
     preview_blueprint = Blueprint('ui_preview', __name__, template_folder="../ui/templates")
 
-    @preview_blueprint.route("/preview/<string:uuid>", methods=['GET'])
+    @preview_blueprint.route("/preview/<uuid_str:uuid>", methods=['GET'])
     @login_optionally_required
     def preview_page(uuid):
         """
@@ -26,10 +26,9 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         Each processor implements processors/{type}/preview.py::render()
         If a processor doesn't have a preview module, falls back to default text preview.
         """
-        # More for testing, possible to return the first/only
+
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
-
         try:
             watch = datastore.data['watching'][uuid]
         except KeyError:
@@ -39,24 +38,21 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's preview module
-            import importlib
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.preview')
+        # Try to get the processor's preview module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'preview')
 
-            # Call the processor's render() function
-            if hasattr(processor_module, 'render'):
-                return processor_module.render(
-                    watch=watch,
-                    datastore=datastore,
-                    request=request,
-                    url_for=url_for,
-                    render_template=render_template,
-                    flash=flash,
-                    redirect=redirect
-                )
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.debug(f"Processor {processor_name} does not have a preview module, using default preview: {e}")
+        # Call the processor's render() function
+        if processor_module and hasattr(processor_module, 'render'):
+            return processor_module.render(
+                watch=watch,
+                datastore=datastore,
+                request=request,
+                url_for=url_for,
+                render_template=render_template,
+                flash=flash,
+                redirect=redirect
+            )
 
         # Fallback: if processor doesn't have preview module, use default text preview
         content = []
@@ -129,7 +125,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
         return output
 
-    @preview_blueprint.route("/preview/<string:uuid>/processor-asset/<string:asset_name>", methods=['GET'])
+    @preview_blueprint.route("/preview/<uuid_str:uuid>/processor-asset/<string:asset_name>", methods=['GET'])
     @login_optionally_required
     def processor_asset(uuid, asset_name):
         """
@@ -150,10 +146,8 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         """
         from flask import make_response
 
-        # More for testing, possible to return the first/only
         if uuid == 'first':
             uuid = list(datastore.data['watching'].keys()).pop()
-
         try:
             watch = datastore.data['watching'][uuid]
         except KeyError:
@@ -163,39 +157,33 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Get the processor type for this watch
         processor_name = watch.get('processor', 'text_json_diff')
 
-        try:
-            # Try to import the processor's preview module
-            import importlib
-            processor_module = importlib.import_module(f'changedetectionio.processors.{processor_name}.preview')
+        # Try to get the processor's preview module (works for both built-in and plugin processors)
+        from changedetectionio.processors import get_processor_submodule
+        processor_module = get_processor_submodule(processor_name, 'preview')
 
-            # Call the processor's get_asset() function
-            if hasattr(processor_module, 'get_asset'):
-                result = processor_module.get_asset(
-                    asset_name=asset_name,
-                    watch=watch,
-                    datastore=datastore,
-                    request=request
-                )
+        # Call the processor's get_asset() function
+        if processor_module and hasattr(processor_module, 'get_asset'):
+            result = processor_module.get_asset(
+                asset_name=asset_name,
+                watch=watch,
+                datastore=datastore,
+                request=request
+            )
 
-                if result is None:
-                    from flask import abort
-                    abort(404, description=f"Asset '{asset_name}' not found")
-
-                binary_data, content_type, cache_control = result
-
-                response = make_response(binary_data)
-                response.headers['Content-Type'] = content_type
-                if cache_control:
-                    response.headers['Cache-Control'] = cache_control
-                return response
-            else:
-                logger.warning(f"Processor {processor_name} does not implement get_asset()")
+            if result is None:
                 from flask import abort
-                abort(404, description=f"Processor '{processor_name}' does not support assets")
+                abort(404, description=f"Asset '{asset_name}' not found")
 
-        except (ImportError, ModuleNotFoundError) as e:
-            logger.warning(f"Processor {processor_name} does not have a preview module: {e}")
+            binary_data, content_type, cache_control = result
+
+            response = make_response(binary_data)
+            response.headers['Content-Type'] = content_type
+            if cache_control:
+                response.headers['Cache-Control'] = cache_control
+            return response
+        else:
+            logger.warning(f"Processor {processor_name} does not implement get_asset()")
             from flask import abort
-            abort(404, description=f"Processor '{processor_name}' not found")
+            abort(404, description=f"Processor '{processor_name}' does not support assets")
 
     return preview_blueprint

changedetectionio/blueprint/ui/templates/edit.html

@@ -45,14 +45,19 @@
     <div class="tabs collapsable">
         <ul>
             <li class="tab"><a href="#general">{{ _('General') }}</a></li>
+            {% if capabilities.supports_request_type %}
             <li class="tab"><a href="#request">{{ _('Request') }}</a></li>
+            {% endif %}
             {% if extra_tab_content %}
             <li class="tab"><a href="#extras_tab">{{ extra_tab_content }}</a></li>
             {% endif %}
+            {% if capabilities.supports_browser_steps %}
             <li class="tab"><a id="browsersteps-tab" href="#browser-steps">{{ _('Browser Steps') }}</a></li>
-        <!-- should goto extra forms? -->
-            {% if watch['processor'] == 'text_json_diff' or watch['processor'] == 'image_ssim_diff' %}
+            {% endif %}
+            {% if capabilities.supports_visual_selector %}
             <li class="tab"><a id="visualselector-tab" href="#visualselector">{{ _('Visual Filter Selector') }}</a></li>
+            {% endif %}
+            {% if capabilities.supports_text_filters_and_triggers %}
             <li class="tab" id="filters-and-triggers-tab"><a href="#filters-and-triggers">{{ _('Filters & Triggers') }}</a></li>
             <li class="tab" id="conditions-tab"><a href="#conditions">{{ _('Conditions') }}</a></li>
             {% endif %}
@@ -110,12 +115,20 @@
                          {{ _('Sends a notification when the filter can no longer be seen on the page, good for knowing when the page changed and your filter will not work anymore.') }}
                         </span>
                     </div>
+                    <div class="pure-control-group">
+                        {{ render_field(form.history_snapshot_max_length, class="history_snapshot_max_length") }}
+                        <span class="pure-form-message-inline">{{ _('Limit collection of history snapshots for each watch to this number of history items.') }}
+                        <br>
+                        {{ _('Set to empty to use system settings default') }}
+                        </span>
+                    </div>
                     <div class="pure-control-group">
                         {{ render_ternary_field(form.use_page_title_in_list) }}
                     </div>
                 </fieldset>
             </div>
 
+            {% if capabilities.supports_request_type %}
             <div class="tab-pane-inner" id="request">
                     <div class="pure-control-group inline-radio">
                         {{ render_field(form.fetch_backend, class="fetch-backend") }}
@@ -203,10 +216,11 @@ Math: {{ 1 + 1 }}") }}
                     </div>
             </fieldset>
             </div>
+            {% endif %}
 
             <div class="tab-pane-inner" id="browser-steps">
             {% if capabilities.supports_browser_steps %}
-               {% if visual_selector_data_ready %}
+               {% if true %}
                 <img class="beta-logo" src="{{url_for('static_content', group='images', filename='beta-logo.png')}}" alt="New beta functionality">
                 <fieldset>
                     <div class="pure-control-group">
@@ -283,8 +297,7 @@ Math: {{ 1 + 1 }}") }}
                 </fieldset>
             </div>
 
-            {% if watch['processor'] == 'text_json_diff' or watch['processor'] == 'image_ssim_diff' %}
-
+            {% if capabilities.supports_text_filters_and_triggers %}
             <div class="tab-pane-inner" id="conditions">
                     <script>
                         const verify_condition_rule_url="{{url_for('conditions.verify_condition_single_rule', watch_uuid=uuid)}}";
@@ -303,7 +316,9 @@ Math: {{ 1 + 1 }}") }}
                 <span id="activate-text-preview" class="pure-button pure-button-primary button-xsmall">{{ _('Activate preview') }}</span>
               <div>
               <div id="edit-text-filter">
-                    <div class="pure-control-group" id="pro-tips">
+
+{% if capabilities.supports_text_filters_and_triggers_elements %}
+                        <div class="pure-control-group" id="pro-tips">
                             <strong>{{ _('Pro-tips:') }}</strong><br>
                             <ul>
                                 <li>
@@ -314,8 +329,8 @@ Math: {{ 1 + 1 }}") }}
                                 </li>
                             </ul>
                     </div>
-
 {% include "edit/include_subtract.html" %}
+{% endif %}
                 <div class="text-filtering border-fieldset">
                 <fieldset class="pure-group" id="text-filtering-type-options">
                     <h3>{{ _('Text filtering') }}</h3>
@@ -374,7 +389,7 @@ Math: {{ 1 + 1 }}") }}
             {{ extra_form_content|safe }}
             </div>
         {% endif %}
-            {% if watch['processor'] == 'text_json_diff' or watch['processor'] == 'image_ssim_diff' %}
+            {% if capabilities.supports_visual_selector %}
             <div class="tab-pane-inner visual-selector-ui" id="visualselector">
                 <img class="beta-logo" src="{{url_for('static_content', group='images', filename='beta-logo.png')}}" alt="New beta functionality">
 
@@ -386,7 +401,7 @@ Math: {{ 1 + 1 }}") }}
                                     {{ _('The Visual Selector tool lets you select the') }} <i>{{ _('text') }}</i> {{ _('elements that will be used for the change detection. It automatically fills-in the filters in the "CSS/JSONPath/JQ/XPath Filters" box of the') }} <a href="#filters-and-triggers">{{ _('Filters & Triggers') }}</a> {{ _('tab. Use') }} <strong>{{ _('Shift+Click') }}</strong> {{ _('to select multiple items.') }}
                                 </span>
 
-                                {% if watch['processor'] == 'image_ssim_diff' %}
+                                {% if watch['processor'] == 'image_ssim_diff' %} {# @todo, integrate with image_ssim_diff selector better, use some extra form ? #}
                                 <div id="selection-mode-controls" style="margin: 10px 0; padding: 10px; background: var(--color-background-tab); border-radius: 5px;">
                                     <label style="font-weight: 600; margin-right: 15px;">{{ _('Selection Mode:') }}</label>
                                     <label style="margin-right: 15px;">
@@ -473,6 +488,7 @@ Math: {{ 1 + 1 }}") }}
                     {% if watch.history_n %}
                         <p>
                              <a href="{{url_for('ui.ui_edit.watch_get_latest_html', uuid=uuid)}}" class="pure-button button-small">{{ _('Download latest HTML snapshot') }}</a>
+                             <a href="{{url_for('ui.ui_edit.watch_get_data_package', uuid=uuid)}}" class="pure-button button-small">{{ _('Download watch data package') }}</a>
                         </p>
                     {% endif %}
 

changedetectionio/blueprint/ui/views.py

@@ -2,7 +2,7 @@ from flask import Blueprint, request, redirect, url_for, flash
 from flask_babel import gettext
 from changedetectionio.store import ChangeDetectionStore
 from changedetectionio.auth_decorator import login_optionally_required
-from changedetectionio import worker_handler
+from changedetectionio import worker_pool
 
 
 def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMetaData, watch_check_update):
@@ -24,8 +24,9 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             flash(gettext('Warning, URL {} already exists').format(url), "notice")
 
         add_paused = request.form.get('edit_and_watch_submit_button') != None
-        processor = request.form.get('processor', 'text_json_diff')
-        new_uuid = datastore.add_watch(url=url, tag=request.form.get('tags').strip(), extras={'paused': add_paused, 'processor': processor})
+        from changedetectionio import processors
+        processor = request.form.get('processor', processors.get_default_processor())
+        new_uuid = datastore.add_watch(url=url, tag=request.form.get('tags','').strip(), extras={'paused': add_paused, 'processor': processor})
 
         if new_uuid:
             if add_paused:
@@ -33,9 +34,9 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                 return redirect(url_for('ui.ui_edit.edit_page', uuid=new_uuid, unpause_on_save=1, tag=request.args.get('tag')))
             else:
                 # Straight into the queue.
-                worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': new_uuid}))
+                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': new_uuid}))
                 flash(gettext("Watch added."))
 
         return redirect(url_for('watchlist.index', tag=request.args.get('tag','')))
 
-    return views_blueprint
+    return views_blueprint

changedetectionio/blueprint/watchlist/__init__.py

@@ -2,8 +2,8 @@ import os
 import time
 
 from flask import Blueprint, request, make_response, render_template, redirect, url_for, flash, session
-from flask_login import current_user
 from flask_paginate import Pagination, get_page_parameter
+from flask_babel import gettext as _
 
 from changedetectionio import forms
 from changedetectionio import processors
@@ -39,7 +39,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             elif op == 'mute':
                 datastore.data['watching'][uuid].toggle_mute()
 
-            datastore.needs_write = True
+            datastore.data['watching'][uuid].commit()
             return redirect(url_for('watchlist.index', tag = active_tag_uuid))
 
         # Sort by last_changed and add the uuid which is usually the key..
@@ -74,10 +74,14 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
         pagination = Pagination(page=page,
                                 total=total_count,
-                                per_page=datastore.data['settings']['application'].get('pager_size', 50), css_framework="semantic")
+                                per_page=datastore.data['settings']['application'].get('pager_size', 50),
+                                css_framework="semantic",
+                                display_msg=_('displaying <b>{start} - {end}</b> {record_name} in total <b>{total}</b>'),
+                                record_name=_('records'))
 
         sorted_tags = sorted(datastore.data['settings']['application'].get('tags').items(), key=lambda x: x[1]['title'])
 
+        proxy_list = datastore.proxy_list
         output = render_template(
             "watch-overview.html",
             active_tag=active_tag,
@@ -85,17 +89,19 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             app_rss_token=datastore.data['settings']['application'].get('rss_access_token'),
             datastore=datastore,
             errored_count=errored_count,
+            extra_classes='has-queue' if not update_q.empty() else '',
             form=form,
             generate_tag_colors=processors.generate_processor_badge_colors,
             guid=datastore.data['app_guid'],
-            has_proxies=datastore.proxy_list,
+            has_proxies=proxy_list,
             hosted_sticky=os.getenv("SALTED_PASS", False) == False,
             now_time_server=round(time.time()),
             pagination=pagination,
+            processor_badge_css=processors.get_processor_badge_css(),
             processor_badge_texts=processors.get_processor_badge_texts(),
             processor_descriptions=processors.get_processor_descriptions(),
-            processor_badge_css=processors.get_processor_badge_css(),
-            queued_uuids=[q_uuid.item['uuid'] for q_uuid in update_q.queue],
+            queue_size=update_q.qsize(),
+            queued_uuids=update_q.get_queued_uuids(),
             search_q=request.args.get('q', '').strip(),
             sort_attribute=request.args.get('sort') if request.args.get('sort') else request.cookies.get('sort'),
             sort_order=request.args.get('order') if request.args.get('order') else request.cookies.get('order'),
@@ -105,6 +111,16 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             watches=sorted_watches
         )
 
+        # Return freed template-building memory to the OS immediately.
+        # render_template allocates ~20MB of intermediate strings that are freed on return,
+        # but glibc keeps those pages mapped in its arenas as RSS. malloc_trim() forces
+        # glibc to release them, preventing RSS growth from concurrent Chrome connections.
+        try:
+            import ctypes
+            ctypes.CDLL('libc.so.6').malloc_trim(0)
+        except Exception:
+            pass
+
         if session.get('share-link'):
             del (session['share-link'])
 

changedetectionio/blueprint/watchlist/templates/watch-overview.html

@@ -1,5 +1,9 @@
 {%- extends 'base.html' -%}
 {%- block content -%}
+{%- set tips = [
+    _("Changedetection.io can monitor more than just web-pages! See our plugins!") ~ ' <a href="https://changedetection.io/plugins">' ~ _('More info') ~ '</a>',
+    _("You can also add 'shared' watches.") ~ ' <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Sharing-a-Watch">' ~ _('More info') ~ '</a>'
+] -%}
 {%- from '_helpers.html' import render_simple_field, render_field, render_nolabel_field, sort_by_title -%}
 <script src="{{url_for('static_content', group='js', filename='jquery-3.6.0.min.js')}}"></script>
 <script src="{{url_for('static_content', group='js', filename='watch-overview.js')}}" defer></script>
@@ -10,6 +14,46 @@
 // Initialize Feather icons after the page loads
 document.addEventListener('DOMContentLoaded', function() {
     feather.replace();
+
+    // Intersection Observer for lazy loading favicons
+    // Only load favicon images when they enter the viewport
+    if ('IntersectionObserver' in window) {
+        const faviconObserver = new IntersectionObserver((entries, observer) => {
+            entries.forEach(entry => {
+                if (entry.isIntersecting) {
+                    const img = entry.target;
+                    const src = img.getAttribute('data-src');
+
+                    if (src) {
+                        // Load the actual favicon
+                        img.src = src;
+                        img.removeAttribute('data-src');
+                    }
+
+                    // Stop observing this image
+                    observer.unobserve(img);
+                }
+            });
+        }, {
+            // Start loading slightly before the image enters viewport
+            rootMargin: '50px',
+            threshold: 0.01
+        });
+
+        // Observe all lazy favicon images
+        document.querySelectorAll('.lazy-favicon').forEach(img => {
+            faviconObserver.observe(img);
+        });
+    } else {
+        // Fallback for older browsers: load all favicons immediately
+        document.querySelectorAll('.lazy-favicon').forEach(img => {
+            const src = img.getAttribute('data-src');
+            if (src) {
+                img.src = src;
+                img.removeAttribute('data-src');
+            }
+        });
+    }
 });
 </script>
 <style>
@@ -62,14 +106,16 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
                     {{ render_nolabel_field(form.edit_and_watch_submit_button, title=_("Edit first then Watch") ) }}
             </div>
             <div id="watch-group-tag">
-               {{ render_field(form.tags, value=active_tag.title if active_tag_uuid else '', placeholder="Watch group / tag", class="transparent-field") }}
+               {{ render_field(form.tags, value=active_tag.title if active_tag_uuid else '', placeholder=_("Watch group / tag"), class="transparent-field") }}
             </div>
             <div id="quick-watch-processor-type">
                 {{ render_simple_field(form.processor) }}
             </div>
 
         </fieldset>
-        <span style="color:#eee; font-size: 80%;"><img alt="{{ _('Create a shareable link') }}" style="height: 1em;display:inline-block;" src="{{url_for('static_content', group='images', filename='spread-white.svg')}}" > {{ _("Tip: You can also add 'shared' watches.") }} <a href="https://github.com/dgtlmoon/changedetection.io/wiki/Sharing-a-Watch">{{ _('More info') }}</a></span>
+        <span style="color:#eee; font-size: 80%;">
+            <strong>Tip: </strong> {{ tips | random | safe }}
+        </span>
     </form>
 </div>
 <div class="box">
@@ -99,9 +145,14 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
                        data-confirm-message="{{ _('<p>Are you sure you want to delete the selected watches?</strong></p><p>This action cannot be undone.</p>') }}"
                        data-confirm-button="{{ _('Delete') }}"><i data-feather="trash" style="width: 14px; height: 14px; stroke: white; margin-right: 4px;"></i>{{ _('Delete') }}</button>
     </div>
-    {%- if watches|length >= pagination.per_page -%}
-        {{ pagination.info }}
-    {%- endif -%}
+
+    <div id="stats_row">
+        <div class="left">{%- if watches|length >= pagination.per_page -%}{{ pagination.info }}{%- endif -%}</div>
+        <div class="right" >{{ _('Queued size') }}: <span id="queue-size-int">{{ queue_size }}</span></div>
+    </div>
+
+
+
     {%- if search_q -%}<div id="search-result-info">{{ _('Searching') }} "<strong><i>{{search_q}}</i></strong>"</div>{%- endif -%}
     <div>
         <a href="{{url_for('watchlist.index')}}" class="pure-button button-tag {{'active' if not active_tag_uuid }}">{{ _('All') }}</a>
@@ -154,7 +205,7 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
             <tbody>
             {%- if not watches|length -%}
             <tr>
-                <td colspan="{{ cols_required }}" style="text-wrap: wrap;">{{ _('No website watches configured, please add a URL in the box above, or') }} <a href="{{ url_for('imports.import_page')}}" >{{ _('import a list') }}</a>.</td>
+                <td colspan="{{ cols_required }}" style="text-wrap: wrap;">{{ _('No web page change detection watches configured, please add a URL in the box above, or') }} <a href="{{ url_for('imports.import_page')}}" >{{ _('import a list') }}</a>.</td>
             </tr>
             {%- endif -%}
 
@@ -162,12 +213,13 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
                 {%- set checking_now = is_checking_now(watch) -%}
                 {%- set history_n = watch.history_n -%}
                 {%- set favicon = watch.get_favicon_filename() -%}
+                {%- set error_texts = watch.compile_error_texts(has_proxies=has_proxies) -%}
                 {%- set system_use_url_watchlist = datastore.data['settings']['application']['ui'].get('use_page_title_in_list')  -%}
                 {#  Class settings mirrored in changedetectionio/static/js/realtime.js for the frontend #}
                 {%- set row_classes = [
                     loop.cycle('pure-table-odd', 'pure-table-even'),
                     'processor-' ~ watch['processor'],
-                    'has-error' if watch.compile_error_texts()|length > 2 else '',
+                    'has-error' if error_texts|length > 2 else '',
                     'paused' if watch.paused is defined and watch.paused != False else '',
                     'unviewed' if watch.has_unviewed else '',
                     'has-restock-info' if watch.has_restock_info else 'no-restock-info',
@@ -195,30 +247,40 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
                 <td class="title-col inline">
                     <div class="flex-wrapper">
                     {% if 'favicons_enabled' not in ui_settings or ui_settings['favicons_enabled'] %}
-                        <div>{# A page might have hundreds of these images, set IMG options for lazy loading, don't set SRC if we dont have it so it doesnt fetch the placeholder'  #}
-                            <img alt="Favicon thumbnail" class="favicon" loading="lazy" decoding="async" fetchpriority="low" {% if favicon %} src="{{url_for('static_content', group='favicon', filename=watch.uuid)}}" {% else %} src='data:image/svg+xml;utf8,%3Csvg xmlns="http://www.w3.org/2000/svg" width="7.087" height="7.087" viewBox="0 0 7.087 7.087"%3E%3Ccircle cx="3.543" cy="3.543" r="3.279" stroke="%23e1e1e1" stroke-width="0.45" fill="none" opacity="0.74"/%3E%3C/svg%3E' {%  endif %} >
+                        <div>
+                            {# Intersection Observer lazy loading: store real URL in data-src, load only when visible in viewport #}
+                            <img alt="Favicon thumbnail"
+                                 class="favicon lazy-favicon"
+                                 loading="lazy"
+                                 decoding="async"
+                                 fetchpriority="low"
+                                 {% if favicon %}
+                                 data-src="{{url_for('static_content', group='favicon', filename=watch.uuid)}}"
+                                 {% endif %}
+                                 src='data:image/svg+xml;utf8,%3Csvg xmlns="http://www.w3.org/2000/svg" width="7.087" height="7.087" viewBox="0 0 7.087 7.087"%3E%3Ccircle cx="3.543" cy="3.543" r="3.279" stroke="%23e1e1e1" stroke-width="0.45" fill="none" opacity="0.74"/%3E%3C/svg%3E'>
                         </div>
                     {%  endif %}
                         <div>
-                        <span class="watch-title">
-                            {% if system_use_url_watchlist or watch.get('use_page_title_in_list') %}
-                                {{ watch.label }}
-                            {% else %}
-                                {{ watch.get('title') or watch.link }}
-                            {% endif %}
-                           <a class="external" target="_blank" rel="noopener" href="{{ watch.link.replace('source:','') }}">&nbsp;</a>
-                        </span>
-                            <div class="error-text" style="display:none;">{{ watch.compile_error_texts(has_proxies=datastore.proxy_list)|safe }}</div>
+                            {%- if watch['processor'] and watch['processor'] in processor_badge_texts -%}
+                                <span class="processor-badge processor-badge-{{ watch['processor'] }}" title="{{ processor_descriptions.get(watch['processor'], watch['processor']) }}">{{ processor_badge_texts[watch['processor']] }}</span>
+                            {%- endif -%}
+                            <span class="watch-title">
+                                {% if system_use_url_watchlist or watch.get('use_page_title_in_list') %}
+                                    {{ watch.label }}
+                                {% else %}
+                                    {{ watch.get('title') or watch.link }}
+                                {% endif %}
+                               <a class="external" target="_blank" rel="noopener" href="{{ watch.link.replace('source:','') }}">&nbsp;</a>
+                            </span>
+                            <div class="error-text" style="display:none;">{{ error_texts|safe }}</div>
                             {%- if watch['processor'] == 'text_json_diff'  -%}
                                 {%- if watch['has_ldjson_price_data'] and not watch['track_ldjson_price_data']  -%}
                                 <div class="ldjson-price-track-offer">Switch to Restock & Price watch mode? <a href="{{url_for('price_data_follower.accept', uuid=watch.uuid)}}" class="pure-button button-xsmall">Yes</a> <a href="{{url_for('price_data_follower.reject', uuid=watch.uuid)}}" class="">No</a></div>
                                 {%- endif -%}
                             {%- endif -%}
-                            {%- if watch['processor'] and watch['processor'] in processor_badge_texts -%}
-                                <span class="processor-badge processor-badge-{{ watch['processor'] }}" title="{{ processor_descriptions.get(watch['processor'], watch['processor']) }}">{{ processor_badge_texts[watch['processor']] }}</span>
-                            {%- endif -%}
+
                             {%- for watch_tag_uuid, watch_tag in datastore.get_all_tags_for_watch(watch['uuid']).items() -%}
-                              <span class="watch-tag-list tag-{{ watch_tag.title|sanitize_tag_class }}">{{ watch_tag.title }}</span>
+                              <a href="{{url_for('watchlist.index', tag=watch_tag_uuid) }}" class="watch-tag-list tag-{{ watch_tag.title|sanitize_tag_class }}">{{ watch_tag.title }}</a>
                             {%- endfor -%}
                         </div>
                     <div class="status-icons">
@@ -243,11 +305,20 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
                             </span>
                         {%- endif -%}
 
-                        {%- if watch.get('restock') and watch['restock']['price'] != None -%}
-                            {%- if watch['restock']['price'] != None -%}
-                                <span class="restock-label price" title="{{ _('Price') }}">
-                                {{ watch['restock']['price']|format_number_locale if watch['restock'].get('price') else '' }} {{ watch['restock'].get('currency','') }}
-                                </span>
+                        {%- if watch.get('restock') and watch['restock'].get('price') -%}
+                            {%- set restock = watch['restock'] -%}
+                            {%- set price = restock.get('price') -%}
+                            {%- set cur = restock.get('currency','') -%}
+
+                            {%- if price is not none and (price|string)|regex_search('\d') -%}
+                              <span class="restock-label price" title="{{ _('Price') }}">
+                              {# @todo: make parse_currency/parse_decimal aware of the locale of the actual web page and use that instead changedetectionio/processors/restock_diff/__init__.py #}
+                                {%- if price is number -%}{# It's a number so we can convert it to their locale' #}
+                                  {{ price|format_number_locale }} {{ cur }}<!-- as number -->
+                                {%- else -%}{# It's totally fine if it arrives as something else, the website might be something weird in this field #}
+                                  {{ price }} {{ cur }}<!-- as string -->
+                                {%- endif -%}
+                              </span>
                             {%- endif -%}
                         {%- elif not watch.has_restock_info -%}
                             <span class="restock-label error">{{ _('No information') }}</span>

changedetectionio/browser_steps/browser_steps.py

@@ -8,6 +8,17 @@ from changedetectionio.content_fetchers import SCREENSHOT_MAX_HEIGHT_DEFAULT
 from changedetectionio.content_fetchers.base import manage_user_agent
 from changedetectionio.jinja2_custom import render as jinja_render
 
+def browser_steps_get_valid_steps(browser_steps: list):
+    if browser_steps is not None and len(browser_steps):
+        valid_steps = list(filter(
+            lambda s: (s['operation'] and len(s['operation']) and s['operation'] != 'Choose one'),browser_steps))
+
+        # Just incase they selected Goto site by accident with older JS
+        if valid_steps and valid_steps[0]['operation'] == 'Goto site':
+            del(valid_steps[0])
+
+        return valid_steps
+    return []
 
 
 # Two flags, tell the JS which of the "Selector" or "Value" field should be enabled in the front end

changedetectionio/content_fetchers/base.py

@@ -38,7 +38,6 @@ def manage_user_agent(headers, current_ua=''):
 
     return None
 
-
 class Fetcher():
     browser_connection_is_custom = None
     browser_connection_url = None
@@ -71,10 +70,19 @@ class Fetcher():
     supports_screenshots = False        # Can capture page screenshots
     supports_xpath_element_data = False # Can extract xpath element positions/data for visual selector
 
+    # Screenshot element locking - prevents layout shifts during screenshot capture
+    # Only needed for visual comparison (image_ssim_diff processor)
+    # Locks element dimensions in the first viewport to prevent headers/ads from resizing
+    lock_viewport_elements = False      # Default: disabled for performance
+
     def __init__(self, **kwargs):
         if kwargs and 'screenshot_format' in kwargs:
             self.screenshot_format = kwargs.get('screenshot_format')
 
+        # Allow lock_viewport_elements to be set via kwargs
+        if kwargs and 'lock_viewport_elements' in kwargs:
+            self.lock_viewport_elements = kwargs.get('lock_viewport_elements')
+
 
     @classmethod
     def get_status_icon_data(cls):
@@ -154,30 +162,16 @@ class Fetcher():
         """
         return {k.lower(): v for k, v in self.headers.items()}
 
-    def browser_steps_get_valid_steps(self):
-        if self.browser_steps is not None and len(self.browser_steps):
-            valid_steps = list(filter(
-                lambda s: (s['operation'] and len(s['operation']) and s['operation'] != 'Choose one'),
-                self.browser_steps))
-
-            # Just incase they selected Goto site by accident with older JS
-            if valid_steps and valid_steps[0]['operation'] == 'Goto site':
-                del(valid_steps[0])
-
-            return valid_steps
-
-        return None
-
     async def iterate_browser_steps(self, start_url=None):
-        from changedetectionio.blueprint.browser_steps.browser_steps import steppable_browser_interface
+        from changedetectionio.browser_steps.browser_steps import steppable_browser_interface, browser_steps_get_valid_steps
         from playwright._impl._errors import TimeoutError, Error
         from changedetectionio.jinja2_custom import render as jinja_render
         step_n = 0
 
-        if self.browser_steps is not None and len(self.browser_steps):
+        if self.browser_steps:
             interface = steppable_browser_interface(start_url=start_url)
             interface.page = self.page
-            valid_steps = self.browser_steps_get_valid_steps()
+            valid_steps = browser_steps_get_valid_steps(self.browser_steps)
 
             for step in valid_steps:
                 step_n += 1

changedetectionio/content_fetchers/playwright.py

@@ -1,3 +1,5 @@
+import asyncio
+import gc
 import json
 import os
 from urllib.parse import urlparse
@@ -7,20 +9,24 @@ from loguru import logger
 from changedetectionio.content_fetchers import SCREENSHOT_MAX_HEIGHT_DEFAULT, visualselector_xpath_selectors, \
     SCREENSHOT_SIZE_STITCH_THRESHOLD, SCREENSHOT_MAX_TOTAL_HEIGHT, XPATH_ELEMENT_JS, INSTOCK_DATA_JS, FAVICON_FETCHER_JS
 from changedetectionio.content_fetchers.base import Fetcher, manage_user_agent
-from changedetectionio.content_fetchers.exceptions import PageUnloadable, Non200ErrorCodeReceived, EmptyReply, ScreenshotUnavailable
+from changedetectionio.content_fetchers.exceptions import PageUnloadable, Non200ErrorCodeReceived, EmptyReply, ScreenshotUnavailable, \
+    BrowserStepsStepException
 
-async def capture_full_page_async(page, screenshot_format='JPEG'):
+
+async def capture_full_page_async(page, screenshot_format='JPEG', watch_uuid=None, lock_viewport_elements=False):
     import os
     import time
-    import multiprocessing
 
     start = time.time()
+    watch_info = f"[{watch_uuid}] " if watch_uuid else ""
 
+    setup_start = time.time()
     page_height = await page.evaluate("document.documentElement.scrollHeight")
     page_width = await page.evaluate("document.documentElement.scrollWidth")
     original_viewport = page.viewport_size
+    dimensions_time = time.time() - setup_start
 
-    logger.debug(f"Playwright viewport size {page.viewport_size} page height {page_height} page width {page_width}")
+    logger.debug(f"{watch_info}Playwright viewport size {page.viewport_size} page height {page_height} page width {page_width} (got dimensions in {dimensions_time:.2f}s)")
 
     # Use an approach similar to puppeteer: set a larger viewport and take screenshots in chunks
     step_size = SCREENSHOT_SIZE_STITCH_THRESHOLD # Size that won't cause GPU to overflow
@@ -28,25 +34,31 @@ async def capture_full_page_async(page, screenshot_format='JPEG'):
     y = 0
     elements_locked = False
 
-    if page_height > page.viewport_size['height']:
-
-        # Lock all element dimensions BEFORE screenshot to prevent CSS media queries from resizing
-        # capture_full_page_async() changes viewport height which triggers @media (min-height) rules
+    # Only lock viewport elements if explicitly enabled (for image_ssim_diff processor)
+    # This prevents headers/ads from resizing when viewport changes
+    if lock_viewport_elements and page_height > page.viewport_size['height']:
+        lock_start = time.time()
         lock_elements_js_path = os.path.join(os.path.dirname(__file__), 'res', 'lock-elements-sizing.js')
         with open(lock_elements_js_path, 'r') as f:
             lock_elements_js = f.read()
         await page.evaluate(lock_elements_js)
         elements_locked = True
+        lock_time = time.time() - lock_start
+        logger.debug(f"{watch_info}Viewport element locking enabled (took {lock_time:.2f}s)")
 
-        logger.debug("Element dimensions locked before screenshot capture")
-
+    if page_height > page.viewport_size['height']:
         if page_height < step_size:
             step_size = page_height # Incase page is bigger than default viewport but smaller than proposed step size
-        logger.debug(f"Setting bigger viewport to step through large page width W{page.viewport_size['width']}xH{step_size} because page_height > viewport_size")
+        viewport_start = time.time()
+        logger.debug(f"{watch_info}Setting bigger viewport to step through large page width W{page.viewport_size['width']}xH{step_size} because page_height > viewport_size")
         # Set viewport to a larger size to capture more content at once
         await page.set_viewport_size({'width': page.viewport_size['width'], 'height': step_size})
+        viewport_time = time.time() - viewport_start
+        logger.debug(f"{watch_info}Viewport changed to {page.viewport_size['width']}x{step_size} (took {viewport_time:.2f}s)")
 
     # Capture screenshots in chunks up to the max total height
+    capture_start = time.time()
+    chunk_times = []
     # Use PNG for better quality (no compression artifacts), JPEG for smaller size
     screenshot_type = screenshot_format.lower() if screenshot_format else 'jpeg'
     # PNG should use quality 100, JPEG uses configurable quality
@@ -68,7 +80,11 @@ async def capture_full_page_async(page, screenshot_format='JPEG'):
         if screenshot_type == 'jpeg':
             screenshot_kwargs['quality'] = screenshot_quality
 
+        chunk_start = time.time()
         screenshot_chunks.append(await page.screenshot(**screenshot_kwargs))
+        chunk_time = time.time() - chunk_start
+        chunk_times.append(chunk_time)
+        logger.debug(f"{watch_info}Chunk {len(screenshot_chunks)} captured in {chunk_time:.2f}s")
         y += step_size
 
     # Restore original viewport size
@@ -80,40 +96,54 @@ async def capture_full_page_async(page, screenshot_format='JPEG'):
         with open(unlock_elements_js_path, 'r') as f:
             unlock_elements_js = f.read()
         await page.evaluate(unlock_elements_js)
-        logger.debug("Element dimensions unlocked after screenshot capture")
+        logger.debug(f"{watch_info}Element dimensions unlocked after screenshot capture")
+
+    capture_time = time.time() - capture_start
+    total_capture_time = sum(chunk_times)
+    logger.debug(f"{watch_info}All {len(screenshot_chunks)} chunks captured in {capture_time:.2f}s (total chunk time: {total_capture_time:.2f}s)")
 
     # If we have multiple chunks, stitch them together
     if len(screenshot_chunks) > 1:
-        logger.debug(f"Screenshot stitching {len(screenshot_chunks)} chunks together")
+        stitch_start = time.time()
+        logger.debug(f"{watch_info}Starting stitching of {len(screenshot_chunks)} chunks")
 
-        # For small number of chunks (2-3), stitch inline to avoid multiprocessing overhead
-        # Only use separate process for many chunks (4+) to avoid blocking the event loop
-        if len(screenshot_chunks) <= 3:
-            from changedetectionio.content_fetchers.screenshot_handler import stitch_images_inline
-            screenshot = stitch_images_inline(screenshot_chunks, page_height, SCREENSHOT_MAX_TOTAL_HEIGHT)
-        else:
-            # Use separate process for many chunks to avoid blocking
-            # Always use spawn for thread safety - consistent behavior in tests and production
-            from changedetectionio.content_fetchers.screenshot_handler import stitch_images_worker
-            ctx = multiprocessing.get_context('spawn')
-            parent_conn, child_conn = ctx.Pipe()
-            p = ctx.Process(target=stitch_images_worker, args=(child_conn, screenshot_chunks, page_height, SCREENSHOT_MAX_TOTAL_HEIGHT))
-            p.start()
-            screenshot = parent_conn.recv_bytes()
-            p.join()
-            # Explicit cleanup
-            del p
-            del parent_conn, child_conn
+        # Always use spawn subprocess for ANY stitching (2+ chunks)
+        # PIL allocates at C level and Python GC never releases it - subprocess exit forces OS to reclaim
+        # Trade-off: 35MB resource_tracker vs 500MB+ PIL leak in main process
+        from changedetectionio.content_fetchers.screenshot_handler import stitch_images_worker_raw_bytes
+        import multiprocessing
+        import struct
 
+        ctx = multiprocessing.get_context('spawn')
+        parent_conn, child_conn = ctx.Pipe()
+        p = ctx.Process(target=stitch_images_worker_raw_bytes, args=(child_conn, page_height, SCREENSHOT_MAX_TOTAL_HEIGHT))
+        p.start()
+
+        # Send via raw bytes (no pickle)
+        parent_conn.send_bytes(struct.pack('I', len(screenshot_chunks)))
+        for chunk in screenshot_chunks:
+            parent_conn.send_bytes(chunk)
+
+        screenshot = parent_conn.recv_bytes()
+        p.join()
+
+        parent_conn.close()
+        child_conn.close()
+        del p, parent_conn, child_conn
+
+        stitch_time = time.time() - stitch_start
+        total_time = time.time() - start
+        setup_time = total_time - capture_time - stitch_time
         logger.debug(
-            f"Screenshot (chunked/stitched) - Page height: {page_height} Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT} - Stitched together in {time.time() - start:.2f}s")
-        # Explicit cleanup
-        del screenshot_chunks
-        screenshot_chunks = None
+            f"{watch_info}Screenshot complete - Page height: {page_height}px, Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT}px | "
+            f"Setup: {setup_time:.2f}s, Capture: {capture_time:.2f}s, Stitching: {stitch_time:.2f}s, Total: {total_time:.2f}s")
         return screenshot
 
+    total_time = time.time() - start
+    setup_time = total_time - capture_time
     logger.debug(
-        f"Screenshot Page height: {page_height} Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT} - Stitched together in {time.time() - start:.2f}s")
+        f"{watch_info}Screenshot complete - Page height: {page_height}px, Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT}px | "
+        f"Setup: {setup_time:.2f}s, Single chunk: {capture_time:.2f}s, Total: {total_time:.2f}s")
 
     return screenshot_chunks[0]
 
@@ -183,22 +213,36 @@ class fetcher(Fetcher):
 
     async def screenshot_step(self, step_n=''):
         super().screenshot_step(step_n=step_n)
-        screenshot = await capture_full_page_async(page=self.page, screenshot_format=self.screenshot_format)
+        watch_uuid = getattr(self, 'watch_uuid', None)
+        screenshot = await capture_full_page_async(page=self.page, screenshot_format=self.screenshot_format, watch_uuid=watch_uuid, lock_viewport_elements=self.lock_viewport_elements)
 
+        # Request GC immediately after screenshot to free memory
+        # Screenshots can be large and browser steps take many of them
+        await self.page.request_gc()
 
         if self.browser_steps_screenshot_path is not None:
             destination = os.path.join(self.browser_steps_screenshot_path, 'step_{}.jpeg'.format(step_n))
             logger.debug(f"Saving step screenshot to {destination}")
             with open(destination, 'wb') as f:
                 f.write(screenshot)
+            # Clear local reference to allow screenshot bytes to be collected
+            del screenshot
+            gc.collect()
 
     async def save_step_html(self, step_n):
         super().save_step_html(step_n=step_n)
         content = await self.page.content()
+
+        # Request GC after getting page content
+        await self.page.request_gc()
+
         destination = os.path.join(self.browser_steps_screenshot_path, 'step_{}.html'.format(step_n))
         logger.debug(f"Saving step HTML to {destination}")
         with open(destination, 'w', encoding='utf-8') as f:
             f.write(content)
+        # Clear local reference
+        del content
+        gc.collect()
 
     async def run(self,
                   fetch_favicon=True,
@@ -219,6 +263,7 @@ class fetcher(Fetcher):
         import playwright._impl._errors
         import time
         self.delete_browser_steps_screenshots()
+        self.watch_uuid = watch_uuid  # Store for use in screenshot_step
         response = None
 
         async with async_playwright() as p:
@@ -250,7 +295,7 @@ class fetcher(Fetcher):
             self.page.on("console", lambda msg: logger.debug(f"Playwright console: Watch URL: {url} {msg.type}: {msg.text} {msg.args}"))
 
             # Re-use as much code from browser steps as possible so its the same
-            from changedetectionio.blueprint.browser_steps.browser_steps import steppable_browser_interface
+            from changedetectionio.browser_steps.browser_steps import steppable_browser_interface
             browsersteps_interface = steppable_browser_interface(start_url=url)
             browsersteps_interface.page = self.page
 
@@ -304,7 +349,8 @@ class fetcher(Fetcher):
                     logger.error(f"Error fetching FavIcon info {str(e)}, continuing.")
 
             if self.status_code != 200 and not ignore_status_codes:
-                screenshot = await capture_full_page_async(self.page, screenshot_format=self.screenshot_format)
+                screenshot = await capture_full_page_async(self.page, screenshot_format=self.screenshot_format, watch_uuid=watch_uuid, lock_viewport_elements=self.lock_viewport_elements)
+                # Finally block will handle cleanup
                 raise Non200ErrorCodeReceived(url=url, status_code=self.status_code, screenshot=screenshot)
 
             if not empty_pages_are_a_change and len((await self.page.content()).strip()) == 0:
@@ -313,81 +359,100 @@ class fetcher(Fetcher):
                 await browser.close()
                 raise EmptyReply(url=url, status_code=response.status)
 
-            # Run Browser Steps here
-            if self.browser_steps_get_valid_steps():
-                await self.iterate_browser_steps(start_url=url)
-
-            await self.page.wait_for_timeout(extra_wait * 1000)
-
-            now = time.time()
-            # So we can find an element on the page where its selector was entered manually (maybe not xPath etc)
-            if current_include_filters is not None:
-                await self.page.evaluate("var include_filters={}".format(json.dumps(current_include_filters)))
-            else:
-                await self.page.evaluate("var include_filters=''")
-            await self.page.request_gc()
-
-            # request_gc before and after evaluate to free up memory
-            # @todo browsersteps etc
-            MAX_TOTAL_HEIGHT = int(os.getenv("SCREENSHOT_MAX_HEIGHT", SCREENSHOT_MAX_HEIGHT_DEFAULT))
-            self.xpath_data = await self.page.evaluate(XPATH_ELEMENT_JS, {
-                "visualselector_xpath_selectors": visualselector_xpath_selectors,
-                "max_height": MAX_TOTAL_HEIGHT
-            })
-            await self.page.request_gc()
-
-            self.instock_data = await self.page.evaluate(INSTOCK_DATA_JS)
-            await self.page.request_gc()
-
-            self.content = await self.page.content()
-            await self.page.request_gc()
-            logger.debug(f"Scrape xPath element data in browser done in {time.time() - now:.2f}s")
-
-
-            # Bug 3 in Playwright screenshot handling
-            # Some bug where it gives the wrong screenshot size, but making a request with the clip set first seems to solve it
-            # JPEG is better here because the screenshots can be very very large
-
-            # Screenshots also travel via the ws:// (websocket) meaning that the binary data is base64 encoded
-            # which will significantly increase the IO size between the server and client, it's recommended to use the lowest
-            # acceptable screenshot quality here
+            # Wrap remaining operations in try/finally to ensure cleanup
             try:
+                # Run Browser Steps here
+                if self.browser_steps:
+                    try:
+                        await self.iterate_browser_steps(start_url=url)
+                    except BrowserStepsStepException:
+                        # Finally block will handle cleanup
+                        raise
+
+                    await self.page.wait_for_timeout(extra_wait * 1000)
+
+                now = time.time()
+                # So we can find an element on the page where its selector was entered manually (maybe not xPath etc)
+                if current_include_filters is not None:
+                    await self.page.evaluate("var include_filters={}".format(json.dumps(current_include_filters)))
+                else:
+                    await self.page.evaluate("var include_filters=''")
+                await self.page.request_gc()
+
+                # request_gc before and after evaluate to free up memory
+                # @todo browsersteps etc
+                MAX_TOTAL_HEIGHT = int(os.getenv("SCREENSHOT_MAX_HEIGHT", SCREENSHOT_MAX_HEIGHT_DEFAULT))
+                self.xpath_data = await self.page.evaluate(XPATH_ELEMENT_JS, {
+                    "visualselector_xpath_selectors": visualselector_xpath_selectors,
+                    "max_height": MAX_TOTAL_HEIGHT
+                })
+                await self.page.request_gc()
+
+                self.instock_data = await self.page.evaluate(INSTOCK_DATA_JS)
+                await self.page.request_gc()
+
+                self.content = await self.page.content()
+                await self.page.request_gc()
+                logger.debug(f"Scrape xPath element data in browser done in {time.time() - now:.2f}s")
+
+
+                # Bug 3 in Playwright screenshot handling
+                # Some bug where it gives the wrong screenshot size, but making a request with the clip set first seems to solve it
+                # JPEG is better here because the screenshots can be very very large
+
+                # Screenshots also travel via the ws:// (websocket) meaning that the binary data is base64 encoded
+                # which will significantly increase the IO size between the server and client, it's recommended to use the lowest
+                # acceptable screenshot quality here
                 # The actual screenshot - this always base64 and needs decoding! horrible! huge CPU usage
-                self.screenshot = await capture_full_page_async(page=self.page, screenshot_format=self.screenshot_format)
+                self.screenshot = await capture_full_page_async(page=self.page, screenshot_format=self.screenshot_format, watch_uuid=watch_uuid, lock_viewport_elements=self.lock_viewport_elements)
 
-            except Exception as e:
-                # It's likely the screenshot was too long/big and something crashed
+                # Force aggressive memory cleanup - screenshots are large and base64 decode creates temporary buffers
+                await self.page.request_gc()
+                gc.collect()
+
+            except ScreenshotUnavailable:
+                # Re-raise screenshot unavailable exceptions
                 raise ScreenshotUnavailable(url=url, status_code=self.status_code)
+
             finally:
-                # Request garbage collection one more time before closing
+                # Clean up resources properly with timeouts to prevent hanging
                 try:
-                    await self.page.request_gc()
-                except:
-                    pass
-                
-                # Clean up resources properly
-                try:
-                    await self.page.request_gc()
-                except:
-                    pass
+                    if hasattr(self, 'page') and self.page:
+                        await self.page.request_gc()
+                        await asyncio.wait_for(self.page.close(), timeout=5.0)
+                        logger.debug(f"Successfully closed page for {url}")
+                except asyncio.TimeoutError:
+                    logger.warning(f"Timed out closing page for {url} (5s)")
+                except Exception as e:
+                    logger.warning(f"Error closing page for {url}: {e}")
+                finally:
+                    self.page = None
 
                 try:
-                    await self.page.close()
-                except:
-                    pass
-                self.page = None
+                    if context:
+                        await asyncio.wait_for(context.close(), timeout=5.0)
+                        logger.debug(f"Successfully closed context for {url}")
+                except asyncio.TimeoutError:
+                    logger.warning(f"Timed out closing context for {url} (5s)")
+                except Exception as e:
+                    logger.warning(f"Error closing context for {url}: {e}")
+                finally:
+                    context = None
 
                 try:
-                    await context.close()
-                except:
-                    pass
-                context = None
+                    if browser:
+                        await asyncio.wait_for(browser.close(), timeout=5.0)
+                        logger.debug(f"Successfully closed browser connection for {url}")
+                except asyncio.TimeoutError:
+                    logger.warning(f"Timed out closing browser connection for {url} (5s)")
+                except Exception as e:
+                    logger.warning(f"Error closing browser for {url}: {e}")
+                finally:
+                    browser = None
 
-                try:
-                    await browser.close()
-                except:
-                    pass
-                browser = None
+                # Force Python GC to release Playwright resources immediately
+                # Playwright objects can have circular references that delay cleanup
+                gc.collect()
 
 
 # Plugin registration for built-in fetcher

changedetectionio/content_fetchers/puppeteer.py

@@ -1,4 +1,5 @@
 import asyncio
+import gc
 import json
 import os
 import websockets.exceptions
@@ -20,18 +21,20 @@ from changedetectionio.content_fetchers.exceptions import PageUnloadable, Non200
 # Screenshots also travel via the ws:// (websocket) meaning that the binary data is base64 encoded
 # which will significantly increase the IO size between the server and client, it's recommended to use the lowest
 # acceptable screenshot quality here
-async def capture_full_page(page, screenshot_format='JPEG'):
+async def capture_full_page(page, screenshot_format='JPEG', watch_uuid=None, lock_viewport_elements=False):
     import os
     import time
-    import multiprocessing
 
     start = time.time()
+    watch_info = f"[{watch_uuid}] " if watch_uuid else ""
 
+    setup_start = time.time()
     page_height = await page.evaluate("document.documentElement.scrollHeight")
     page_width = await page.evaluate("document.documentElement.scrollWidth")
     original_viewport = page.viewport
+    dimensions_time = time.time() - setup_start
 
-    logger.debug(f"Puppeteer viewport size {page.viewport} page height {page_height} page width {page_width}")
+    logger.debug(f"{watch_info}Puppeteer viewport size {page.viewport} page height {page_height} page width {page_width} (got dimensions in {dimensions_time:.2f}s)")
 
     # Bug 3 in Playwright screenshot handling
     # Some bug where it gives the wrong screenshot size, but making a request with the clip set first seems to solve it
@@ -50,26 +53,41 @@ async def capture_full_page(page, screenshot_format='JPEG'):
     screenshot_chunks = []
     y = 0
     elements_locked = False
-    if page_height > page.viewport['height']:
-        # Lock all element dimensions BEFORE screenshot to prevent CSS media queries from resizing
-        # capture_full_page() changes viewport height which triggers @media (min-height) rules
+
+    # Only lock viewport elements if explicitly enabled (for image_ssim_diff processor)
+    # This prevents headers/ads from resizing when viewport changes
+    if lock_viewport_elements and page_height > page.viewport['height']:
+        lock_start = time.time()
         lock_elements_js_path = os.path.join(os.path.dirname(__file__), 'res', 'lock-elements-sizing.js')
+        file_read_start = time.time()
         with open(lock_elements_js_path, 'r') as f:
             lock_elements_js = f.read()
-        await page.evaluate(lock_elements_js)
-        elements_locked = True
-        logger.debug("Element dimensions locked before screenshot capture")
+        file_read_time = time.time() - file_read_start
 
+        evaluate_start = time.time()
+        await page.evaluate(lock_elements_js)
+        evaluate_time = time.time() - evaluate_start
+
+        elements_locked = True
+        lock_time = time.time() - lock_start
+        logger.debug(f"{watch_info}Viewport element locking enabled - File read: {file_read_time:.3f}s, Browser evaluate: {evaluate_time:.2f}s, Total: {lock_time:.2f}s")
+
+    if page_height > page.viewport['height']:
         if page_height < step_size:
             step_size = page_height # Incase page is bigger than default viewport but smaller than proposed step size
+        viewport_start = time.time()
         await page.setViewport({'width': page.viewport['width'], 'height': step_size})
+        viewport_time = time.time() - viewport_start
+        logger.debug(f"{watch_info}Viewport changed to {page.viewport['width']}x{step_size} (took {viewport_time:.2f}s)")
 
+    capture_start = time.time()
+    chunk_times = []
     while y < min(page_height, SCREENSHOT_MAX_TOTAL_HEIGHT):
         # better than scrollTo incase they override it in the page
         await page.evaluate(
             """(y) => {
-                document.documentElement.scrollTop = y;
-                document.body.scrollTop = y;
+                const el = document.scrollingElement;
+                if (el) el.scrollTop = y;
             }""",
             y
         )
@@ -82,7 +100,11 @@ async def capture_full_page(page, screenshot_format='JPEG'):
         if screenshot_type == 'jpeg':
             screenshot_kwargs['quality'] = screenshot_quality
 
+        chunk_start = time.time()
         screenshot_chunks.append(await page.screenshot(**screenshot_kwargs))
+        chunk_time = time.time() - chunk_start
+        chunk_times.append(chunk_time)
+        logger.debug(f"{watch_info}Chunk {len(screenshot_chunks)} captured in {chunk_time:.2f}s")
         y += step_size
 
     await page.setViewport({'width': original_viewport['width'], 'height': original_viewport['height']})
@@ -93,26 +115,53 @@ async def capture_full_page(page, screenshot_format='JPEG'):
         with open(unlock_elements_js_path, 'r') as f:
             unlock_elements_js = f.read()
         await page.evaluate(unlock_elements_js)
-        logger.debug("Element dimensions unlocked after screenshot capture")
+        logger.debug(f"{watch_info}Element dimensions unlocked after screenshot capture")
+
+    capture_time = time.time() - capture_start
+    total_capture_time = sum(chunk_times)
+    logger.debug(f"{watch_info}All {len(screenshot_chunks)} chunks captured in {capture_time:.2f}s (total chunk time: {total_capture_time:.2f}s)")
 
     if len(screenshot_chunks) > 1:
-        # Always use spawn for thread safety - consistent behavior in tests and production
-        from changedetectionio.content_fetchers.screenshot_handler import stitch_images_worker
-        logger.debug(f"Screenshot stitching {len(screenshot_chunks)} chunks together")
+        stitch_start = time.time()
+        logger.debug(f"{watch_info}Starting stitching of {len(screenshot_chunks)} chunks")
+
+        # Always use spawn subprocess for ANY stitching (2+ chunks)
+        # PIL allocates at C level and Python GC never releases it - subprocess exit forces OS to reclaim
+        # Trade-off: 35MB resource_tracker vs 500MB+ PIL leak in main process
+        from changedetectionio.content_fetchers.screenshot_handler import stitch_images_worker_raw_bytes
+        import multiprocessing
+        import struct
+
         ctx = multiprocessing.get_context('spawn')
         parent_conn, child_conn = ctx.Pipe()
-        p = ctx.Process(target=stitch_images_worker, args=(child_conn, screenshot_chunks, page_height, SCREENSHOT_MAX_TOTAL_HEIGHT))
+        p = ctx.Process(target=stitch_images_worker_raw_bytes, args=(child_conn, page_height, SCREENSHOT_MAX_TOTAL_HEIGHT))
         p.start()
+
+        # Send via raw bytes (no pickle)
+        parent_conn.send_bytes(struct.pack('I', len(screenshot_chunks)))
+        for chunk in screenshot_chunks:
+            parent_conn.send_bytes(chunk)
+
         screenshot = parent_conn.recv_bytes()
         p.join()
-        logger.debug(
-            f"Screenshot (chunked/stitched) - Page height: {page_height} Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT} - Stitched together in {time.time() - start:.2f}s")
 
-        screenshot_chunks = None
+        parent_conn.close()
+        child_conn.close()
+        del p, parent_conn, child_conn
+
+        stitch_time = time.time() - stitch_start
+        total_time = time.time() - start
+        setup_time = total_time - capture_time - stitch_time
+        logger.debug(
+            f"{watch_info}Screenshot complete - Page height: {page_height}px, Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT}px | "
+            f"Setup: {setup_time:.2f}s, Capture: {capture_time:.2f}s, Stitching: {stitch_time:.2f}s, Total: {total_time:.2f}s")
         return screenshot
 
+    total_time = time.time() - start
+    setup_time = total_time - capture_time
     logger.debug(
-        f"Screenshot Page height: {page_height} Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT} - Stitched together in {time.time() - start:.2f}s")
+        f"{watch_info}Screenshot complete - Page height: {page_height}px, Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT}px | "
+        f"Setup: {setup_time:.2f}s, Single chunk: {capture_time:.2f}s, Total: {total_time:.2f}s")
     return screenshot_chunks[0]
 
 
@@ -173,19 +222,36 @@ class fetcher(Fetcher):
                 self.browser_connection_url += f"{r}--proxy-server={proxy_url}"
 
     async def quit(self, watch=None):
-        try:
-            await self.page.close()
-            del self.page
-        except Exception as e:
-            pass
+        watch_uuid = watch.get('uuid') if watch else 'unknown'
 
+        # Close page
         try:
-            await self.browser.close()
-            del self.browser
+            if hasattr(self, 'page') and self.page:
+                await asyncio.wait_for(self.page.close(), timeout=5.0)
+                logger.debug(f"[{watch_uuid}] Page closed successfully")
+        except asyncio.TimeoutError:
+            logger.warning(f"[{watch_uuid}] Timed out closing page (5s)")
         except Exception as e:
-            pass
+            logger.warning(f"[{watch_uuid}] Error closing page: {e}")
+        finally:
+            self.page = None
 
-        logger.info("Cleanup puppeteer complete.")
+        # Close browser connection
+        try:
+            if hasattr(self, 'browser') and self.browser:
+                await asyncio.wait_for(self.browser.close(), timeout=5.0)
+                logger.debug(f"[{watch_uuid}] Browser closed successfully")
+        except asyncio.TimeoutError:
+            logger.warning(f"[{watch_uuid}] Timed out closing browser (5s)")
+        except Exception as e:
+            logger.warning(f"[{watch_uuid}] Error closing browser: {e}")
+        finally:
+            self.browser = None
+
+        logger.info(f"[{watch_uuid}] Cleanup puppeteer complete")
+
+        # Force garbage collection to release resources
+        gc.collect()
 
     async def fetch_page(self,
                          current_include_filters,
@@ -215,9 +281,11 @@ class fetcher(Fetcher):
         # Connect directly using the specified browser_ws_endpoint
         # @todo timeout
         try:
+            logger.debug(f"[{watch_uuid}] Connecting to browser at {self.browser_connection_url}")
             self.browser = await pyppeteer_instance.connect(browserWSEndpoint=self.browser_connection_url,
                                                             ignoreHTTPSErrors=True
                                                             )
+            logger.debug(f"[{watch_uuid}] Browser connected successfully")
         except websockets.exceptions.InvalidStatusCode as e:
             raise BrowserConnectError(msg=f"Error while trying to connect the browser, Code {e.status_code} (check your access, whitelist IP, password etc)")
         except websockets.exceptions.InvalidURI:
@@ -226,7 +294,20 @@ class fetcher(Fetcher):
             raise BrowserConnectError(msg=f"Error connecting to the browser - Exception '{str(e)}'")
 
         # more reliable is to just request a new page
-        self.page = await self.browser.newPage()
+        try:
+            logger.debug(f"[{watch_uuid}] Creating new page")
+            self.page = await self.browser.newPage()
+            logger.debug(f"[{watch_uuid}] Page created successfully")
+        except Exception as e:
+            logger.error(f"[{watch_uuid}] Failed to create new page: {e}")
+            # Browser is connected but page creation failed - must cleanup browser
+            try:
+                await asyncio.wait_for(self.browser.close(), timeout=3.0)
+            except Exception as cleanup_error:
+                logger.error(f"[{watch_uuid}] Failed to cleanup browser after page creation failure: {cleanup_error}")
+            finally:
+                self.browser = None
+            raise
         
         # Add console handler to capture console.log from favicon fetcher
         #self.page.on('console', lambda msg: logger.debug(f"Browser console [{msg.type}]: {msg.text}"))
@@ -295,6 +376,12 @@ class fetcher(Fetcher):
             w = extra_wait - 2 if extra_wait > 4 else 2
             logger.debug(f"Waiting {w} seconds before calling Page.stopLoading...")
             await asyncio.sleep(w)
+
+            # Check if page still exists (might have been closed due to error during sleep)
+            if not self.page or not hasattr(self.page, '_client'):
+                logger.debug("Page already closed, skipping stopLoading")
+                return
+
             logger.debug("Issuing stopLoading command...")
             await self.page._client.send('Page.stopLoading')
             logger.debug("stopLoading command sent!")
@@ -320,7 +407,9 @@ class fetcher(Fetcher):
             asyncio.create_task(handle_frame_navigation())
             response = await self.page.goto(url, timeout=0)
             await asyncio.sleep(1 + extra_wait)
-            await self.page._client.send('Page.stopLoading')
+            # Check if page still exists before sending command
+            if self.page and hasattr(self.page, '_client'):
+                await self.page._client.send('Page.stopLoading')
 
             if response:
                 break
@@ -357,7 +446,7 @@ class fetcher(Fetcher):
                 logger.error(f"Error fetching FavIcon info {str(e)}, continuing.")
 
         if self.status_code != 200 and not ignore_status_codes:
-            screenshot = await capture_full_page(page=self.page, screenshot_format=self.screenshot_format)
+            screenshot = await capture_full_page(page=self.page, screenshot_format=self.screenshot_format, watch_uuid=watch_uuid, lock_viewport_elements=self.lock_viewport_elements)
 
             raise Non200ErrorCodeReceived(url=url, status_code=self.status_code, screenshot=screenshot)
 
@@ -369,7 +458,7 @@ class fetcher(Fetcher):
 
         # Run Browser Steps here
         # @todo not yet supported, we switch to playwright in this case
-        #            if self.browser_steps_get_valid_steps():
+        #            if self.browser_steps:
         #                self.iterate_browser_steps()
 
 
@@ -387,7 +476,11 @@ class fetcher(Fetcher):
 
         # Now take screenshot (scrolling may trigger layout changes, but measurements are already captured)
         logger.debug(f"Screenshot format {self.screenshot_format}")
-        self.screenshot = await capture_full_page(page=self.page, screenshot_format=self.screenshot_format)
+        self.screenshot = await capture_full_page(page=self.page, screenshot_format=self.screenshot_format, watch_uuid=watch_uuid, lock_viewport_elements=self.lock_viewport_elements)
+
+        # Force garbage collection - pyppeteer base64 decode creates temporary buffers
+        import gc
+        gc.collect()
         self.xpath_data = await self.page.evaluate(XPATH_ELEMENT_JS, {
             "visualselector_xpath_selectors": visualselector_xpath_selectors,
             "max_height": MAX_TOTAL_HEIGHT
@@ -441,6 +534,14 @@ class fetcher(Fetcher):
             )
         except asyncio.TimeoutError:
             raise (BrowserFetchTimedOut(msg=f"Browser connected but was unable to process the page in {max_time} seconds."))
+        finally:
+            # Internal cleanup on any exception/timeout - call quit() immediately
+            # This prevents connection leaks during exception bursts
+            # Worker.py's quit() call becomes a redundant safety net (idempotent)
+            try:
+                await self.quit(watch={'uuid': watch_uuid} if watch_uuid else None)
+            except Exception as cleanup_error:
+                logger.error(f"[{watch_uuid}] Error during internal quit() cleanup: {cleanup_error}")
 
 
 # Plugin registration for built-in fetcher

changedetectionio/content_fetchers/requests.py

@@ -1,12 +1,14 @@
 from loguru import logger
+from urllib.parse import urljoin, urlparse
 import hashlib
 import os
 import re
 import asyncio
-from functools import partial
+
 from changedetectionio import strtobool
 from changedetectionio.content_fetchers.exceptions import BrowserStepsInUnsupportedFetcher, EmptyReply, Non200ErrorCodeReceived
 from changedetectionio.content_fetchers.base import Fetcher
+from changedetectionio.validate_url import is_private_hostname
 
 
 # "html_requests" is listed as the default fetcher in store.py!
@@ -36,7 +38,7 @@ class fetcher(Fetcher):
         import requests
         from requests.exceptions import ProxyError, ConnectionError, RequestException
 
-        if self.browser_steps_get_valid_steps():
+        if self.browser_steps:
             raise BrowserStepsInUnsupportedFetcher(url=url)
 
         proxies = {}
@@ -55,18 +57,72 @@ class fetcher(Fetcher):
 
         session = requests.Session()
 
+        # Configure retry adapter for low-level network errors only
+        # Retries connection timeouts, read timeouts, connection resets - not HTTP status codes
+        # Especially helpful in parallel test execution when servers are slow/overloaded
+        # Configurable via REQUESTS_RETRY_MAX_COUNT (default: 3 attempts)
+        from requests.adapters import HTTPAdapter
+        from urllib3.util.retry import Retry
+
+        max_retries = int(os.getenv("REQUESTS_RETRY_MAX_COUNT", "6"))
+        retry_strategy = Retry(
+            total=max_retries,
+            connect=max_retries,  # Retry connection timeouts
+            read=max_retries,     # Retry read timeouts
+            status=0,             # Don't retry on HTTP status codes
+            backoff_factor=0.5,   # Wait 0.3s, 0.6s, 1.2s between retries
+            allowed_methods=["HEAD", "GET", "OPTIONS", "POST"],
+            raise_on_status=False
+        )
+        adapter = HTTPAdapter(max_retries=retry_strategy)
+        session.mount("http://", adapter)
+        session.mount("https://", adapter)
 
         if strtobool(os.getenv('ALLOW_FILE_URI', 'false')) and url.startswith('file://'):
             from requests_file import FileAdapter
             session.mount('file://', FileAdapter())
+
+        allow_iana_restricted = strtobool(os.getenv('ALLOW_IANA_RESTRICTED_ADDRESSES', 'false'))
+
         try:
+            # Fresh DNS check at fetch time — catches DNS rebinding regardless of add-time cache.
+            if not allow_iana_restricted:
+                parsed_initial = urlparse(url)
+                if parsed_initial.hostname and is_private_hostname(parsed_initial.hostname):
+                    raise Exception(f"Fetch blocked: '{url}' resolves to a private/reserved IP address. "
+                                    f"Set ALLOW_IANA_RESTRICTED_ADDRESSES=true to allow.")
+
             r = session.request(method=request_method,
                                 data=request_body.encode('utf-8') if type(request_body) is str else request_body,
                                 url=url,
                                 headers=request_headers,
                                 timeout=timeout,
                                 proxies=proxies,
-                                verify=False)
+                                verify=False,
+                                allow_redirects=False)
+
+            # Manually follow redirects so each hop's resolved IP can be validated,
+            # preventing SSRF via an open redirect on a public host.
+            current_url = url
+            for _ in range(10):
+                if not r.is_redirect:
+                    break
+                location = r.headers.get('Location', '')
+                redirect_url = urljoin(current_url, location)
+                if not allow_iana_restricted:
+                    parsed_redirect = urlparse(redirect_url)
+                    if parsed_redirect.hostname and is_private_hostname(parsed_redirect.hostname):
+                        raise Exception(f"Redirect blocked: '{redirect_url}' resolves to a private/reserved IP address.")
+                current_url = redirect_url
+                r = session.request('GET', redirect_url,
+                                    headers=request_headers,
+                                    timeout=timeout,
+                                    proxies=proxies,
+                                    verify=False,
+                                    allow_redirects=False)
+            else:
+                raise Exception("Too many redirects")
+
         except Exception as e:
             msg = str(e)
             if proxies and 'SOCKSHTTPSConnectionPool' in msg:
@@ -92,10 +148,32 @@ class fetcher(Fetcher):
                         # Default to UTF-8 for XML if no encoding found
                         r.encoding = 'utf-8'
                 else:
-                    # For other content types, use chardet
-                    encoding = chardet.detect(r.content)['encoding']
-                    if encoding:
-                        r.encoding = encoding
+                    # No charset in HTTP header - sniff encoding in priority order matching browsers
+                    # (WHATWG encoding sniffing algorithm):
+                    # 1. BOM - highest confidence, check before anything else
+                    # 2. <meta charset> in first 2kb
+                    # 3. chardet statistical detection - last resort
+                    # See: https://github.com/dgtlmoon/changedetection.io/issues/3952
+                    boms = [
+                        (b'\xef\xbb\xbf', 'utf-8-sig'),
+                        (b'\xff\xfe', 'utf-16-le'),
+                        (b'\xfe\xff', 'utf-16-be'),
+                    ]
+                    bom_encoding = next((enc for bom, enc in boms if r.content.startswith(bom)), None)
+                    if bom_encoding:
+                        logger.info(f"URL: {url} Using encoding '{bom_encoding}' detected from BOM")
+                        r.encoding = bom_encoding
+                    else:
+                        meta_charset_match = re.search(rb'<meta[^>]+charset\s*=\s*["\']?\s*([^"\'\s;>]+)', r.content[:2000], re.IGNORECASE)
+                        if meta_charset_match:
+                            encoding = meta_charset_match.group(1).decode('ascii', errors='ignore')
+                            logger.info(f"URL: {url} No content-type encoding in HTTP headers - Using encoding '{encoding}' from HTML meta charset tag")
+                            r.encoding = encoding
+                        else:
+                            encoding = chardet.detect(r.content)['encoding']
+                            logger.warning(f"URL: {url} No charset in headers or meta tag, guessed encoding as '{encoding}' via chardet")
+                            if encoding:
+                                r.encoding = encoding
 
         self.headers = r.headers
 
@@ -142,10 +220,11 @@ class fetcher(Fetcher):
                   watch_uuid=None,
                   ):
         """Async wrapper that runs the synchronous requests code in a thread pool"""
-        
+
         loop = asyncio.get_event_loop()
-        
+
         # Run the synchronous _run_sync in a thread pool to avoid blocking the event loop
+        # Retry logic is handled by requests' HTTPAdapter (see _run_sync for configuration)
         await loop.run_in_executor(
             None,  # Use default ThreadPoolExecutor
             lambda: self._run_sync(
@@ -163,7 +242,6 @@ class fetcher(Fetcher):
         )
 
     async def quit(self, watch=None):
-
         # In case they switched to `requests` fetcher from something else
         # Then the screenshot could be old, in any case, it's not used here.
         # REMOVE_REQUESTS_OLD_SCREENSHOTS - Mainly used for testing

changedetectionio/content_fetchers/res/lock-elements-sizing.js

@@ -1,5 +1,5 @@
 /**
- * Lock Element Dimensions for Screenshot Capture
+ * Lock Element Dimensions for Screenshot Capture (First Viewport Only)
  *
  * THE PROBLEM:
  * When taking full-page screenshots of tall pages, Chrome/Puppeteer/Playwright need to:
@@ -10,40 +10,31 @@
  * However, changing the viewport height triggers CSS media queries like:
  *   @media (min-height: 860px) { .ad { height: 250px; } }
  *
- * This causes elements (especially ads) to resize during screenshot capture, creating a mismatch:
- * - Screenshot shows element at NEW size (after media query triggered)
- * - xpath element coordinates measured at OLD size (before viewport change)
- * - Visual selector overlays don't align with screenshot
- *
- * EXAMPLE BUG:
- * - Initial viewport: 1280x800, ad height: 138px, article position: 279px ✓
- * - Viewport changes to 1280x3809 for screenshot
- * - Media query triggers: ad expands to 250px
- * - All content below shifts down by 112px (250-138)
- * - Article now at position: 391px (279+112)
- * - But xpath data says 279px → 112px mismatch! ✗
+ * This causes elements (especially ads/headers) to resize during screenshot capture.
  *
  * THE SOLUTION:
- * Before changing viewport, lock ALL element dimensions with !important inline styles.
- * Inline styles with !important override media query CSS, preventing layout changes.
+ * Lock element dimensions in the FIRST VIEWPORT ONLY with !important inline styles.
+ * This prevents headers, navigation, and top ads from resizing when viewport changes.
+ * We only lock the visible portion because:
+ * - Most layout shifts happen in headers/navbars/top ads
+ * - Locking only visible elements is 100x+ faster (100-200 elements vs 10,000+)
+ * - Below-fold content shifts don't affect visual comparison accuracy
  *
  * WHAT THIS SCRIPT DOES:
- * 1. Iterates through every element on the page
- * 2. Captures current computed dimensions (width, height)
- * 3. Sets inline styles with !important to freeze those dimensions
+ * 1. Gets current viewport height
+ * 2. Finds elements within first viewport (top of page to bottom of screen)
+ * 3. Locks their dimensions with !important inline styles
  * 4. Disables ResizeObserver API (for JS-based resizing)
- * 5. When viewport changes for screenshot, media queries can't resize anything
- * 6. Layout remains consistent → xpath coordinates match screenshot ✓
  *
  * USAGE:
  * Execute this script BEFORE calling capture_full_page() / screenshot functions.
- * The page must be fully loaded and settled at its initial viewport size.
- * No need to restore state afterward - page is closed after screenshot.
+ * Only enabled for image_ssim_diff processor (visual comparison).
+ * Default: OFF for performance.
  *
  * PERFORMANCE:
- * - Iterates all DOM elements (can be 1000s on complex pages)
- * - Typically completes in 50-200ms
- * - One-time cost before screenshot, well worth it for coordinate accuracy
+ * - Only processes 100-300 elements (first viewport) vs 10,000+ (entire page)
+ * - Typically completes in 10-50ms
+ * - 100x+ faster than locking entire page
  *
  * @see https://github.com/dgtlmoon/changedetection.io/issues/XXXX
  */
@@ -52,11 +43,34 @@
     // Store original styles in a global WeakMap for later restoration
     window.__elementSizingRestore = new WeakMap();
 
-    // Lock ALL element dimensions to prevent media query layout changes
-    document.querySelectorAll('*').forEach(el => {
-        const computed = window.getComputedStyle(el);
-        const rect = el.getBoundingClientRect();
+    const start = performance.now();
 
+    // Get current viewport height (visible portion of page)
+    const viewportHeight = window.innerHeight;
+
+    // Get all elements and filter to FIRST VIEWPORT ONLY
+    // This dramatically reduces elements to process (100-300 vs 10,000+)
+    const allElements = Array.from(document.querySelectorAll('*'));
+
+    // BATCH READ PHASE: Get bounding rects and filter to viewport
+    const measurements = allElements.map(el => {
+        const rect = el.getBoundingClientRect();
+        const computed = window.getComputedStyle(el);
+
+        // Only lock elements in the first viewport (visible on initial page load)
+        // rect.top < viewportHeight means element starts within visible area
+        const inViewport = rect.top < viewportHeight && rect.top >= 0;
+        const hasSize = rect.height > 0 && rect.width > 0;
+
+        return inViewport && hasSize ? { el, computed, rect } : null;
+    }).filter(Boolean);  // Remove null entries
+
+    const elapsed = performance.now() - start;
+    console.log(`Locked first viewport elements: ${measurements.length} of ${allElements.length} total elements (viewport height: ${viewportHeight}px, took ${elapsed.toFixed(0)}ms)`);
+
+    // BATCH WRITE PHASE: Apply all inline styles without triggering layout
+    // No interleaved reads means browser can optimize style application
+    measurements.forEach(({el, computed, rect}) => {
         // Save original inline style values BEFORE locking
         const properties = ['height', 'min-height', 'max-height', 'width', 'min-width', 'max-width'];
         const originalStyles = {};
@@ -89,5 +103,5 @@
         disconnect() {}
     };
 
-    console.log('✓ Element dimensions locked to prevent media query changes during screenshot');
+    console.log(`✓ Element dimensions locked (${measurements.length} elements) to prevent media query changes during screenshot`);
 })();

changedetectionio/content_fetchers/screenshot_handler.py

@@ -8,92 +8,42 @@ from loguru import logger
 
 from changedetectionio.content_fetchers import SCREENSHOT_MAX_HEIGHT_DEFAULT, SCREENSHOT_DEFAULT_QUALITY
 
-# Cache font to avoid loading on every stitch
-_cached_font = None
-
-def _get_caption_font():
-    """Get or create cached font for caption text."""
-    global _cached_font
-    if _cached_font is None:
-        from PIL import ImageFont
-        try:
-            _cached_font = ImageFont.truetype("arial.ttf", 35)
-        except IOError:
-            _cached_font = ImageFont.load_default()
-    return _cached_font
-
-
-def stitch_images_inline(chunks_bytes, original_page_height, capture_height):
-    """
-    Stitch image chunks together inline (no multiprocessing).
-    Optimized for small number of chunks (2-3) to avoid process creation overhead.
-
-    Args:
-        chunks_bytes: List of JPEG image bytes
-        original_page_height: Original page height in pixels
-        capture_height: Maximum capture height
-
-    Returns:
-        bytes: Stitched JPEG image
-    """
-    import os
-    import io
-    from PIL import Image, ImageDraw
-
-    # Load images from byte chunks
-    images = [Image.open(io.BytesIO(b)) for b in chunks_bytes]
-    total_height = sum(im.height for im in images)
-    max_width = max(im.width for im in images)
-
-    # Create stitched image
-    stitched = Image.new('RGB', (max_width, total_height))
-    y_offset = 0
-    for im in images:
-        stitched.paste(im, (0, y_offset))
-        y_offset += im.height
-        im.close()  # Close immediately after pasting
-
-    # Draw caption only if page was trimmed
-    if original_page_height > capture_height:
-        draw = ImageDraw.Draw(stitched)
-        caption_text = f"WARNING: Screenshot was {original_page_height}px but trimmed to {capture_height}px because it was too long"
-        padding = 10
-        font = _get_caption_font()
-
-        bbox = draw.textbbox((0, 0), caption_text, font=font)
-        text_width = bbox[2] - bbox[0]
-        text_height = bbox[3] - bbox[1]
-
-        # Draw white background rectangle
-        draw.rectangle([(0, 0), (max_width, text_height + 2 * padding)], fill=(255, 255, 255))
-
-        # Draw text centered
-        text_x = (max_width - text_width) // 2
-        draw.text((text_x, padding), caption_text, font=font, fill=(255, 0, 0))
-
-    # Encode to JPEG
-    output = io.BytesIO()
-    stitched.save(output, format="JPEG", quality=int(os.getenv("SCREENSHOT_QUALITY", SCREENSHOT_DEFAULT_QUALITY)), optimize=True)
-    result = output.getvalue()
-
-    # Cleanup
-    stitched.close()
-
-    return result
-
-
-def stitch_images_worker(pipe_conn, chunks_bytes, original_page_height, capture_height):
+def stitch_images_worker_raw_bytes(pipe_conn, original_page_height, capture_height):
     """
     Stitch image chunks together in a separate process.
-    Used for large number of chunks (4+) to avoid blocking the main event loop.
+
+    Uses spawn multiprocessing to isolate PIL's C-level memory allocation.
+    When the subprocess exits, the OS reclaims ALL memory including C-level allocations
+    that Python's GC cannot release. This prevents the ~50MB per stitch from accumulating
+    in the main process.
+
+    Trade-off: Adds 35MB resource_tracker subprocess, but prevents 500MB+ memory leak
+    in main process (much better at scale: 35GB vs 500GB for 1000 instances).
+
+    Args:
+        pipe_conn: Pipe connection to receive data and send result
+        original_page_height: Original page height in pixels
+        capture_height: Maximum capture height
     """
     import os
     import io
+    import struct
     from PIL import Image, ImageDraw, ImageFont
 
     try:
+        # Receive chunk count as 4-byte integer (no pickle!)
+        count_bytes = pipe_conn.recv_bytes()
+        chunk_count = struct.unpack('I', count_bytes)[0]
+
+        # Receive each chunk as raw bytes (no pickle!)
+        chunks_bytes = []
+        for _ in range(chunk_count):
+            chunks_bytes.append(pipe_conn.recv_bytes())
+
         # Load images from byte chunks
         images = [Image.open(io.BytesIO(b)) for b in chunks_bytes]
+        del chunks_bytes
+
         total_height = sum(im.height for im in images)
         max_width = max(im.width for im in images)
 
@@ -103,15 +53,14 @@ def stitch_images_worker(pipe_conn, chunks_bytes, original_page_height, capture_
         for im in images:
             stitched.paste(im, (0, y_offset))
             y_offset += im.height
-            im.close()  # Close immediately after pasting
+            im.close()
+        del images
 
         # Draw caption only if page was trimmed
         if original_page_height > capture_height:
             draw = ImageDraw.Draw(stitched)
             caption_text = f"WARNING: Screenshot was {original_page_height}px but trimmed to {capture_height}px because it was too long"
             padding = 10
-
-            # Try to load font
             try:
                 font = ImageFont.truetype("arial.ttf", 35)
             except IOError:
@@ -120,23 +69,26 @@ def stitch_images_worker(pipe_conn, chunks_bytes, original_page_height, capture_
             bbox = draw.textbbox((0, 0), caption_text, font=font)
             text_width = bbox[2] - bbox[0]
             text_height = bbox[3] - bbox[1]
-
-            # Draw white background rectangle
             draw.rectangle([(0, 0), (max_width, text_height + 2 * padding)], fill=(255, 255, 255))
-
-            # Draw text centered
             text_x = (max_width - text_width) // 2
             draw.text((text_x, padding), caption_text, font=font, fill=(255, 0, 0))
 
-        # Encode and send image with optimization
+        # Encode and send
         output = io.BytesIO()
         stitched.save(output, format="JPEG", quality=int(os.getenv("SCREENSHOT_QUALITY", SCREENSHOT_DEFAULT_QUALITY)), optimize=True)
-        pipe_conn.send_bytes(output.getvalue())
+        result_bytes = output.getvalue()
 
         stitched.close()
+        del stitched
+        output.close()
+        del output
+
+        pipe_conn.send_bytes(result_bytes)
+        del result_bytes
+
     except Exception as e:
-        pipe_conn.send(f"error:{e}")
+        logger.error(f"Error in stitch_images_worker_raw_bytes: {e}")
+        error_msg = f"error:{e}".encode('utf-8')
+        pipe_conn.send_bytes(error_msg)
     finally:
         pipe_conn.close()
-
-

changedetectionio/content_fetchers/webdriver_selenium.py

@@ -15,7 +15,7 @@ class fetcher(Fetcher):
     proxy_url = None
 
     # Capability flags
-    supports_browser_steps = True
+    supports_browser_steps = False
     supports_screenshots = True
     supports_xpath_element_data = True
 
@@ -156,6 +156,19 @@ class fetcher(Fetcher):
                     from PIL import Image
                     import io
                     img = Image.open(io.BytesIO(screenshot_png))
+                    # Convert to RGB if needed (JPEG doesn't support transparency)
+                    # Always convert non-RGB modes to RGB to ensure JPEG compatibility
+                    if img.mode in ('RGBA', 'LA', 'P', 'PA'):
+                        # Handle transparency by compositing onto white background
+                        if img.mode == 'P':
+                            img = img.convert('RGBA')
+                        background = Image.new('RGB', img.size, (255, 255, 255))
+                        if img.mode in ('RGBA', 'LA', 'PA'):
+                            background.paste(img, mask=img.split()[-1])  # Use alpha channel as mask
+                        img = background
+                    elif img.mode != 'RGB':
+                        # For other modes, direct conversion
+                        img = img.convert('RGB')
                     jpeg_buffer = io.BytesIO()
                     img.save(jpeg_buffer, format='JPEG', quality=int(os.getenv("SCREENSHOT_QUALITY", 72)))
                     self.screenshot = jpeg_buffer.getvalue()

changedetectionio/custom_queue.py

@@ -57,14 +57,15 @@ class SignalPriorityQueue(queue.PriorityQueue):
     def put(self, item, block=True, timeout=None):
         # Call the parent's put method first
         super().put(item, block, timeout)
-        
+
         # After putting the item in the queue, check if it has a UUID and emit signal
         if hasattr(item, 'item') and isinstance(item.item, dict) and 'uuid' in item.item:
             uuid = item.item['uuid']
             # Get the signal and send it if it exists
             watch_check_update = signal('watch_check_update')
             if watch_check_update:
-                # Send the watch_uuid parameter
+                # NOTE: This would block other workers from .put/.get while this signal sends
+                # Signal handlers may iterate the queue/datastore while holding locks
                 watch_check_update.send(watch_uuid=uuid)
         
         # Send queue_length signal with current queue size
@@ -312,14 +313,15 @@ class AsyncSignalPriorityQueue(asyncio.PriorityQueue):
     async def put(self, item):
         # Call the parent's put method first
         await super().put(item)
-        
+
         # After putting the item in the queue, check if it has a UUID and emit signal
         if hasattr(item, 'item') and isinstance(item.item, dict) and 'uuid' in item.item:
             uuid = item.item['uuid']
             # Get the signal and send it if it exists
             watch_check_update = signal('watch_check_update')
             if watch_check_update:
-                # Send the watch_uuid parameter
+                # NOTE: This would block other workers from .put/.get while this signal sends
+                # Signal handlers may iterate the queue/datastore while holding locks
                 watch_check_update.send(watch_uuid=uuid)
         
         # Send queue_length signal with current queue size

changedetectionio/favicon_utils.py

@@ -0,0 +1,43 @@
+"""
+Favicon utilities for changedetection.io
+Handles favicon MIME type detection with caching
+"""
+
+from functools import lru_cache
+
+
+@lru_cache(maxsize=1000)
+def get_favicon_mime_type(filepath):
+    """
+    Detect MIME type of favicon by reading file content using puremagic.
+    Results are cached to avoid repeatedly reading the same files.
+
+    Args:
+        filepath: Full path to the favicon file
+
+    Returns:
+        MIME type string (e.g., 'image/png')
+    """
+    mime = None
+
+    try:
+        import puremagic
+        with open(filepath, 'rb') as f:
+            content_bytes = f.read(200)  # Read first 200 bytes
+
+        detections = puremagic.magic_string(content_bytes)
+        if detections:
+            mime = detections[0].mime_type
+    except Exception:
+        pass
+
+    # Fallback to mimetypes if puremagic fails
+    if not mime:
+        import mimetypes
+        mime, _ = mimetypes.guess_type(filepath)
+
+    # Final fallback based on extension
+    if not mime:
+        mime = 'image/x-icon' if filepath.endswith('.ico') else 'image/png'
+
+    return mime

changedetectionio/flask_app.py

@@ -4,16 +4,18 @@ import flask_login
 import locale
 import os
 import queue
+import re
 import sys
 import threading
 import time
 import timeago
 from blinker import signal
+from pathlib import Path
 
 from changedetectionio.strtobool import strtobool
 from threading import Event
 from changedetectionio.queue_handlers import RecheckPriorityQueue, NotificationQueue
-from changedetectionio import worker_handler
+from changedetectionio import worker_pool
 
 from flask import (
     Flask,
@@ -26,9 +28,6 @@ from flask import (
     session,
     url_for,
 )
-from urllib.parse import urlparse
-from flask_compress import Compress as FlaskCompress
-from flask_login import current_user
 from flask_restful import abort, Api
 from flask_cors import CORS
 
@@ -41,10 +40,13 @@ from loguru import logger
 
 from changedetectionio import __version__
 from changedetectionio import queuedWatchMetaData
-from changedetectionio.api import Watch, WatchHistory, WatchSingleHistory, WatchHistoryDiff, CreateWatch, Import, SystemInfo, Tag, Tags, Notifications, WatchFavicon
+from changedetectionio.api import Watch, WatchHistory, WatchSingleHistory, WatchHistoryDiff, CreateWatch, Import, SystemInfo, Tag, Tags, Notifications, WatchFavicon, Spec
 from changedetectionio.api.Search import Search
 from .time_handler import is_within_schedule
 from changedetectionio.languages import get_available_languages, get_language_codes, get_flag_for_locale, get_timeago_locale
+from changedetectionio.favicon_utils import get_favicon_mime_type
+
+IN_PYTEST = "pytest" in sys.modules or "PYTEST_CURRENT_TEST" in os.environ
 
 datastore = None
 
@@ -55,7 +57,7 @@ extra_stylesheets = []
 # Use bulletproof janus-based queues for sync/async reliability  
 update_q = RecheckPriorityQueue()
 notification_q = NotificationQueue()
-MAX_QUEUE_SIZE = 2000
+MAX_QUEUE_SIZE = 5000
 
 app = Flask(__name__,
             static_url_path="",
@@ -67,11 +69,43 @@ socketio_server = None
 
 # Enable CORS, especially useful for the Chrome extension to operate from anywhere
 CORS(app)
+from werkzeug.routing import BaseConverter, ValidationError
+from uuid import UUID
+
+class StrictUUIDConverter(BaseConverter):
+    # Special sentinel values allowed in addition to strict UUIDs
+    _ALLOWED_SENTINELS = frozenset({'first'})
+
+    def to_python(self, value: str) -> str:
+        if value in self._ALLOWED_SENTINELS:
+            return value
+        try:
+            u = UUID(value)
+        except ValueError as e:
+            raise ValidationError() from e
+        # Reject non-standard formats (braces, URNs, no-hyphens)
+        if str(u) != value.lower():
+            raise ValidationError()
+        return str(u)
+
+    def to_url(self, value) -> str:
+        return str(value)
+
+# app setup (once)
+app.url_map.converters["uuid_str"] = StrictUUIDConverter
+
+# Flask-Compress handles HTTP compression, Socket.IO compression disabled to prevent memory leak.
+# There's also a bug between flask compress and socketio that causes some kind of slow memory leak
+# It's better to use compression on your reverse proxy (nginx etc) instead.
+if strtobool(os.getenv("FLASK_ENABLE_COMPRESSION")):
+    from flask_compress import Compress as FlaskCompress
+    app.config['COMPRESS_MIN_SIZE'] = 2096
+    app.config['COMPRESS_MIMETYPES'] = ['text/html', 'text/css', 'text/javascript', 'application/json', 'application/javascript', 'image/svg+xml']
+    # Use gzip only - smaller memory footprint than zstd/brotli (4-8KB vs 200-500KB contexts)
+    app.config['COMPRESS_ALGORITHM'] = ['gzip']
+    compress = FlaskCompress()
+    compress.init_app(app)
 
-# Super handy for compressing large BrowserSteps responses and others
-FlaskCompress(app)
-app.config['COMPRESS_MIN_SIZE'] = 4096
-app.config['COMPRESS_MIMETYPES'] = ['text/html', 'text/css', 'text/javascript', 'application/json', 'application/javascript', 'image/svg+xml']
 app.config['TEMPLATES_AUTO_RELOAD'] = False
 
 
@@ -84,6 +118,18 @@ app.config['NEW_VERSION_AVAILABLE'] = False
 if os.getenv('FLASK_SERVER_NAME'):
     app.config['SERVER_NAME'] = os.getenv('FLASK_SERVER_NAME')
 
+# Babel/i18n configuration
+app.config['BABEL_TRANSLATION_DIRECTORIES'] = str(Path(__file__).parent / 'translations')
+app.config['BABEL_DEFAULT_LOCALE'] = 'en_GB'
+
+# Session configuration
+# NOTE: Flask session (for locale, etc.) is separate from Flask-Login's remember-me cookie
+# - Flask session stores data like session['locale'] in a signed cookie
+# - Flask-Login's remember=True creates a separate authentication cookie
+# - Setting PERMANENT_SESSION_LIFETIME controls how long the Flask session cookie lasts
+from datetime import timedelta
+app.config['PERMANENT_SESSION_LIFETIME'] = timedelta(days=3650)  # ~10 years (effectively unlimited)
+
 #app.config["EXPLAIN_TEMPLATE_LOADING"] = True
 
 
@@ -172,12 +218,16 @@ def _jinja2_filter_format_number_locale(value: float) -> str:
     "Formats for example 4000.10 to the local locale default of 4,000.10"
     # Format the number with two decimal places (locale format string will return 6 decimal)
     formatted_value = locale.format_string("%.2f", value, grouping=True)
-
     return formatted_value
 
+@app.template_filter('regex_search')
+def _jinja2_filter_regex_search(value, pattern):
+    import re
+    return re.search(pattern, str(value)) is not None
+
 @app.template_global('is_checking_now')
 def _watch_is_checking_now(watch_obj, format="%Y-%m-%d %H:%M:%S"):
-    return worker_handler.is_watch_running(watch_obj['uuid'])
+    return worker_pool.is_watch_running(watch_obj['uuid'])
 
 @app.template_global('get_watch_queue_position')
 def _get_watch_queue_position(watch_obj):
@@ -188,13 +238,13 @@ def _get_watch_queue_position(watch_obj):
 @app.template_global('get_current_worker_count')
 def _get_current_worker_count():
     """Get the current number of operational workers"""
-    return worker_handler.get_worker_count()
+    return worker_pool.get_worker_count()
 
 @app.template_global('get_worker_status_info')
 def _get_worker_status_info():
     """Get detailed worker status information for display"""
-    status = worker_handler.get_worker_status()
-    running_uuids = worker_handler.get_running_uuids()
+    status = worker_pool.get_worker_status()
+    running_uuids = worker_pool.get_running_uuids()
     
     return {
         'count': status['worker_count'],
@@ -248,6 +298,47 @@ def _jinja2_filter_seconds_precise(timestamp):
 
     return format(int(time.time()-timestamp), ',d')
 
+@app.template_filter('format_duration')
+def _jinja2_filter_format_duration(seconds):
+    """Format a duration in seconds into human readable string like '5 days, 3 hours, 30 minutes'"""
+    from datetime import timedelta
+
+    if not seconds or seconds < 0:
+        return gettext('0 seconds')
+
+    td = timedelta(seconds=int(seconds))
+
+    # Calculate components
+    years = td.days // 365
+    remaining_days = td.days % 365
+    months = remaining_days // 30
+    remaining_days = remaining_days % 30
+    weeks = remaining_days // 7
+    days = remaining_days % 7
+
+    hours = td.seconds // 3600
+    minutes = (td.seconds % 3600) // 60
+    secs = td.seconds % 60
+
+    # Build parts list
+    parts = []
+    if years > 0:
+        parts.append(f"{years} {gettext('year') if years == 1 else gettext('years')}")
+    if months > 0:
+        parts.append(f"{months} {gettext('month') if months == 1 else gettext('months')}")
+    if weeks > 0:
+        parts.append(f"{weeks} {gettext('week') if weeks == 1 else gettext('weeks')}")
+    if days > 0:
+        parts.append(f"{days} {gettext('day') if days == 1 else gettext('days')}")
+    if hours > 0:
+        parts.append(f"{hours} {gettext('hour') if hours == 1 else gettext('hours')}")
+    if minutes > 0:
+        parts.append(f"{minutes} {gettext('minute') if minutes == 1 else gettext('minutes')}")
+    if secs > 0 or not parts:
+        parts.append(f"{secs} {gettext('second') if secs == 1 else gettext('seconds')}")
+
+    return ", ".join(parts)
+
 @app.template_filter('fetcher_status_icons')
 def _jinja2_filter_fetcher_status_icons(fetcher_name):
     """Get status icon HTML for a given fetcher.
@@ -297,6 +388,8 @@ def _jinja2_filter_fetcher_status_icons(fetcher_name):
 
     return ''
 
+_RE_SANITIZE_TAG = re.compile(r'[^a-zA-Z0-9]')
+
 @app.template_filter('sanitize_tag_class')
 def _jinja2_filter_sanitize_tag_class(tag_title):
     """Sanitize a tag title to create a valid CSS class name.
@@ -308,9 +401,8 @@ def _jinja2_filter_sanitize_tag_class(tag_title):
     Returns:
         str: A sanitized string suitable for use as a CSS class name
     """
-    import re
     # Remove all non-alphanumeric characters and convert to lowercase
-    sanitized = re.sub(r'[^a-zA-Z0-9]', '', tag_title).lower()
+    sanitized = _RE_SANITIZE_TAG.sub('', tag_title).lower()
     # Ensure it starts with a letter (CSS requirement)
     if sanitized and not sanitized[0].isalpha():
         sanitized = 'tag' + sanitized
@@ -370,6 +462,9 @@ def changedetection_app(config=None, datastore_o=None):
     global datastore, socketio_server
     datastore = datastore_o
 
+    # Set datastore reference in notification queue for all_muted checking
+    notification_q.set_datastore(datastore)
+
     # Import and create a wrapper for is_safe_url that has access to app
     from changedetectionio.is_safe_url import is_safe_url as _is_safe_url
 
@@ -380,7 +475,10 @@ def changedetection_app(config=None, datastore_o=None):
     # so far just for read-only via tests, but this will be moved eventually to be the main source
     # (instead of the global var)
     app.config['DATASTORE'] = datastore_o
-    
+
+    # Store batch mode flag to skip background threads when running in batch mode
+    app.config['batch_mode'] = config.get('batch_mode', False) if config else False
+
     # Store the signal in the app config to ensure it's accessible everywhere
     app.config['watch_check_update_SIGNAL'] = watch_check_update
 
@@ -392,16 +490,21 @@ def changedetection_app(config=None, datastore_o=None):
     available_languages = get_available_languages()
     language_codes = get_language_codes()
 
+    _locale_aliases = {
+        'zh-TW': 'zh_Hant_TW',  # Traditional Chinese: browser sends zh-TW, we use zh_Hant_TW
+        'zh_TW': 'zh_Hant_TW',  # Also handle underscore variant
+    }
+    _locale_match_list = language_codes + list(_locale_aliases.keys())
+
     def get_locale():
         # 1. Try to get locale from session (user explicitly selected)
         if 'locale' in session:
-            locale = session['locale']
-            logger.trace(f"DEBUG: get_locale() returning from session: {locale}")
-            return locale
+            return session['locale']
+
         # 2. Fall back to Accept-Language header
-        locale = request.accept_languages.best_match(language_codes)
-        logger.trace(f"DEBUG: get_locale() returning from Accept-Language: {locale}")
-        return locale
+        browser_locale = request.accept_languages.best_match(_locale_match_list)
+        # 3. Map browser locale to our internal locale if needed
+        return _locale_aliases.get(browser_locale, browser_locale)
 
     # Initialize Babel with locale selector
     babel = Babel(app, locale_selector=get_locale)
@@ -453,22 +556,22 @@ def changedetection_app(config=None, datastore_o=None):
 
 
     watch_api.add_resource(WatchHistoryDiff,
-                           '/api/v1/watch/<string:uuid>/difference/<string:from_timestamp>/<string:to_timestamp>',
+                           '/api/v1/watch/<uuid_str:uuid>/difference/<string:from_timestamp>/<string:to_timestamp>',
                            resource_class_kwargs={'datastore': datastore})
     watch_api.add_resource(WatchSingleHistory,
-                           '/api/v1/watch/<string:uuid>/history/<string:timestamp>',
+                           '/api/v1/watch/<uuid_str:uuid>/history/<string:timestamp>',
                            resource_class_kwargs={'datastore': datastore, 'update_q': update_q})
     watch_api.add_resource(WatchFavicon,
-                           '/api/v1/watch/<string:uuid>/favicon',
+                           '/api/v1/watch/<uuid_str:uuid>/favicon',
                            resource_class_kwargs={'datastore': datastore})
     watch_api.add_resource(WatchHistory,
-                           '/api/v1/watch/<string:uuid>/history',
+                           '/api/v1/watch/<uuid_str:uuid>/history',
                            resource_class_kwargs={'datastore': datastore})
 
     watch_api.add_resource(CreateWatch, '/api/v1/watch',
                            resource_class_kwargs={'datastore': datastore, 'update_q': update_q})
 
-    watch_api.add_resource(Watch, '/api/v1/watch/<string:uuid>',
+    watch_api.add_resource(Watch, '/api/v1/watch/<uuid_str:uuid>',
                            resource_class_kwargs={'datastore': datastore, 'update_q': update_q})
 
     watch_api.add_resource(SystemInfo, '/api/v1/systeminfo',
@@ -481,7 +584,7 @@ def changedetection_app(config=None, datastore_o=None):
     watch_api.add_resource(Tags, '/api/v1/tags',
                            resource_class_kwargs={'datastore': datastore})
 
-    watch_api.add_resource(Tag, '/api/v1/tag', '/api/v1/tag/<string:uuid>',
+    watch_api.add_resource(Tag, '/api/v1/tag', '/api/v1/tag/<uuid_str:uuid>',
                            resource_class_kwargs={'datastore': datastore, 'update_q': update_q})
                            
     watch_api.add_resource(Search, '/api/v1/search',
@@ -490,6 +593,8 @@ def changedetection_app(config=None, datastore_o=None):
     watch_api.add_resource(Notifications, '/api/v1/notifications',
                            resource_class_kwargs={'datastore': datastore})
 
+    watch_api.add_resource(Spec, '/api/v1/full-spec')
+
     @login_manager.user_loader
     def user_loader(email):
         user = User()
@@ -518,9 +623,23 @@ def changedetection_app(config=None, datastore_o=None):
     @app.route('/set-language/<locale>')
     def set_language(locale):
         """Set the user's preferred language in the session"""
+        if not request.cookies:
+            logger.error("Cannot set language without session cookie")
+            flash("Cannot set language without session cookie", 'error')
+            return redirect(url_for('watchlist.index'))
+
         # Validate the locale against available languages
         if locale in language_codes:
+            # Make session permanent so language preference persists across browser sessions
+            # NOTE: This is the Flask session cookie (separate from Flask-Login's remember-me auth cookie)
+            session.permanent = True
             session['locale'] = locale
+
+            # CRITICAL: Flask-Babel caches the locale in the request context (ctx.babel_locale)
+            # We must refresh to clear this cache so the new locale takes effect immediately
+            # This is especially important for tests where multiple requests happen rapidly
+            from flask_babel import refresh
+            refresh()
         else:
             logger.error(f"Invalid locale {locale}, available: {language_codes}")
 
@@ -617,8 +736,14 @@ def changedetection_app(config=None, datastore_o=None):
     def static_content(group, filename):
         from flask import make_response
         import re
-        group = re.sub(r'[^\w.-]+', '', group.lower())
-        filename = re.sub(r'[^\w.-]+', '', filename.lower())
+
+        # Strict sanitization: only allow a-z, 0-9, and underscore (blocks .. and other traversal)
+        group = re.sub(r'[^a-z0-9_-]+', '', group.lower())
+        filename = filename
+
+        # Additional safety: reject if sanitization resulted in empty strings
+        if not group or not filename:
+            abort(404)
 
         if group == 'screenshot':
             # Could be sensitive, follow password requirements
@@ -652,18 +777,11 @@ def changedetection_app(config=None, datastore_o=None):
 
             favicon_filename = watch.get_favicon_filename()
             if favicon_filename:
-                try:
-                    import magic
-                    mime = magic.from_file(
-                        os.path.join(watch.watch_data_dir, favicon_filename),
-                        mime=True
-                    )
-                except ImportError:
-                    # Fallback, no python-magic
-                    import mimetypes
-                    mime, encoding = mimetypes.guess_type(favicon_filename)
+                # Use cached MIME type detection
+                filepath = os.path.join(watch.data_dir, favicon_filename)
+                mime = get_favicon_mime_type(filepath)
 
-                response = make_response(send_from_directory(watch.watch_data_dir, favicon_filename))
+                response = make_response(send_from_directory(watch.data_dir, favicon_filename))
                 response.headers['Content-type'] = mime
                 response.headers['Cache-Control'] = 'max-age=300, must-revalidate'  # Cache for 5 minutes, then revalidate
                 return response
@@ -758,13 +876,15 @@ def changedetection_app(config=None, datastore_o=None):
 
     # watchlist UI buttons etc
     import changedetectionio.blueprint.ui as ui
-    app.register_blueprint(ui.construct_blueprint(datastore, update_q, worker_handler, queuedWatchMetaData, watch_check_update))
+    app.register_blueprint(ui.construct_blueprint(datastore, update_q, worker_pool, queuedWatchMetaData, watch_check_update))
 
     import changedetectionio.blueprint.watchlist as watchlist
     app.register_blueprint(watchlist.construct_blueprint(datastore=datastore, update_q=update_q, queuedWatchMetaData=queuedWatchMetaData), url_prefix='')
 
     # Initialize Socket.IO server conditionally based on settings
-    socket_io_enabled = datastore.data['settings']['application']['ui'].get('socket_io_enabled', True)
+    socket_io_enabled = datastore.data['settings']['application'].get('ui', {}).get('socket_io_enabled', True)
+    if socket_io_enabled and app.config.get('batch_mode'):
+        socket_io_enabled = False
     if socket_io_enabled:
         from changedetectionio.realtime.socket_server import init_socketio
         global socketio_server
@@ -793,10 +913,10 @@ def changedetection_app(config=None, datastore_o=None):
         expected_workers = int(os.getenv("FETCH_WORKERS", datastore.data['settings']['requests']['workers']))
         
         # Get basic status
-        status = worker_handler.get_worker_status()
+        status = worker_pool.get_worker_status()
         
         # Perform health check
-        health_result = worker_handler.check_worker_health(
+        health_result = worker_pool.check_worker_health(
             expected_count=expected_workers,
             update_q=update_q,
             notification_q=notification_q,
@@ -860,16 +980,31 @@ def changedetection_app(config=None, datastore_o=None):
     # Can be overridden by ENV or use the default settings
     n_workers = int(os.getenv("FETCH_WORKERS", datastore.data['settings']['requests']['workers']))
     logger.info(f"Starting {n_workers} workers during app initialization")
-    worker_handler.start_workers(n_workers, update_q, notification_q, app, datastore)
+    worker_pool.start_workers(n_workers, update_q, notification_q, app, datastore)
 
-    # @todo handle ctrl break
-    ticker_thread = threading.Thread(target=ticker_thread_check_time_launch_checks).start()
-    threading.Thread(target=notification_runner).start()
+    # Skip background threads in batch mode (just process queue and exit)
+    batch_mode = app.config.get('batch_mode', False)
+    if not batch_mode:
+        # @todo handle ctrl break
+        ticker_thread = threading.Thread(target=ticker_thread_check_time_launch_checks, daemon=True, name="TickerThread-ScheduleChecker").start()
 
-    in_pytest = "pytest" in sys.modules or "PYTEST_CURRENT_TEST" in os.environ
-    # Check for new release version, but not when running in test/build or pytest
-    if not os.getenv("GITHUB_REF", False) and not strtobool(os.getenv('DISABLE_VERSION_CHECK', 'no')) and not in_pytest:
-        threading.Thread(target=check_for_new_version).start()
+        # Start configurable number of notification workers (default 1)
+        notification_workers = int(os.getenv("NOTIFICATION_WORKERS", "1"))
+        for i in range(notification_workers):
+            threading.Thread(
+                target=notification_runner,
+                args=(i,),
+                daemon=True,
+                name=f"NotificationRunner-{i}"
+            ).start()
+        logger.info(f"Started {notification_workers} notification worker(s)")
+
+        in_pytest = "pytest" in sys.modules or "PYTEST_CURRENT_TEST" in os.environ
+        # Check for new release version, but not when running in test/build or pytest
+        if not os.getenv("GITHUB_REF", False) and not strtobool(os.getenv('DISABLE_VERSION_CHECK', 'no')) and not in_pytest:
+            threading.Thread(target=check_for_new_version, daemon=True, name="VersionChecker").start()
+    else:
+        logger.info("Batch mode: Skipping ticker thread, notification runner, and version checker")
 
     # Return the Flask app - the Socket.IO will be attached to it but initialized separately
     # This avoids circular dependencies
@@ -882,15 +1017,16 @@ def check_for_new_version():
     import urllib3
     urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
 
+    session = requests.Session()
+    session.verify = False
+
     while not app.config.exit.is_set():
         try:
-            r = requests.post("https://changedetection.io/check-ver.php",
+            r = session.post("https://changedetection.io/check-ver.php",
                               data={'version': __version__,
                                     'app_guid': datastore.data['app_guid'],
                                     'watch_count': len(datastore.data['watching'])
-                                    },
-
-                              verify=False)
+                                    })
         except:
             pass
 
@@ -904,14 +1040,14 @@ def check_for_new_version():
         app.config.exit.wait(86400)
 
 
-def notification_runner():
+def notification_runner(worker_id=0):
     global notification_debug_log
     from datetime import datetime
     import json
     with app.app_context():
         while not app.config.exit.is_set():
             try:
-                # At the moment only one thread runs (single runner)
+                # Multiple workers can run concurrently (configurable via NOTIFICATION_WORKERS)
                 n_object = notification_q.get(block=False)
             except queue.Empty:
                 app.config.exit.wait(1)
@@ -937,7 +1073,7 @@ def notification_runner():
                         sent_obj = process_notification(n_object, datastore)
 
                 except Exception as e:
-                    logger.error(f"Watch URL: {n_object['watch_url']}  Error {str(e)}")
+                    logger.error(f"Notification worker {worker_id} - Watch URL: {n_object['watch_url']}  Error {str(e)}")
 
                     # UUID wont be present when we submit a 'test' from the global settings
                     if 'uuid' in n_object:
@@ -967,6 +1103,10 @@ def ticker_thread_check_time_launch_checks():
     logger.debug(f"System env MINIMUM_SECONDS_RECHECK_TIME {recheck_time_minimum_seconds}")
 
     # Workers are now started during app initialization, not here
+    WAIT_TIME_BETWEEN_LOOP = 1.0 if not IN_PYTEST else 0.01
+    if IN_PYTEST:
+        # The time between loops should be less than the first .sleep/wait in def wait_for_all_checks() of tests/util.py
+        logger.warning(f"Looks like we're in PYTEST! Setting time between searching for items to add to the queue to {WAIT_TIME_BETWEEN_LOOP}s")
 
     while not app.config.exit.is_set():
 
@@ -974,7 +1114,7 @@ def ticker_thread_check_time_launch_checks():
         now = time.time()
         if now - last_health_check > 60:
             expected_workers = int(os.getenv("FETCH_WORKERS", datastore.data['settings']['requests']['workers']))
-            health_result = worker_handler.check_worker_health(
+            health_result = worker_pool.check_worker_health(
                 expected_count=expected_workers,
                 update_q=update_q,
                 notification_q=notification_q,
@@ -984,11 +1124,19 @@ def ticker_thread_check_time_launch_checks():
             
             if health_result['status'] != 'healthy':
                 logger.warning(f"Worker health check: {health_result['message']}")
-                
+
             last_health_check = now
 
+        # Check if all checks are paused
+        if datastore.data['settings']['application'].get('all_paused', False):
+            app.config.exit.wait(1)
+            continue
+
         # Get a list of watches by UUID that are currently fetching data
-        running_uuids = worker_handler.get_running_uuids()
+        running_uuids = worker_pool.get_running_uuids()
+
+        # Build set of queued UUIDs once for O(1) lookup instead of O(n) per watch
+        queued_uuids = {q_item.item['uuid'] for q_item in update_q.queue}
 
         # Re #232 - Deepcopy the data incase it changes while we're iterating through it all
         watch_uuid_list = []
@@ -1006,16 +1154,17 @@ def ticker_thread_check_time_launch_checks():
             else:
                 break
 
-        # Re #438 - Don't place more watches in the queue to be checked if the queue is already large
-        while update_q.qsize() >= 2000:
-            logger.warning(f"Recheck watches queue size limit reached ({MAX_QUEUE_SIZE}), skipping adding more items")
-            app.config.exit.wait(10.0)
-
-
         recheck_time_system_seconds = int(datastore.threshold_seconds)
 
         # Check for watches outside of the time threshold to put in the thread queue.
-        for uuid in watch_uuid_list:
+        for watch_index, uuid in enumerate(watch_uuid_list):
+            # Re #438 - Check queue size every 100 watches for CPU efficiency (not every watch)
+            if watch_index % 100 == 0:
+                current_queue_size = update_q.qsize()
+                if current_queue_size >= MAX_QUEUE_SIZE:
+                    logger.debug(f"Queue size limit reached ({current_queue_size}/{MAX_QUEUE_SIZE}), stopping scheduler this iteration.")
+                    break
+
             now = time.time()
             watch = datastore.data['watching'].get(uuid)
             if not watch:
@@ -1065,7 +1214,7 @@ def ticker_thread_check_time_launch_checks():
             seconds_since_last_recheck = now - watch['last_checked']
 
             if seconds_since_last_recheck >= (threshold + watch.jitter_seconds) and seconds_since_last_recheck >= recheck_time_minimum_seconds:
-                if not uuid in running_uuids and uuid not in [q_uuid.item['uuid'] for q_uuid in update_q.queue]:
+                if not uuid in running_uuids and uuid not in queued_uuids:
 
                     # Proxies can be set to have a limit on seconds between which they can be called
                     watch_proxy = datastore.get_preferred_proxy_for_watch(uuid=uuid)
@@ -1090,7 +1239,7 @@ def ticker_thread_check_time_launch_checks():
                     priority = int(time.time())
 
                     # Into the queue with you
-                    queued_successfully = worker_handler.queue_item_async_safe(update_q,
+                    queued_successfully = worker_pool.queue_item_async_safe(update_q,
                                                                                queuedWatchMetaData.PrioritizedItem(priority=priority,
                                                                                                                    item={'uuid': uuid})
                                                                                )
@@ -1108,4 +1257,4 @@ def ticker_thread_check_time_launch_checks():
                     watch.jitter_seconds = 0
 
         # Should be low so we can break this out in testing
-        app.config.exit.wait(1)
+        app.config.exit.wait(WAIT_TIME_BETWEEN_LOOP)

changedetectionio/forms.py

@@ -7,8 +7,6 @@ from flask_babel import lazy_gettext as _l, gettext
 from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES, RSS_TEMPLATE_TYPE_OPTIONS, RSS_TEMPLATE_HTML_DEFAULT
 from changedetectionio.conditions.form import ConditionFormRow
 from changedetectionio.notification_service import NotificationContextData
-from changedetectionio.processors.image_ssim_diff import SCREENSHOT_COMPARISON_THRESHOLD_OPTIONS, \
-    SCREENSHOT_COMPARISON_THRESHOLD_OPTIONS_DEFAULT
 from changedetectionio.strtobool import strtobool
 from changedetectionio import processors
 
@@ -37,7 +35,7 @@ from changedetectionio.widgets import TernaryNoneBooleanField
 
 # default
 # each select <option data-enabled="enabled-0-0"
-from changedetectionio.blueprint.browser_steps.browser_steps import browser_step_ui_config
+from changedetectionio.browser_steps.browser_steps import browser_step_ui_config
 
 from changedetectionio import html_tools, content_fetchers
 
@@ -494,7 +492,6 @@ class ValidateJinja2Template(object):
     Validates that a {token} is from a valid set
     """
     def __call__(self, form, field):
-        from changedetectionio import notification
         from changedetectionio.jinja2_custom import create_jinja_env
         from jinja2 import BaseLoader, TemplateSyntaxError, UndefinedError
         from jinja2.meta import find_undeclared_variables
@@ -611,13 +608,12 @@ class ValidateCSSJSONXPATHInput(object):
                     raise ValidationError("XPath not permitted in this field!")
                 from lxml import etree, html
                 import elementpath
-                # xpath 2.0-3.1
-                from elementpath.xpath3 import XPath3Parser
+                from changedetectionio.html_tools import SafeXPath3Parser
                 tree = html.fromstring("<html></html>")
                 line = line.replace('xpath:', '')
 
                 try:
-                    elementpath.select(tree, line.strip(), parser=XPath3Parser)
+                    elementpath.select(tree, line.strip(), parser=SafeXPath3Parser)
                 except elementpath.ElementPathError as e:
                     message = field.gettext('\'%s\' is not a valid XPath expression. (%s)')
                     raise ValidationError(message % (line, str(e)))
@@ -727,10 +723,10 @@ class ValidateStartsWithRegex(object):
                 raise ValidationError(self.message or _l("Invalid value."))
 
 class quickWatchForm(Form):
-    url = fields.URLField('URL', validators=[validateURL()])
-    tags = StringTagUUID('Group tag', [validators.Optional()])
+    url = fields.URLField(_l('URL'), validators=[validateURL()])
+    tags = StringTagUUID(_l('Group tag'), validators=[validators.Optional()])
     watch_submit_button = SubmitField(_l('Watch'), render_kw={"class": "pure-button pure-button-primary"})
-    processor = RadioField(_l('Processor'), choices=lambda: processors.available_processors(), default="text_json_diff")
+    processor = RadioField(_l('Processor'), choices=lambda: processors.available_processors(), default=processors.get_default_processor)
     edit_and_watch_submit_button = SubmitField(_l('Edit > Watch'), render_kw={"class": "pure-button pure-button-primary"})
 
 
@@ -749,7 +745,7 @@ class commonSettingsForm(Form):
     notification_format = SelectField(_l('Notification format'), choices=list(valid_notification_formats.items()))
     notification_title = StringField(_l('Notification Title'), default='ChangeDetection.io Notification - {{ watch_url }}', validators=[validators.Optional(), ValidateJinja2Template()])
     notification_urls = StringListField(_l('Notification URL List'), validators=[validators.Optional(), ValidateAppRiseServers(), ValidateJinja2Template()])
-    processor = RadioField( label=_l("Processor - What do you want to achieve?"), choices=lambda: processors.available_processors(), default="text_json_diff")
+    processor = RadioField( label=_l("Processor - What do you want to achieve?"), choices=lambda: processors.available_processors(), default=processors.get_default_processor)
     scheduler_timezone_default = StringField(_l("Default timezone for watch check scheduler"), render_kw={"list": "timezones"}, validators=[validateTimeZoneName()])
     webdriver_delay = IntegerField(_l('Wait seconds before extracting text'), validators=[validators.Optional(), validators.NumberRange(min=1, message=_l("Should contain one or more seconds"))])
 
@@ -763,7 +759,7 @@ class commonSettingsForm(Form):
 
 
 class importForm(Form):
-    processor = RadioField(_l('Processor'), choices=lambda: processors.available_processors(), default="text_json_diff")
+    processor = RadioField(_l('Processor'), choices=lambda: processors.available_processors(), default=processors.get_default_processor)
     urls = TextAreaField(_l('URLs'))
     xlsx_file = FileField(_l('Upload .xlsx file'), validators=[FileAllowed(['xlsx'], _l('Must be .xlsx file!'))])
     file_mapping = SelectField(_l('File mapping'), [validators.DataRequired()], choices={('wachete', 'Wachete mapping'), ('custom','Custom mapping')})
@@ -786,6 +782,7 @@ class processor_text_json_diff_form(commonSettingsForm):
 
     time_between_check = EnhancedFormField(
         TimeBetweenCheckForm,
+        label=_l('Time Between Check'),
         conditional_field='time_between_check_use_default',
         conditional_message=REQUIRE_ATLEAST_ONE_TIME_PART_WHEN_NOT_GLOBAL_DEFAULT,
         conditional_test_function=validate_time_between_check_has_values
@@ -819,8 +816,7 @@ class processor_text_json_diff_form(commonSettingsForm):
     filter_text_removed = BooleanField(_l('Removed lines'), default=True)
 
     trigger_text = StringListField(_l('Keyword triggers - Trigger/wait for text'), [validators.Optional(), ValidateListRegex()])
-    if os.getenv("PLAYWRIGHT_DRIVER_URL"):
-        browser_steps = FieldList(FormField(SingleBrowserStep), min_entries=10)
+    browser_steps = FieldList(FormField(SingleBrowserStep), min_entries=10)
     text_should_not_be_present = StringListField(_l('Block change-detection while text matches'), [validators.Optional(), ValidateListRegex()])
     webdriver_js_execute_code = TextAreaField(_l('Execute JavaScript before change detection'), render_kw={"rows": "5"}, validators=[validators.Optional()])
 
@@ -836,6 +832,8 @@ class processor_text_json_diff_form(commonSettingsForm):
     conditions = FieldList(FormField(ConditionFormRow), min_entries=1)  # Add rule logic here
     use_page_title_in_list = TernaryNoneBooleanField(_l('Use page <title> in list'), default=None)
 
+    history_snapshot_max_length = IntegerField(_l('Number of history items per watch to keep'), render_kw={"style": "width: 5em;"}, validators=[validators.Optional(), validators.NumberRange(min=2)])
+
     def extra_tab_content(self):
         return None
 
@@ -947,7 +945,7 @@ class DefaultUAInputForm(Form):
 
 # datastore.data['settings']['requests']..
 class globalSettingsRequestForm(Form):
-    time_between_check = RequiredFormField(TimeBetweenCheckForm)
+    time_between_check = RequiredFormField(TimeBetweenCheckForm, label=_l('Time Between Check'))
     time_schedule_limit = FormField(ScheduleLimitForm)
     proxy = RadioField(_l('Default proxy'))
     jitter_seconds = IntegerField(_l('Random jitter seconds ± check'),
@@ -1007,7 +1005,7 @@ class globalSettingsApplicationForm(commonSettingsForm):
         render_kw={"placeholder": "0.1", "style": "width: 8em;"}
     )
 
-    password = SaltyPasswordField()
+    password = SaltyPasswordField(_l('Password'))
     pager_size = IntegerField(_l('Pager size'),
                               render_kw={"style": "width: 5em;"},
                               validators=[validators.NumberRange(min=0,
@@ -1033,6 +1031,8 @@ class globalSettingsApplicationForm(commonSettingsForm):
                                                                   render_kw={"style": "width: 5em;"},
                                                                   validators=[validators.NumberRange(min=0,
                                                                                                      message=_l("Should contain zero or more attempts"))])
+
+    history_snapshot_max_length = IntegerField(_l('Number of history items per watch to keep'), render_kw={"style": "width: 5em;"}, validators=[validators.Optional(), validators.NumberRange(min=2)])
     ui = FormField(globalSettingsApplicationUIForm)
 
 

changedetectionio/html_tools.py

@@ -23,6 +23,53 @@ class JSONNotFound(ValueError):
     def __init__(self, msg):
         ValueError.__init__(self, msg)
 
+
+_DEFAULT_UNSAFE_XPATH3_FUNCTIONS = [
+    'unparsed-text',
+    'unparsed-text-lines',
+    'unparsed-text-available',
+    'doc',
+    'doc-available',
+    'environment-variable',
+    'available-environment-variables',
+]
+
+
+def _build_safe_xpath3_parser():
+    """Return an XPath3Parser subclass with filesystem/environment access functions removed.
+
+    XPath 3.0 includes functions that can read arbitrary files or environment variables:
+      - unparsed-text / unparsed-text-lines / unparsed-text-available  (file read)
+      - doc / doc-available                                             (XML fetch from URI)
+      - environment-variable / available-environment-variables         (env var leakage)
+
+    Subclassing gives us an independent symbol_table copy (not shared with the parent class),
+    so removing entries here does not affect XPath3Parser itself.
+
+    Override the blocked list via the XPATH_BLOCKED_FUNCTIONS environment variable
+    (comma-separated, e.g. "unparsed-text,doc,environment-variable").
+    """
+    import os
+    from elementpath.xpath3 import XPath3Parser
+
+    class SafeXPath3Parser(XPath3Parser):
+        pass
+
+    env_override = os.getenv('XPATH_BLOCKED_FUNCTIONS')
+    if env_override is not None:
+        blocked = [f.strip() for f in env_override.split(',') if f.strip()]
+    else:
+        blocked = _DEFAULT_UNSAFE_XPATH3_FUNCTIONS
+
+    for _fn in blocked:
+        SafeXPath3Parser.symbol_table.pop(_fn, None)
+
+    return SafeXPath3Parser
+
+
+# Module-level singleton — built once, reused everywhere.
+SafeXPath3Parser = _build_safe_xpath3_parser()
+
 # Doesn't look like python supports forward slash auto enclosure in re.findall
 # So convert it to inline flag "(?i)foobar" type configuration
 @lru_cache(maxsize=100)
@@ -183,8 +230,6 @@ def xpath_filter(xpath_filter, html_content, append_pretty_line_formatting=False
     """
     from lxml import etree, html
     import elementpath
-    # xpath 2.0-3.1
-    from elementpath.xpath3 import XPath3Parser
 
     parser = etree.HTMLParser()
     tree = None
@@ -210,7 +255,7 @@ def xpath_filter(xpath_filter, html_content, append_pretty_line_formatting=False
             # This allows //title to match elements in the default namespace
             namespaces[''] = tree.nsmap[None]
 
-        r = elementpath.select(tree, xpath_filter.strip(), namespaces=namespaces, parser=XPath3Parser)
+        r = elementpath.select(tree, xpath_filter.strip(), namespaces=namespaces, parser=SafeXPath3Parser)
         #@note: //title/text() now works with default namespaces (fixed by registering '' prefix)
         #@note: //title/text() wont work where <title>CDATA.. (use cdata_in_document_to_text first)
 
@@ -235,6 +280,9 @@ def xpath_filter(xpath_filter, html_content, append_pretty_line_formatting=False
             else:
                 html_block += elementpath_tostring(element)
 
+        # Drop element references before the finally block so tree.clear() can release
+        # the libxml2 document immediately (elements pin the C-level doc via refcount).
+        del r
         return html_block
     finally:
         # Explicitly clear the tree to free memory
@@ -539,6 +587,18 @@ def cdata_in_document_to_text(html_content: str, render_anchor_tag_content=False
 
 
 def html_to_text(html_content: str, render_anchor_tag_content=False, is_rss=False, timeout=10) -> str:
+    """
+    Convert HTML content to plain text using inscriptis.
+
+    Thread-Safety: This function uses inscriptis.get_text() which internally calls
+    lxml.html.fromstring() with the default parser. Testing with 50 concurrent threads
+    confirms this approach is thread-safe and produces deterministic output.
+
+    Alternative Approach Rejected: An explicit HTMLParser instance (thread-local or fresh)
+    would also be thread-safe, but was found to break change detection logic in subtle ways
+    (test_check_basic_change_detection_functionality). The default parser provides correct
+    and reliable behavior.
+    """
     from inscriptis import get_text
     from inscriptis.model.config import ParserConfig
 
@@ -549,10 +609,33 @@ def html_to_text(html_content: str, render_anchor_tag_content=False, is_rss=Fals
         )
     else:
         parser_config = None
-
     if is_rss:
         html_content = re.sub(r'<title([\s>])', r'<h1\1', html_content)
         html_content = re.sub(r'</title>', r'</h1>', html_content)
+    else:
+        # Use BS4 html.parser to strip bloat — SPA's often dump 10MB+ of CSS/JS into <head>,
+        # causing inscriptis to silently give up. Regex-based stripping is unsafe because tags
+        # can appear inside JSON data attributes with JS-escaped closing tags (e.g. <\/script>),
+        # causing the regex to scan past the intended close and eat real page content.
+        from bs4 import BeautifulSoup
+        soup = BeautifulSoup(html_content, 'html.parser')
+        # Strip tags that inscriptis cannot render as meaningful text and which can be very large.
+        # svg/math: produce path-data/MathML garbage; canvas/iframe/template: no inscriptis handlers.
+        # video/audio/picture are kept — they may contain meaningful fallback text or captions.
+        for tag in soup.find_all(['head', 'script', 'style', 'noscript', 'svg',
+                                  'math', 'canvas', 'iframe', 'template']):
+            tag.decompose()
+
+        # SPAs often use <body style="display:none"> to hide content until JS loads.
+        # inscriptis respects CSS display rules, so strip hiding styles from the body tag.
+        body_tag = soup.find('body')
+        if body_tag and body_tag.get('style'):
+            style = body_tag['style']
+            if re.search(r'\b(?:display\s*:\s*none|visibility\s*:\s*hidden)\b', style, re.IGNORECASE):
+                logger.debug(f"html_to_text: Removing hiding styles from body tag (found: '{style}')")
+                del body_tag['style']
+
+        html_content = str(soup)
 
     text_content = get_text(html_content, config=parser_config)
     return text_content

changedetectionio/jinja2_custom/safe_jinja.py

@@ -52,7 +52,13 @@ def render(template_str, **args: t.Any) -> str:
     return output[:JINJA2_MAX_RETURN_PAYLOAD_SIZE]
 
 def render_fully_escaped(content):
-    env = jinja2.sandbox.ImmutableSandboxedEnvironment(autoescape=True)
-    template = env.from_string("{{ some_html|e }}")
-    return template.render(some_html=content)
+    """
+    Escape HTML content safely.
+
+    MEMORY LEAK FIX: Use markupsafe.escape() directly instead of creating
+    Jinja2 environments (was causing 1M+ compilations per page load).
+    Simpler, faster, and no concerns about environment state.
+    """
+    from markupsafe import escape
+    return str(escape(content))
 

changedetectionio/languages.py

@@ -29,11 +29,15 @@ def get_timeago_locale(flask_locale):
     """
     locale_map = {
         'zh': 'zh_CN',      # Chinese Simplified
+        # timeago library just hasn't been updated to use the more modern locale naming convention, before BCP 47 / RFC 5646.
+        'zh_TW': 'zh_TW',   # Chinese Traditional (timeago uses zh_TW)
+        'zh_Hant_TW': 'zh_TW',  # Flask-Babel normalizes zh_TW to zh_Hant_TW, map back to timeago's zh_TW
         'pt': 'pt_PT',      # Portuguese (Portugal)
         'sv': 'sv_SE',      # Swedish
         'no': 'nb_NO',      # Norwegian Bokmål
         'hi': 'in_HI',      # Hindi
         'cs': 'en',         # Czech not supported by timeago, fallback to English
+        'uk': 'uk',         # Ukrainian
         'en_GB': 'en',      # British English - timeago uses 'en'
         'en_US': 'en',      # American English - timeago uses 'en'
     }
@@ -53,7 +57,7 @@ LANGUAGE_DATA = {
     'it': {'flag': 'fi fi-it fis', 'name': 'Italiano'},
     'ja': {'flag': 'fi fi-jp fis', 'name': '日本語'},
     'zh': {'flag': 'fi fi-cn fis', 'name': '中文 (简体)'},
-    'zh_TW': {'flag': 'fi fi-tw fis', 'name': '繁體中文'},
+    'zh_Hant_TW': {'flag': 'fi fi-tw fis', 'name': '繁體中文'},
     'ru': {'flag': 'fi fi-ru fis', 'name': 'Русский'},
     'pl': {'flag': 'fi fi-pl fis', 'name': 'Polski'},
     'nl': {'flag': 'fi fi-nl fis', 'name': 'Nederlands'},
@@ -64,6 +68,7 @@ LANGUAGE_DATA = {
     'tr': {'flag': 'fi fi-tr fis', 'name': 'Türkçe'},
     'ar': {'flag': 'fi fi-sa fis', 'name': 'العربية'},
     'hi': {'flag': 'fi fi-in fis', 'name': 'हिन्दी'},
+    'uk': {'flag': 'fi fi-ua fis', 'name': 'Українська'},
 }
 
 

changedetectionio/model/App.py

@@ -2,6 +2,7 @@ from os import getenv
 from copy import deepcopy
 
 from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES, RSS_CONTENT_FORMAT_DEFAULT
+from changedetectionio.model.Tags import TagsDict
 
 from changedetectionio.notification import (
     default_notification_body,
@@ -29,7 +30,7 @@ class model(dict):
                     'proxy': None, # Preferred proxy connection
                     'time_between_check': {'weeks': None, 'days': None, 'hours': 3, 'minutes': None, 'seconds': None},
                     'timeout': int(getenv("DEFAULT_SETTINGS_REQUESTS_TIMEOUT", "45")),  # Default 45 seconds
-                    'workers': int(getenv("DEFAULT_SETTINGS_REQUESTS_WORKERS", "10")),  # Number of threads, lower is better for slow connections
+                    'workers': int(getenv("DEFAULT_SETTINGS_REQUESTS_WORKERS", "5")),  # Number of threads, lower is better for slow connections
                     'default_ua': {
                         'html_requests': getenv("DEFAULT_SETTINGS_HEADERS_USERAGENT", DEFAULT_SETTINGS_HEADERS_USERAGENT),
                         'html_webdriver': None,
@@ -37,6 +38,8 @@ class model(dict):
                 },
                 'application': {
                     # Custom notification content
+                    'all_paused': False,
+                    'all_muted': False,
                     'api_access_token_enabled': True,
                     'base_url' : None,
                     'empty_pages_are_a_change': False,
@@ -44,6 +47,7 @@ class model(dict):
                     'filter_failure_notification_threshold_attempts': _FILTER_FAILURE_THRESHOLD_ATTEMPTS_DEFAULT,
                     'global_ignore_text': [], # List of text to ignore when calculating the comparison checksum
                     'global_subtractive_selectors': [],
+                    'history_snapshot_max_length': None,
                     'ignore_whitespace': True,
                     'ignore_status_codes': False, #@todo implement, as ternary.
                     'ssim_threshold': '0.96',  # Default SSIM threshold for screenshot comparison
@@ -65,7 +69,7 @@ class model(dict):
                     'schema_version' : 0,
                     'shared_diff_access': False,
                     'strip_ignored_lines': False,
-                    'tags': {}, #@todo use Tag.model initialisers
+                    'tags': None,  # Initialized in __init__ with real datastore_path
                     'webdriver_delay': None , # Extra delay in seconds before extracting text
                     'ui': {
                         'use_page_title_in_list': True,
@@ -77,10 +81,16 @@ class model(dict):
             }
         }
 
-    def __init__(self, *arg, **kw):
+    def __init__(self, *arg, datastore_path=None, **kw):
         super(model, self).__init__(*arg, **kw)
+        # Capture any tags data passed in before base_config overwrites the structure
+        existing_tags = self.get('settings', {}).get('application', {}).get('tags') or {}
         # CRITICAL: deepcopy to avoid sharing mutable objects between instances
         self.update(deepcopy(self.base_config))
+        # TagsDict requires the real datastore_path at runtime (cannot be set at class-definition time)
+        if datastore_path is None:
+            raise ValueError("App.model() requires 'datastore_path' keyword argument")
+        self['settings']['application']['tags'] = TagsDict(existing_tags, datastore_path=datastore_path)
 
 
 def parse_headers_from_text_file(filepath):

changedetectionio/model/Tag.py

@@ -1,10 +1,48 @@
+"""
+Tag/Group domain model for organizing and overriding watch settings.
+
+ARCHITECTURE NOTE: Configuration Override Hierarchy
+===================================================
+
+Tags can override Watch settings when overrides_watch=True.
+Current implementation requires manual checking in processors:
+
+    for tag_uuid in watch.get('tags'):
+        tag = datastore['settings']['application']['tags'][tag_uuid]
+        if tag.get('overrides_watch'):
+            restock_settings = tag.get('restock_settings', {})
+            break
+
+With Pydantic, this would be automatic via chain resolution:
+    Watch → Tag (first with overrides_watch) → Global
+
+See: Watch.py model docstring for full Pydantic architecture explanation
+See: processors/restock_diff/processor.py:184-192 for current manual implementation
+"""
 
 from changedetectionio.model import watch_base
+from changedetectionio.model.persistence import EntityPersistenceMixin
 
+class model(EntityPersistenceMixin, watch_base):
+    """
+    Tag domain model - groups watches and can override their settings.
 
-class model(watch_base):
+    Tags inherit from watch_base to reuse all the same fields as Watch.
+    When overrides_watch=True, tag settings take precedence over watch settings
+    for all watches in this tag/group.
+
+    Fields:
+        overrides_watch (bool): If True, this tag's settings override watch settings
+        title (str): Display name for this tag/group
+        uuid (str): Unique identifier
+        ... (all fields from watch_base can be set as tag-level overrides)
+
+    Resolution order when overrides_watch=True:
+        Watch.field → Tag.field (if overrides_watch) → Global.field
+    """
 
     def __init__(self, *arg, **kw):
+        # Parent class (watch_base) handles __datastore and __datastore_path
         super(model, self).__init__(*arg, **kw)
 
         self['overrides_watch'] = kw.get('default', {}).get('overrides_watch')
@@ -12,3 +50,7 @@ class model(watch_base):
         if kw.get('default'):
             self.update(kw['default'])
             del kw['default']
+
+    # _save_to_disk() method provided by EntityPersistenceMixin
+    # commit() and _get_commit_data() methods inherited from watch_base
+    # Tag uses default _get_commit_data() (includes all keys)

changedetectionio/model/Tags.py

@@ -0,0 +1,39 @@
+import os
+import shutil
+from pathlib import Path
+from loguru import logger
+
+_SENTINEL = object()
+
+
+class TagsDict(dict):
+    """Dict subclass that removes the corresponding tag.json file when a tag is deleted."""
+
+    def __init__(self, *args, datastore_path: str | os.PathLike, **kwargs) -> None:
+        self._datastore_path = Path(datastore_path)
+        super().__init__(*args, **kwargs)
+
+    def __delitem__(self, key: str) -> None:
+        super().__delitem__(key)
+        tag_dir = self._datastore_path / key
+        tag_json_file = tag_dir / "tag.json"
+        if not os.path.exists(tag_json_file):
+            logger.critical(f"Aborting deletion of directory '{tag_dir}' because '{tag_json_file}' does not exist.")
+            return
+        try:
+            shutil.rmtree(tag_dir)
+            logger.info(f"Deleted tag directory for tag {key!r}")
+        except FileNotFoundError:
+            pass
+        except OSError as e:
+            logger.error(f"Failed to delete tag directory for tag {key!r}: {e}")
+
+    def pop(self, key: str, default=_SENTINEL):
+        """Remove and return tag, deleting its tag.json file. Raises KeyError if missing and no default given."""
+        if key in self:
+            value = self[key]
+            del self[key]
+            return value
+        if default is _SENTINEL:
+            raise KeyError(key)
+        return default

changedetectionio/model/Watch.py

@@ -1,63 +1,66 @@
+"""
+Watch domain model for change detection monitoring.
+
+ARCHITECTURE NOTE: Configuration Override Hierarchy
+===================================================
+
+This module implements Watch objects that inherit from dict (technical debt).
+The dream architecture would use Pydantic for:
+
+1. CHAIN RESOLUTION (Watch → Tag → Global Settings)
+   - Current: Manual resolution scattered across codebase
+   - Future: @computed_field properties with automatic resolution
+   - Examples: resolved_fetch_backend, resolved_restock_settings, etc.
+
+2. DATABASE BACKEND ABSTRACTION
+   - Current: Domain model tightly coupled to file-based JSON storage
+   - Future: Domain model (Pydantic) separate from persistence layer
+   - Enables: Easy migration to PostgreSQL, MongoDB, etc.
+
+3. TYPE SAFETY & VALIDATION
+   - Current: Dict access with no compile-time checks
+   - Future: Type hints, IDE autocomplete, validation at boundaries
+
+See class model docstring for detailed explanation and examples.
+See: processors/restock_diff/processor.py:184-192 for manual resolution example
+"""
+
 from blinker import signal
 from changedetectionio.validate_url import is_safe_valid_url
 
 from changedetectionio.strtobool import strtobool
 from changedetectionio.jinja2_custom import render as jinja_render
 from . import watch_base
+from .persistence import EntityPersistenceMixin
 import os
 import re
 from pathlib import Path
 from loguru import logger
 
 from .. import jinja2_custom as safe_jinja
-from ..diff import ADDED_PLACEMARKER_OPEN
 from ..html_tools import TRANSLATE_WHITESPACE_TABLE
 
+FAVICON_RESAVE_THRESHOLD_SECONDS=86400
+BROTLI_COMPRESS_SIZE_THRESHOLD = int(os.getenv('SNAPSHOT_BROTLI_COMPRESSION_THRESHOLD', 1024*20))
 
-def _brotli_compress_worker(conn, filepath, mode=None):
+# Module-level favicon filename cache: data_dir → basename (or None)
+# Keyed by data_dir so it survives Watch object recreation, deepcopy, and concurrent requests.
+# Invalidated explicitly in bump_favicon() when a new favicon is saved.
+_FAVICON_FILENAME_CACHE: dict = {}
+
+minimum_seconds_recheck_time = int(os.getenv('MINIMUM_SECONDS_RECHECK_TIME', 3))
+mtable = {'seconds': 1, 'minutes': 60, 'hours': 3600, 'days': 86400, 'weeks': 86400 * 7}
+
+def _brotli_save(contents, filepath, mode=None, fallback_uncompressed=False):
     """
-    Worker function to compress data with brotli in a separate process.
-    This isolates memory - when process exits, OS reclaims all memory.
-
-    Args:
-        conn: multiprocessing.Pipe connection to receive data
-        filepath: destination file path
-        mode: brotli compression mode (e.g., brotli.MODE_TEXT)
-    """
-    import brotli
-
-    try:
-        # Receive data from parent process via pipe (avoids pickle overhead)
-        contents = conn.recv()
-
-        if mode is not None:
-            compressed_data = brotli.compress(contents, mode=mode)
-        else:
-            compressed_data = brotli.compress(contents)
-
-        with open(filepath, 'wb') as f:
-            f.write(compressed_data)
-
-        # Send success status back
-        conn.send(True)
-        # No need for explicit cleanup - process exit frees all memory
-    except Exception as e:
-        logger.error(f"Brotli compression worker failed: {e}")
-        conn.send(False)
-    finally:
-        conn.close()
-
-
-def _brotli_subprocess_save(contents, filepath, mode=None, timeout=30, fallback_uncompressed=False):
-    """
-    Save compressed data using subprocess to isolate memory.
-    Uses Pipe to avoid pickle overhead for large data.
+    Save compressed data using native brotli with streaming compression.
+    Uses chunked compression to minimize peak memory usage and malloc_trim()
+    to force release of C-level memory back to the OS.
 
     Args:
         contents: data to compress (str or bytes)
         filepath: destination file path
         mode: brotli compression mode (e.g., brotli.MODE_TEXT)
-        timeout: subprocess timeout in seconds
         fallback_uncompressed: if True, save uncompressed on failure; if False, raise exception
 
     Returns:
@@ -67,100 +70,175 @@ def _brotli_subprocess_save(contents, filepath, mode=None, timeout=30, fallback_
         Exception: if compression fails and fallback_uncompressed is False
     """
     import brotli
-    import multiprocessing
-    import sys
+    import gc
+    import ctypes
 
     # Ensure contents are bytes
     if isinstance(contents, str):
         contents = contents.encode('utf-8')
 
-    # Use explicit spawn context for thread safety (avoids fork() with multi-threaded parent)
-    # Always use spawn - consistent behavior in tests and production
-    ctx = multiprocessing.get_context('spawn')
-    parent_conn, child_conn = ctx.Pipe()
-
-    # Run compression in subprocess using spawn (not fork)
-    proc = ctx.Process(target=_brotli_compress_worker, args=(child_conn, filepath, mode))
-
-    # Windows-safe: Set daemon=False explicitly to avoid issues with process cleanup
-    proc.daemon = False
-    proc.start()
-
     try:
-        # Send data to subprocess via pipe (avoids pickle)
-        parent_conn.send(contents)
+        original_size = len(contents)
+        logger.debug(f"Starting brotli streaming compression of {original_size} bytes.")
 
-        # Wait for result with timeout
-        if parent_conn.poll(timeout):
-            success = parent_conn.recv()
-        else:
-            success = False
-            logger.warning(f"Brotli compression subprocess timed out after {timeout}s")
-            # Graceful termination with platform-aware cleanup
-            try:
-                proc.terminate()
-            except Exception as term_error:
-                logger.debug(f"Process termination issue (may be normal on Windows): {term_error}")
+        # Create streaming compressor
+        compressor = brotli.Compressor(quality=6, mode=mode if mode is not None else brotli.MODE_GENERIC)
 
-        parent_conn.close()
-        proc.join(timeout=5)
+        # Stream compress in chunks to minimize memory usage
+        chunk_size = 65536  # 64KB chunks
+        total_compressed_size = 0
 
-        # Force kill if still alive after graceful termination
-        if proc.is_alive():
-            try:
-                if sys.platform == 'win32':
-                    # Windows: use kill() which is more forceful
-                    proc.kill()
-                else:
-                    # Unix: terminate() already sent SIGTERM, now try SIGKILL
-                    proc.kill()
-                proc.join(timeout=2)
-            except Exception as kill_error:
-                logger.warning(f"Failed to kill brotli compression process: {kill_error}")
+        with open(filepath, 'wb') as f:
+            # Process data in chunks
+            offset = 0
+            while offset < len(contents):
+                chunk = contents[offset:offset + chunk_size]
+                compressed_chunk = compressor.process(chunk)
+                if compressed_chunk:
+                    f.write(compressed_chunk)
+                    total_compressed_size += len(compressed_chunk)
+                offset += chunk_size
 
-        # Check if file was created successfully
-        if success and os.path.exists(filepath):
-            return filepath
+            # Finalize compression - critical for proper cleanup
+            final_chunk = compressor.finish()
+            if final_chunk:
+                f.write(final_chunk)
+                total_compressed_size += len(final_chunk)
+
+        logger.debug(f"Finished brotli compression - From {original_size} to {total_compressed_size} bytes.")
+
+        # Cleanup: Delete compressor, force Python GC, then force C-level memory release
+        del compressor
+        gc.collect()
+
+        # Force release of C-level memory back to OS (since brotli is a C library)
+        try:
+            ctypes.CDLL('libc.so.6').malloc_trim(0)
+        except Exception:
+            pass  # malloc_trim not available on all systems (e.g., macOS)
+
+        return filepath
 
     except Exception as e:
         logger.error(f"Brotli compression error: {e}")
-        try:
-            parent_conn.close()
-        except:
-            pass
-        try:
-            proc.terminate()
-            proc.join(timeout=2)
-        except:
-            pass
 
-    # Compression failed
-    if fallback_uncompressed:
-        logger.warning(f"Brotli compression failed for {filepath}, saving uncompressed")
-        fallback_path = filepath.replace('.br', '')
-        with open(fallback_path, 'wb') as f:
-            f.write(contents)
-        return fallback_path
-    else:
-        raise Exception(f"Brotli compression subprocess failed for {filepath}")
-
-FAVICON_RESAVE_THRESHOLD_SECONDS=86400
+        # Compression failed
+        if fallback_uncompressed:
+            logger.warning(f"Brotli compression failed for {filepath}, saving uncompressed")
+            fallback_path = filepath.replace('.br', '')
+            with open(fallback_path, 'wb') as f:
+                f.write(contents)
+            return fallback_path
+        else:
+            raise Exception(f"Brotli compression failed for {filepath}: {e}")
 
 
-minimum_seconds_recheck_time = int(os.getenv('MINIMUM_SECONDS_RECHECK_TIME', 3))
-mtable = {'seconds': 1, 'minutes': 60, 'hours': 3600, 'days': 86400, 'weeks': 86400 * 7}
+class model(EntityPersistenceMixin, watch_base):
+    """
+    Watch domain model for monitoring URL changes.
 
-class model(watch_base):
+    Inherits from watch_base (which inherits dict) - see watch_base docstring for field documentation.
+
+    ## Configuration Override Hierarchy (Chain Resolution)
+
+    The dream architecture uses a 3-level resolution chain:
+        Watch settings → Tag/Group settings → Global settings
+
+    Current implementation is MANUAL (see processor.py:184-192 for example):
+        - Processors manually check watch.get('field')
+        - Then loop through watch.tags to find first tag with overrides_watch=True
+        - Finally fall back to datastore['settings']['application']['field']
+
+    FUTURE: Pydantic-based chain resolution would enable:
+
+        ```python
+        # Instead of manual resolution in every processor:
+        restock_settings = watch.get('restock_settings', {})
+        for tag_uuid in watch.get('tags'):
+            tag = datastore['settings']['application']['tags'][tag_uuid]
+            if tag.get('overrides_watch'):
+                restock_settings = tag.get('restock_settings', {})
+                break
+
+        # Clean computed properties with automatic resolution:
+        @computed_field
+        def resolved_restock_settings(self) -> dict:
+            if self.restock_settings:
+                return self.restock_settings
+            for tag_uuid in self.tags:
+                tag = self._datastore.get_tag(tag_uuid)
+                if tag.overrides_watch and tag.restock_settings:
+                    return tag.restock_settings
+            return self._datastore.settings.restock_settings or {}
+
+        # Usage: watch.resolved_restock_settings (automatic, type-safe, tested once)
+        ```
+
+    Benefits of Pydantic migration:
+        1. Single source of truth for resolution logic (not scattered across processors)
+        2. Type safety + IDE autocomplete (watch.resolved_fetch_backend vs dict navigation)
+        3. Database backend abstraction (domain model separate from persistence)
+        4. Automatic validation at boundaries
+        5. Self-documenting via type hints
+        6. Easy to test resolution independently
+
+    Resolution chain examples that would benefit:
+        - fetch_backend: watch → tag → global (see get_fetch_backend property)
+        - notification_urls: watch → tag → global
+        - time_between_check: watch → global (see threshold_seconds)
+        - restock_settings: watch → tag (see processors/restock_diff/processor.py:184-192)
+        - history_snapshot_max_length: watch → global (see save_history_blob:550-556)
+        - All processor_config_* settings could use tag overrides
+
+    ## Database Backend Abstraction with Pydantic
+
+    Current: Watch inherits dict, tightly coupled to file-based JSON storage
+    Future: Domain model (Watch) separate from persistence layer
+
+        ```python
+        # Domain model (database-agnostic)
+        class Watch(BaseModel):
+            uuid: str
+            url: str
+            # ... validation, business logic
+
+        # Pluggable backends
+        class DataStoreBackend(ABC):
+            def save_watch(self, watch: Watch): ...
+            def load_watch(self, uuid: str) -> Watch: ...
+
+        # Implementations: FileBackend, MongoBackend, PostgresBackend, etc.
+        ```
+
+    This would enable:
+        - Easy migration between storage backends (file → postgres → mongodb)
+        - Pydantic handles serialization/deserialization automatically
+        - Domain logic stays clean (no storage concerns in Watch methods)
+
+    ## Migration Path
+
+    Given existing codebase, incremental migration recommended:
+        1. Create Pydantic models alongside existing dict-based models
+        2. Add .to_pydantic() / .from_pydantic() bridge methods
+        3. Gradually migrate code to use Pydantic models
+        4. Remove dict inheritance once migration complete
+
+    See: watch_base docstring for technical debt discussion
+    See: processors/restock_diff/processor.py:184-192 for manual resolution example
+    See: Watch.py:550-556 for nested dict navigation that would become watch.resolved_*
+    """
     __newest_history_key = None
     __history_n = 0
     jitter_seconds = 0
 
     def __init__(self, *arg, **kw):
-        self.__datastore_path = kw.get('datastore_path')
-        if kw.get('datastore_path'):
-            del kw['datastore_path']
-            
+        # Validate __datastore before calling parent (Watch requires it)
+        if not kw.get('__datastore'):
+            raise ValueError("Watch object requires '__datastore' reference - cannot access global settings without it")
+
+        # Parent class (watch_base) handles __datastore and __datastore_path
         super(model, self).__init__(*arg, **kw)
+
         if kw.get('default'):
             self.update(kw['default'])
             del kw['default']
@@ -171,6 +249,9 @@ class model(watch_base):
         # Be sure the cached timestamp is ready
         bump = self.history
 
+    # Note: __deepcopy__, __getstate__, and __setstate__ are inherited from watch_base
+    # This prevents memory leaks by sharing __datastore reference instead of copying it
+
     @property
     def viewed(self):
         # Don't return viewed when last_viewed is 0 and newest_key is 0
@@ -183,11 +264,6 @@ class model(watch_base):
     def has_unviewed(self):
         return int(self.newest_history_key) > int(self['last_viewed']) and self.__history_n >= 2
 
-    def ensure_data_dir_exists(self):
-        if not os.path.isdir(self.watch_data_dir):
-            logger.debug(f"> Creating data dir {self.watch_data_dir}")
-            os.mkdir(self.watch_data_dir)
-
     @property
     def link(self):
 
@@ -243,7 +319,8 @@ class model(watch_base):
 
         # JSON Data, Screenshots, Textfiles (history index and snapshots), HTML in the future etc
         # But preserve processor config files (they're configuration, not history data)
-        for item in pathlib.Path(str(self.watch_data_dir)).rglob("*.*"):
+        # Use glob not rglob here for safety.
+        for item in pathlib.Path(str(self.data_dir)).glob("*.*"):
             # Skip processor config files
             if item.name in processor_config_files:
                 continue
@@ -263,7 +340,6 @@ class model(watch_base):
             'last_notification_error': False,
             'last_viewed': 0,
             'previous_md5': False,
-            'previous_md5_before_filters': False,
             'remote_server_reply': None,
             'track_ldjson_price_data': None
         })
@@ -280,8 +356,30 @@ class model(watch_base):
     @property
     def get_fetch_backend(self):
         """
-        Like just using the `fetch_backend` key but there could be some logic
-        :return:
+        Get the fetch backend for this watch with special case handling.
+
+        CHAIN RESOLUTION OPPORTUNITY:
+        Currently returns watch.fetch_backend directly, but doesn't implement
+        Watch → Tag → Global resolution chain. With Pydantic:
+
+        @computed_field
+        def resolved_fetch_backend(self) -> str:
+            # Special case: PDFs always use html_requests
+            if self.is_pdf:
+                return 'html_requests'
+
+            # Watch override
+            if self.fetch_backend and self.fetch_backend != 'system':
+                return self.fetch_backend
+
+            # Tag override (first tag with overrides_watch=True wins)
+            for tag_uuid in self.tags:
+                tag = self._datastore.get_tag(tag_uuid)
+                if tag.overrides_watch and tag.fetch_backend:
+                    return tag.fetch_backend
+
+            # Global default
+            return self._datastore.settings.fetch_backend
         """
         # Maybe also if is_image etc?
         # This is because chrome/playwright wont render the PDF in the browser and we will just fetch it and use pdf2html to see the text.
@@ -292,10 +390,16 @@ class model(watch_base):
 
     @property
     def is_pdf(self):
-        # content_type field is set in the future
-        # https://github.com/dgtlmoon/changedetection.io/issues/1392
-        # Not sure the best logic here
-        return self.get('url', '').lower().endswith('.pdf') or 'pdf' in self.get('content_type', '').lower()
+        url = str(self.get("url") or "").lower()
+        content_type = str(self.get("content-type") or "").lower()
+
+        if content_type in ("none", "null", ""):
+            content_type = ""
+
+        return (
+                url.endswith(".pdf")
+                or content_type.split(";")[0].strip() == "application/pdf"
+        )
 
     @property
     def label(self):
@@ -330,11 +434,11 @@ class model(watch_base):
         tmp_history = {}
 
         # In the case we are only using the watch for processing without history
-        if not self.watch_data_dir:
+        if not self.data_dir:
             return []
 
         # Read the history file as a dict
-        fname = os.path.join(self.watch_data_dir, self.history_index_filename)
+        fname = os.path.join(self.data_dir, self.history_index_filename)
         if os.path.isfile(fname):
             logger.debug(f"Reading watch history index for {self.get('uuid')}")
             with open(fname, "r", encoding='utf-8') as f:
@@ -347,13 +451,13 @@ class model(watch_base):
                         # Cross-platform: check for any path separator (works on Windows and Unix)
                         if os.sep not in v and '/' not in v and '\\' not in v:
                             # Relative filename only, no path separators
-                            v = os.path.join(self.watch_data_dir, v)
+                            v = os.path.join(self.data_dir, v)
                         else:
                             # It's possible that they moved the datadir on older versions
                             # So the snapshot exists but is in a different path
                             # Cross-platform: use os.path.basename instead of split('/')
                             snapshot_fname = os.path.basename(v)
-                            proposed_new_path = os.path.join(self.watch_data_dir, snapshot_fname)
+                            proposed_new_path = os.path.join(self.data_dir, snapshot_fname)
                             if not os.path.exists(v) and os.path.exists(proposed_new_path):
                                 v = proposed_new_path
 
@@ -370,7 +474,7 @@ class model(watch_base):
 
     @property
     def has_history(self):
-        fname = os.path.join(self.watch_data_dir, self.history_index_filename)
+        fname = os.path.join(self.data_dir, self.history_index_filename)
         return os.path.isfile(fname)
 
     @property
@@ -473,16 +577,49 @@ class model(watch_base):
         with open(filepath, 'r', encoding='utf-8', errors='ignore') as f:
             return f.read()
 
-    def _write_atomic(self, dest, data):
+    def _write_atomic(self, dest, data, mode='wb'):
         """Write data atomically to dest using a temp file"""
-        if not os.path.exists(dest):
-            import tempfile
-            with tempfile.NamedTemporaryFile('wb', delete=False, dir=self.watch_data_dir) as tmp:
-                tmp.write(data)
-                tmp.flush()
-                os.fsync(tmp.fileno())
-                tmp_path = tmp.name
-            os.replace(tmp_path, dest)
+        import tempfile
+        with tempfile.NamedTemporaryFile(mode, delete=False, dir=self.data_dir) as tmp:
+            tmp.write(data)
+            tmp.flush()
+            os.fsync(tmp.fileno())
+            tmp_path = tmp.name
+        os.replace(tmp_path, dest)
+
+    def history_trim(self, newest_n_items):
+        from pathlib import Path
+        import gc
+        # Sort by timestamp (key)
+        sorted_items = sorted(self.history.items(), key=lambda x: int(x[0]))
+
+        keep_part = dict(sorted_items[-newest_n_items:])
+        delete_part = dict(sorted_items[:-newest_n_items])
+        logger.info( f"[{self.get('uuid')}] Trimming history to most recent {newest_n_items} items, keeping {len(keep_part)} items deleting {len(delete_part)} items.")
+
+        if delete_part:
+            for item in delete_part.items():
+                try:
+                    Path(item[1]).unlink(missing_ok=True)
+                except Exception as e:
+                    logger.critical(f"{str(e)}")
+                finally:
+                    logger.debug(f"[{self.get('uuid')}] Deleted {item[1]} history snapshot")
+        try:
+            dest = os.path.join(self.data_dir, self.history_index_filename)
+            output = "\r\n".join(
+                f"{k},{Path(v).name}"
+                for k, v in keep_part.items()
+            )+"\r\n"
+            self._write_atomic(dest=dest, data=output, mode='w')
+        except Exception as e:
+            logger.critical(f"{str(e)}")
+        finally:
+            logger.debug(f"[{self.get('uuid')}] Updated history index {dest}")
+
+        # reimport
+        bump = self.history
+        gc.collect()
 
     # Save some text file to the appropriate path and bump the history
     # result_obj from fetch_site_status.run()
@@ -491,8 +628,6 @@ class model(watch_base):
         logger.trace(f"{self.get('uuid')} - Updating {self.history_index_filename} with timestamp {timestamp}")
 
         self.ensure_data_dir_exists()
-
-        threshold = int(os.getenv('SNAPSHOT_BROTLI_COMPRESSION_THRESHOLD', 1024))
         skip_brotli = strtobool(os.getenv('DISABLE_BROTLI_TEXT_SNAPSHOT', 'False'))
 
         # Binary data - detect file type and save without compression
@@ -510,37 +645,37 @@ class model(watch_base):
                 ext = 'bin'
 
             snapshot_fname = f"{snapshot_id}.{ext}"
-            dest = os.path.join(self.watch_data_dir, snapshot_fname)
+            dest = os.path.join(self.data_dir, snapshot_fname)
             self._write_atomic(dest, contents)
             logger.trace(f"Saved binary snapshot as {snapshot_fname} ({len(contents)} bytes)")
 
         # Text data - use brotli compression if enabled and above threshold
         else:
-            if not skip_brotli and len(contents) > threshold:
+            if not skip_brotli and len(contents) > BROTLI_COMPRESS_SIZE_THRESHOLD:
                 # Compressed text
                 import brotli
                 snapshot_fname = f"{snapshot_id}.txt.br"
-                dest = os.path.join(self.watch_data_dir, snapshot_fname)
+                dest = os.path.join(self.data_dir, snapshot_fname)
 
                 if not os.path.exists(dest):
                     try:
-                        actual_dest = _brotli_subprocess_save(contents, dest, mode=brotli.MODE_TEXT, fallback_uncompressed=True)
+                        actual_dest = _brotli_save(contents, dest, mode=brotli.MODE_TEXT, fallback_uncompressed=True)
                         if actual_dest != dest:
                             snapshot_fname = os.path.basename(actual_dest)
                     except Exception as e:
                         logger.error(f"{self.get('uuid')} - Brotli compression failed: {e}")
                         # Fallback to uncompressed
                         snapshot_fname = f"{snapshot_id}.txt"
-                        dest = os.path.join(self.watch_data_dir, snapshot_fname)
+                        dest = os.path.join(self.data_dir, snapshot_fname)
                         self._write_atomic(dest, contents.encode('utf-8'))
             else:
                 # Plain text
                 snapshot_fname = f"{snapshot_id}.txt"
-                dest = os.path.join(self.watch_data_dir, snapshot_fname)
+                dest = os.path.join(self.data_dir, snapshot_fname)
                 self._write_atomic(dest, contents.encode('utf-8'))
 
         # Append to history.txt atomically
-        index_fname = os.path.join(self.watch_data_dir, self.history_index_filename)
+        index_fname = os.path.join(self.data_dir, self.history_index_filename)
         index_line = f"{timestamp},{snapshot_fname}\n"
 
         with open(index_fname, 'a', encoding='utf-8') as f:
@@ -552,6 +687,17 @@ class model(watch_base):
         self.__newest_history_key = timestamp
         self.__history_n += 1
 
+        # MANUAL CHAIN RESOLUTION: Watch → Global
+        # With Pydantic, this would become: maxlen = watch.resolved_history_snapshot_max_length
+        # @computed_field def resolved_history_snapshot_max_length(self) -> Optional[int]:
+        #     if self.history_snapshot_max_length: return self.history_snapshot_max_length
+        #     if tag := self._get_override_tag(): return tag.history_snapshot_max_length
+        #     return self._datastore.settings.history_snapshot_max_length
+        maxlen = self.get('history_snapshot_max_length') or self.get_global_setting('application', 'history_snapshot_max_length')
+
+        if maxlen and self.__history_n and self.__history_n > maxlen:
+            self.history_trim(newest_n_items=maxlen)
+
         # @todo bump static cache of the last timestamp so we dont need to examine the file to set a proper ''viewed'' status
         return snapshot_fname
 
@@ -604,7 +750,7 @@ class model(watch_base):
         return not local_lines.issubset(existing_history)
 
     def get_screenshot(self):
-        fname = os.path.join(self.watch_data_dir, "last-screenshot.png")
+        fname = os.path.join(self.data_dir, "last-screenshot.png")
         if os.path.isfile(fname):
             return fname
 
@@ -619,7 +765,7 @@ class model(watch_base):
         if not favicon_fname:
             return True
         try:
-            fname = next(iter(glob.glob(os.path.join(self.watch_data_dir, "favicon.*"))), None)
+            fname = next(iter(glob.glob(os.path.join(self.data_dir, "favicon.*"))), None)
             logger.trace(f"Favicon file maybe found at {fname}")
             if os.path.isfile(fname):
                 file_age = int(time.time() - os.path.getmtime(fname))
@@ -652,7 +798,7 @@ class model(watch_base):
             base = "favicon"
             extension = "ico"
 
-        fname = os.path.join(self.watch_data_dir, f"favicon.{extension}")
+        fname = os.path.join(self.data_dir, f"favicon.{extension}")
 
         try:
             # validate=True makes sure the string only contains valid base64 chars
@@ -664,6 +810,10 @@ class model(watch_base):
                 try:
                     with open(fname, 'wb') as f:
                         f.write(decoded)
+
+                    # Invalidate module-level favicon filename cache for this watch
+                    _FAVICON_FILENAME_CACHE.pop(self.data_dir, None)
+
                     # A signal that could trigger the socket server to update the browser also
                     watch_check_update = signal('watch_favicon_bump')
                     if watch_check_update:
@@ -677,23 +827,23 @@ class model(watch_base):
 
     def get_favicon_filename(self) -> str | None:
         """
-        Find any favicon.* file in the current working directory
-        and return the contents of the newest one.
+        Find any favicon.* file in the watch data directory.
+
+        Uses a module-level cache keyed by data_dir to survive Watch object recreation,
+        deepcopy (which drops instance attrs), and concurrent request races.
+        Invalidated by bump_favicon() when a new favicon is saved.
 
         Returns:
-            bytes: Contents of the newest favicon file, or None if not found.
+            str: Basename of the favicon file, or None if not found.
         """
+        if self.data_dir in _FAVICON_FILENAME_CACHE:
+            return _FAVICON_FILENAME_CACHE[self.data_dir]
+
         import glob
-
-        # Search for all favicon.* files
-        files = glob.glob(os.path.join(self.watch_data_dir, "favicon.*"))
-
-        if not files:
-            return None
-
-        # Find the newest by modification time
-        newest_file = max(files, key=os.path.getmtime)
-        return os.path.basename(newest_file)
+        files = glob.glob(os.path.join(self.data_dir, "favicon.*"))
+        fname = os.path.basename(files[0]) if files else None
+        _FAVICON_FILENAME_CACHE[self.data_dir] = fname
+        return fname
 
     def get_screenshot_as_thumbnail(self, max_age=3200):
         """Return path to a square thumbnail of the most recent screenshot.
@@ -709,7 +859,7 @@ class model(watch_base):
         import os
         import time
 
-        thumbnail_path = os.path.join(self.watch_data_dir, "thumbnail.jpeg")
+        thumbnail_path = os.path.join(self.data_dir, "thumbnail.jpeg")
         top_trim = 500  # Pixels from top of screenshot to use
 
         screenshot_path = self.get_screenshot()
@@ -760,7 +910,7 @@ class model(watch_base):
             return None
 
     def __get_file_ctime(self, filename):
-        fname = os.path.join(self.watch_data_dir, filename)
+        fname = os.path.join(self.data_dir, filename)
         if os.path.isfile(fname):
             return int(os.path.getmtime(fname))
         return False
@@ -785,14 +935,9 @@ class model(watch_base):
     def snapshot_error_screenshot_ctime(self):
         return self.__get_file_ctime('last-error-screenshot.png')
 
-    @property
-    def watch_data_dir(self):
-        # The base dir of the watch data
-        return os.path.join(self.__datastore_path, self['uuid']) if self.__datastore_path else None
-
     def get_error_text(self):
         """Return the text saved from a previous request that resulted in a non-200 error"""
-        fname = os.path.join(self.watch_data_dir, "last-error.txt")
+        fname = os.path.join(self.data_dir, "last-error.txt")
         if os.path.isfile(fname):
             with open(fname, 'r', encoding='utf-8') as f:
                 return f.read()
@@ -800,7 +945,7 @@ class model(watch_base):
 
     def get_error_snapshot(self):
         """Return path to the screenshot that resulted in a non-200 error"""
-        fname = os.path.join(self.watch_data_dir, "last-error-screenshot.png")
+        fname = os.path.join(self.data_dir, "last-error-screenshot.png")
         if os.path.isfile(fname):
             return fname
         return False
@@ -824,6 +969,37 @@ class model(watch_base):
     def toggle_mute(self):
         self['notification_muted'] ^= True
 
+    def _get_commit_data(self):
+        """
+        Prepare watch data for commit.
+
+        Excludes processor_config_* keys (stored in separate files).
+        Normalizes browser_steps to empty list if no meaningful steps.
+        """
+        import copy
+
+        # Get base snapshot with lock
+        lock = self._datastore.lock if self._datastore and hasattr(self._datastore, 'lock') else None
+
+        if lock:
+            with lock:
+                snapshot = dict(self)
+        else:
+            snapshot = dict(self)
+
+        # Exclude processor config keys (stored separately)
+        watch_dict = {k: copy.deepcopy(v) for k, v in snapshot.items() if not k.startswith('processor_config_')}
+
+        # Normalize browser_steps: if no meaningful steps, save as empty list
+        if not self.has_browser_steps:
+            watch_dict['browser_steps'] = []
+
+        return watch_dict
+
+    # _save_to_disk() method provided by EntityPersistenceMixin
+    # commit() method inherited from watch_base
+
+
     def extra_notification_token_values(self):
         # Used for providing extra tokens
         # return {'widget': 555}
@@ -853,7 +1029,7 @@ class model(watch_base):
                         if not csv_writer:
                             # A file on the disk can be transferred much faster via flask than a string reply
                             csv_output_filename = f"report-{self.get('uuid')}.csv"
-                            f = open(os.path.join(self.watch_data_dir, csv_output_filename), 'w')
+                            f = open(os.path.join(self.data_dir, csv_output_filename), 'w')
                             # @todo some headers in the future
                             #fieldnames = ['Epoch seconds', 'Date']
                             csv_writer = csv.writer(f,
@@ -895,7 +1071,7 @@ class model(watch_base):
 
     def save_error_text(self, contents):
         self.ensure_data_dir_exists()
-        target_path = os.path.join(self.watch_data_dir, "last-error.txt")
+        target_path = os.path.join(self.data_dir, "last-error.txt")
         with open(target_path, 'w', encoding='utf-8') as f:
             f.write(contents)
 
@@ -904,9 +1080,9 @@ class model(watch_base):
         import zlib
 
         if as_error:
-            target_path = os.path.join(str(self.watch_data_dir), "elements-error.deflate")
+            target_path = os.path.join(str(self.data_dir), "elements-error.deflate")
         else:
-            target_path = os.path.join(str(self.watch_data_dir), "elements.deflate")
+            target_path = os.path.join(str(self.data_dir), "elements.deflate")
 
         self.ensure_data_dir_exists()
 
@@ -921,9 +1097,9 @@ class model(watch_base):
     def save_screenshot(self, screenshot: bytes, as_error=False):
 
         if as_error:
-            target_path = os.path.join(self.watch_data_dir, "last-error-screenshot.png")
+            target_path = os.path.join(self.data_dir, "last-error-screenshot.png")
         else:
-            target_path = os.path.join(self.watch_data_dir, "last-screenshot.png")
+            target_path = os.path.join(self.data_dir, "last-screenshot.png")
 
         self.ensure_data_dir_exists()
 
@@ -934,7 +1110,7 @@ class model(watch_base):
 
     def get_last_fetched_text_before_filters(self):
         import brotli
-        filepath = os.path.join(self.watch_data_dir, 'last-fetched.br')
+        filepath = os.path.join(self.data_dir, 'last-fetched.br')
 
         if not os.path.isfile(filepath) or os.path.getsize(filepath) == 0:
             # If a previous attempt doesnt yet exist, just snarf the previous snapshot instead
@@ -949,21 +1125,21 @@ class model(watch_base):
 
     def save_last_text_fetched_before_filters(self, contents):
         import brotli
-        filepath = os.path.join(self.watch_data_dir, 'last-fetched.br')
-        _brotli_subprocess_save(contents, filepath, mode=brotli.MODE_TEXT, fallback_uncompressed=False)
+        filepath = os.path.join(self.data_dir, 'last-fetched.br')
+        _brotli_save(contents, filepath, mode=brotli.MODE_TEXT, fallback_uncompressed=False)
 
     def save_last_fetched_html(self, timestamp, contents):
         self.ensure_data_dir_exists()
         snapshot_fname = f"{timestamp}.html.br"
-        filepath = os.path.join(self.watch_data_dir, snapshot_fname)
-        _brotli_subprocess_save(contents, filepath, mode=None, fallback_uncompressed=True)
+        filepath = os.path.join(self.data_dir, snapshot_fname)
+        _brotli_save(contents, filepath, mode=None, fallback_uncompressed=True)
         self._prune_last_fetched_html_snapshots()
 
     def get_fetched_html(self, timestamp):
         import brotli
 
         snapshot_fname = f"{timestamp}.html.br"
-        filepath = os.path.join(self.watch_data_dir, snapshot_fname)
+        filepath = os.path.join(self.data_dir, snapshot_fname)
         if os.path.isfile(filepath):
             with open(filepath, 'rb') as f:
                 return (brotli.decompress(f.read()).decode('utf-8'))
@@ -978,7 +1154,7 @@ class model(watch_base):
 
         for index, timestamp in enumerate(dates):
             snapshot_fname = f"{timestamp}.html.br"
-            filepath = os.path.join(self.watch_data_dir, snapshot_fname)
+            filepath = os.path.join(self.data_dir, snapshot_fname)
 
             # Keep only the first 2
             if index > 1 and os.path.isfile(filepath):
@@ -989,7 +1165,7 @@ class model(watch_base):
     def get_browsersteps_available_screenshots(self):
         "For knowing which screenshots are available to show the user in BrowserSteps UI"
         available = []
-        for f in Path(self.watch_data_dir).glob('step_before-*.jpeg'):
+        for f in Path(self.data_dir).glob('step_before-*.jpeg'):
             step_n=re.search(r'step_before-(\d+)', f.name)
             if step_n:
                 available.append(step_n.group(1))
@@ -998,18 +1174,13 @@ class model(watch_base):
     def compile_error_texts(self, has_proxies=None):
         """Compile error texts for this watch.
         Accepts has_proxies parameter to ensure it works even outside app context"""
-        from flask import url_for
+        from flask import url_for, has_request_context
         from markupsafe import Markup
 
         output = []  # Initialize as list since we're using append
         last_error = self.get('last_error','')
 
-        try:
-            url_for('settings.settings_page')
-        except Exception as e:
-            has_app_context = False
-        else:
-            has_app_context = True
+        has_app_context = has_request_context()
 
         # has app+request context, we can use url_for()
         if has_app_context:

changedetectionio/model/__init__.py

@@ -2,12 +2,175 @@ import os
 import uuid
 
 from changedetectionio import strtobool
+from .persistence import EntityPersistenceMixin, _determine_entity_type
+
+__all__ = ['EntityPersistenceMixin', 'watch_base']
+
+from ..browser_steps.browser_steps import browser_steps_get_valid_steps
+
 USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH = 'System default'
 CONDITIONS_MATCH_LOGIC_DEFAULT = 'ALL'
 
+
 class watch_base(dict):
+    """
+    Base watch domain model (inherits from dict for backward compatibility).
+
+    WARNING: This class inherits from dict, which violates proper encapsulation.
+    Dict inheritance is legacy technical debt that should be refactored to a proper
+    domain model (e.g., Pydantic BaseModel) for better type safety and validation.
+
+    TODO: Migrate to Pydantic BaseModel for:
+          - Type safety and IDE autocomplete
+          - Automatic validation
+          - Clear separation between domain model and serialization
+          - Database backend abstraction (file → postgres → mongodb)
+          - Configuration override chain resolution (Watch → Tag → Global)
+          - Immutability options
+          - Better testing
+          - USE https://docs.pydantic.dev/latest/integrations/datamodel_code_generator TO BUILD THE MODEL FROM THE API-SPEC!!!
+
+    CHAIN RESOLUTION ARCHITECTURE:
+        The dream is a 3-level override hierarchy:
+            Watch settings → Tag/Group settings → Global settings
+
+        Current implementation: MANUAL resolution scattered across codebase
+        - Processors manually check watch.get('field')
+        - Loop through tags to find overrides_watch=True
+        - Fall back to datastore['settings']['application']['field']
+
+        Pydantic implementation: AUTOMATIC resolution via @computed_field
+        - Single source of truth for each setting's resolution logic
+        - Type-safe, testable, self-documenting
+        - Example: watch.resolved_fetch_backend (instead of nested dict navigation)
+
+        See: Watch.py model docstring for detailed Pydantic architecture plan
+        See: Tag.py model docstring for tag override explanation
+        See: processors/restock_diff/processor.py:184-192 for current manual example
+
+    Core Fields:
+        uuid (str): Unique identifier for this watch (auto-generated)
+        url (str): Target URL to monitor for changes
+        title (str|None): Custom display name (overrides page_title if set)
+        page_title (str|None): Title extracted from <title> tag of monitored page
+        tags (List[str]): List of tag UUIDs for categorization
+        tag (str): DEPRECATED - Old single-tag system, use tags instead
+
+    Check Configuration:
+        processor (str): Processor type ('text_json_diff', 'restock_diff', etc.)
+        fetch_backend (str): Fetcher to use ('system', 'html_requests', 'playwright', etc.)
+        method (str): HTTP method ('GET', 'POST', etc.)
+        headers (dict): Custom HTTP headers to send
+        proxy (str|None): Preferred proxy server
+        paused (bool): Whether change detection is paused
+
+    Scheduling:
+        time_between_check (dict): Check interval {'weeks': int, 'days': int, 'hours': int, 'minutes': int, 'seconds': int}
+        time_between_check_use_default (bool): Use global default interval if True
+        time_schedule_limit (dict): Weekly schedule limiting when checks can run
+            Structure: {
+                'enabled': bool,
+                'monday/tuesday/.../sunday': {
+                    'enabled': bool,
+                    'start_time': str ('HH:MM'),
+                    'duration': {'hours': str, 'minutes': str}
+                }
+            }
+
+    Content Filtering:
+        include_filters (List[str]): CSS/XPath selectors to extract content
+        subtractive_selectors (List[str]): Selectors to remove from content
+        ignore_text (List[str]): Text patterns to ignore in change detection
+        trigger_text (List[str]): Text/regex that must be present to trigger change
+        text_should_not_be_present (List[str]): Text that should NOT be present
+        extract_text (List[str]): Regex patterns to extract specific text after filtering
+
+    Text Processing:
+        trim_text_whitespace (bool): Strip leading/trailing whitespace
+        sort_text_alphabetically (bool): Sort lines alphabetically before comparison
+        remove_duplicate_lines (bool): Remove duplicate lines
+        check_unique_lines (bool): Compare against all history for unique lines
+        strip_ignored_lines (bool|None): Remove lines matching ignore patterns
+
+    Change Detection Filters:
+        filter_text_added (bool): Include added text in change detection
+        filter_text_removed (bool): Include removed text in change detection
+        filter_text_replaced (bool): Include replaced text in change detection
+
+    Browser Automation:
+        browser_steps (List[dict]): Browser automation steps for JS-heavy sites
+        browser_steps_last_error_step (int|None): Last step that caused error
+        webdriver_delay (int|None): Seconds to wait after page load
+        webdriver_js_execute_code (str|None): JavaScript to execute before extraction
+
+    Restock Detection:
+        in_stock_only (bool): Only trigger on in-stock transitions
+        follow_price_changes (bool): Monitor price changes
+        has_ldjson_price_data (bool|None): Whether page has LD-JSON price data
+        track_ldjson_price_data (str|None): Track LD-JSON price data ('ACCEPT', 'REJECT', None)
+        price_change_threshold_percent (float|None): Minimum price change % to trigger
+
+    Notifications:
+        notification_urls (List[str]): Apprise URLs for notifications
+        notification_title (str|None): Custom notification title template
+        notification_body (str|None): Custom notification body template
+        notification_format (str): Notification format (e.g., 'System default', 'Text', 'HTML')
+        notification_muted (bool): Disable notifications for this watch
+        notification_screenshot (bool): Include screenshot in notifications
+        notification_alert_count (int): Number of notifications sent
+        last_notification_error (str|None): Last notification error message
+        body (str|None): DEPRECATED? Legacy notification body field
+        filter_failure_notification_send (bool): Send notification on filter failures
+
+    History & State:
+        date_created (int|None): Unix timestamp of watch creation
+        last_checked (int): Unix timestamp of last check
+        last_viewed (int): History snapshot key of last user view
+        last_error (str|bool): Last error message or False if no error
+        check_count (int): Total number of checks performed
+        fetch_time (float): Duration of last fetch in seconds
+        consecutive_filter_failures (int): Counter for consecutive filter match failures
+        previous_md5 (str|bool): MD5 hash of previous content
+        history_snapshot_max_length (int|None): Max history snapshots to keep (None = use global)
+
+    Conditions:
+        conditions (dict): Custom conditions for change detection logic
+        conditions_match_logic (str): Logic operator ('ALL', 'ANY') for conditions
+
+    Metadata:
+        content-type (str|None): Content-Type from last fetch
+        remote_server_reply (str|None): Server header from last response
+        ignore_status_codes (List[int]|None): HTTP status codes to ignore
+        use_page_title_in_list (bool|None): Display page title in watch list (None = use system default)
+
+    Instance Attributes (not serialized):
+        __datastore: Reference to parent DataStore (set externally after creation)
+        data_dir: Filesystem path for this watch's data directory
+
+    Notes:
+        - Many fields default to None to distinguish "not set" from "set to default"
+        - When field is None, system-level defaults are used
+        - Processor-specific configs (e.g., processor_config_*) are NOT stored in watch.json
+          They are stored in separate {processor_name}.json files
+        - This class is used for both Watch and Tag objects (tags reuse the structure)
+    """
 
     def __init__(self, *arg, **kw):
+        # Store datastore reference (common to Watch and Tag)
+        # Use single underscore to avoid name mangling issues in subclasses
+        self._datastore = kw.get('__datastore')
+        if kw.get('__datastore'):
+            del kw['__datastore']
+
+        # Store datastore_path (common to Watch and Tag)
+        self._datastore_path = kw.get('datastore_path')
+        if kw.get('datastore_path'):
+            del kw['datastore_path']
+
+        # IMPORTANT: Don't initialize __watch_was_edited yet!
+        # We'll initialize it AFTER the initial update() call below
+        # This prevents marking the watch as edited during initialization
+
         self.update({
             # Custom notification content
             # Re #110, so then if this is set to None, we know to use the default value instead
@@ -16,7 +179,7 @@ class watch_base(dict):
             'body': None,
             'browser_steps': [],
             'browser_steps_last_error_step': None,
-            'conditions' : {},
+            'conditions' : [],
             'conditions_match_logic': CONDITIONS_MATCH_LOGIC_DEFAULT,
             'check_count': 0,
             'check_unique_lines': False,  # On change-detected, compare against all history if its something new
@@ -32,6 +195,7 @@ class watch_base(dict):
             'filter_text_replaced': True,
             'follow_price_changes': True,
             'has_ldjson_price_data': None,
+            'history_snapshot_max_length': None,
             'headers': {},  # Extra headers to send
             'ignore_text': [],  # List of text to ignore when calculating the comparison checksum
             'ignore_status_codes': None,
@@ -52,7 +216,6 @@ class watch_base(dict):
             'page_title': None, # <title> from the page
             'paused': False,
             'previous_md5': False,
-            'previous_md5_before_filters': False,  # Used for skipping changedetection entirely
             'processor': 'text_json_diff',  # could be restock_diff or others from .processors
             'price_change_threshold_percent': None,
             'proxy': None,  # Preferred proxy connection
@@ -138,5 +301,372 @@ class watch_base(dict):
 
         super(watch_base, self).__init__(*arg, **kw)
 
+        # Check if we're being initialized from an existing watch object
+        # that has was_edited=True, so we can preserve the flag
+        preserve_edited_flag = False
         if self.get('default'):
-            del self['default']
+            # When creating a new watch object from an existing one (e.g., changing processor),
+            # preserve the was_edited flag if it was True
+            default_watch = self.get('default')
+            if hasattr(default_watch, 'was_edited') and default_watch.was_edited:
+                preserve_edited_flag = True
+            del self['default']
+
+        # NOW initialize the edited flag after all initial setup is complete
+        # This ensures initialization doesn't trigger the edited flag
+        # But preserve it if the source watch had it set to True
+        self.__watch_was_edited = preserve_edited_flag
+
+    def _mark_field_as_edited(self, key):
+        """
+        Helper to mark a field as edited if it's writable.
+
+        Internal method used by __setitem__, update(), pop(), etc.
+        """
+        # Don't track edits during initial load or if already edited
+        if not hasattr(self, '_watch_base__watch_was_edited'):
+            return
+        if self.__watch_was_edited:
+            return  # Already marked as edited
+
+        # Import from shared schema utilities (no circular dependency)
+        from .schema_utils import get_readonly_watch_fields
+        readonly_fields = get_readonly_watch_fields()
+
+        # Additional system-managed fields not in OpenAPI spec (yet)
+        # These are set by processors/workers and should not trigger edited flag
+        additional_system_fields = {
+            'last_check_status',  # Set by processors
+            'restock',  # Set by restock processor
+            'last_viewed',  # Set by mark_all_viewed endpoint
+        }
+
+        # Only mark as edited if this is a user-writable field
+        if key not in readonly_fields and key not in additional_system_fields:
+            self.__watch_was_edited = True
+
+    def __setitem__(self, key, value):
+        """
+        Override dict.__setitem__ to track when writable watch fields are modified.
+
+        This enables skipping reprocessing when:
+        1. HTML content is unchanged (checksumFromPreviousCheckWasTheSame)
+        2. AND watch configuration was not edited
+
+        Only sets the edited flag when field is NOT in readonly_fields (from OpenAPI spec).
+        """
+        # Set the value first (always)
+        super().__setitem__(key, value)
+        # Mark as edited if writable field
+        self._mark_field_as_edited(key)
+
+    def __delitem__(self, key):
+        """Override dict.__delitem__ to track deletions of writable fields."""
+        super().__delitem__(key)
+        self._mark_field_as_edited(key)
+
+    def update(self, *args, **kwargs):
+
+        if args and args[0].get('browser_steps'):
+            args[0]['browser_steps'] = browser_steps_get_valid_steps(args[0].get('browser_steps'))
+
+        """Override dict.update() to track modifications to writable fields."""
+        # Call parent update first
+        super().update(*args, **kwargs)
+
+        # Mark as edited for any writable fields that were updated
+        # Handle both update(dict) and update(key=value) forms
+        if args:
+            for key in args[0].keys():
+                self._mark_field_as_edited(key)
+        for key in kwargs.keys():
+            self._mark_field_as_edited(key)
+
+
+    def pop(self, key, *args):
+        """Override dict.pop() to track removal of writable fields."""
+        result = super().pop(key, *args)
+        self._mark_field_as_edited(key)
+        return result
+
+    def setdefault(self, key, default=None):
+        """Override dict.setdefault() to track modifications to writable fields."""
+        # Only marks as edited if key didn't exist (i.e., a new value was set)
+        existed = key in self
+        result = super().setdefault(key, default)
+        if not existed:
+            self._mark_field_as_edited(key)
+        return result
+
+    @property
+    def was_edited(self):
+        """
+        Check if watch configuration was edited since last processing.
+
+        Returns:
+            bool: True if writable fields were modified, False otherwise
+        """
+        return getattr(self, '_watch_base__watch_was_edited', False)
+
+    def reset_watch_edited_flag(self):
+        """
+        Reset the watch edited flag after successful processing.
+
+        Call this after processing completes to allow future content-only change detection.
+        """
+        self.__watch_was_edited = False
+
+    @classmethod
+    def get_property_names(cls):
+        """
+        Get all @property attribute names from this model class using introspection.
+
+        This discovers computed/derived properties that are not stored in the datastore.
+        These properties should be filtered out during PUT/POST requests.
+
+        Returns:
+            frozenset: Immutable set of @property attribute names from the model class
+        """
+        import functools
+
+        # Create a cached version if it doesn't exist
+        if not hasattr(cls, '_cached_get_property_names'):
+            @functools.cache
+            def _get_props():
+                properties = set()
+                # Use introspection to find all @property attributes
+                for name in dir(cls):
+                    # Skip private/magic attributes
+                    if name.startswith('_'):
+                        continue
+                    try:
+                        attr = getattr(cls, name)
+                        # Check if it's a property descriptor
+                        if isinstance(attr, property):
+                            properties.add(name)
+                    except (AttributeError, TypeError):
+                        continue
+                return frozenset(properties)
+
+            cls._cached_get_property_names = _get_props
+
+        return cls._cached_get_property_names()
+
+    def __deepcopy__(self, memo):
+        """
+        Custom deepcopy for all watch_base subclasses (Watch, Tag, etc.).
+
+        CRITICAL FIX: Prevents copying large reference objects like __datastore
+        which would cause exponential memory growth when Watch objects are deepcopied.
+
+        This is called by:
+        - api/Watch.py:76 (API endpoint)
+        - api/Tags.py:28 (Tags API)
+        - processors/base.py:26 (EVERY processor run)
+        - store/__init__.py:544 (clone watch)
+        - And other locations
+        """
+        from copy import deepcopy
+
+        # Create new instance without calling __init__
+        cls = self.__class__
+        new_obj = cls.__new__(cls)
+        memo[id(self)] = new_obj
+
+        # Copy the dict data (all the settings)
+        for key, value in self.items():
+            new_obj[key] = deepcopy(value, memo)
+
+        # Copy instance attributes dynamically
+        # This handles Watch-specific attrs (like __datastore) and any future subclass attrs
+        for attr_name in dir(self):
+            # Skip methods, special attrs, and dict keys
+            if attr_name.startswith('_') and not attr_name.startswith('__'):
+                # This catches _model__datastore, _model__history_n, etc.
+                try:
+                    attr_value = getattr(self, attr_name)
+
+                    # Special handling: Share references to large objects instead of copying
+                    # Examples: _datastore, __datastore, __app_reference, __global_settings, etc.
+                    if (attr_name == '_datastore' or
+                        attr_name.endswith('__datastore') or
+                        attr_name.endswith('__app')):
+                        # Share the reference (don't copy!) to prevent memory leaks
+                        setattr(new_obj, attr_name, attr_value)
+                    # Skip cache attributes - let them regenerate on demand
+                    elif 'cache' in attr_name.lower():
+                        pass  # Don't copy caches
+                    # Copy regular instance attributes
+                    elif not callable(attr_value):
+                        setattr(new_obj, attr_name, attr_value)
+                except AttributeError:
+                    pass  # Attribute doesn't exist in this instance
+
+        return new_obj
+
+    def __getstate__(self):
+        """
+        Custom pickle serialization for all watch_base subclasses.
+
+        Excludes large reference objects (like __datastore) from serialization.
+        """
+        # Get the dict data
+        state = dict(self)
+
+        # Collect instance attributes (excluding methods and large references)
+        instance_attrs = {}
+        for attr_name in dir(self):
+            if attr_name.startswith('_') and not attr_name.startswith('__'):
+                try:
+                    attr_value = getattr(self, attr_name)
+                    # Exclude large reference objects and caches from serialization
+                    if not (attr_name == '_datastore' or
+                           attr_name.endswith('__datastore') or
+                           attr_name.endswith('__app') or
+                           'cache' in attr_name.lower() or
+                           callable(attr_value)):
+                        instance_attrs[attr_name] = attr_value
+                except AttributeError:
+                    pass
+
+        if instance_attrs:
+            state['__instance_metadata__'] = instance_attrs
+
+        return state
+
+    def __setstate__(self, state):
+        """
+        Custom pickle deserialization for all watch_base subclasses.
+
+        WARNING: Large reference objects (like __datastore) are NOT restored!
+        Caller must restore these references after unpickling if needed.
+        """
+        # Extract metadata
+        metadata = state.pop('__instance_metadata__', {})
+
+        # Restore dict data
+        self.update(state)
+
+        # Restore instance attributes
+        for attr_name, attr_value in metadata.items():
+            setattr(self, attr_name, attr_value)
+
+    @property
+    def data_dir(self):
+        """
+        The base directory for this watch/tag data (property, computed from UUID).
+
+        Common property for both Watch and Tag objects.
+        Returns path like: /datastore/{uuid}/
+        """
+        return os.path.join(self._datastore_path, self['uuid']) if self._datastore_path else None
+
+    def ensure_data_dir_exists(self):
+        """
+        Create the data directory if it doesn't exist.
+
+        Common method for both Watch and Tag objects.
+        """
+        from loguru import logger
+        if not os.path.isdir(self.data_dir):
+            logger.debug(f"> Creating data dir {self.data_dir}")
+            os.mkdir(self.data_dir)
+
+    def get_global_setting(self, *path):
+        """
+        Get a setting from the global datastore configuration.
+
+        Args:
+            *path: Path to the setting (e.g., 'application', 'history_snapshot_max_length')
+
+        Returns:
+            The setting value, or None if not found
+
+        Example:
+            maxlen = self.get_global_setting('application', 'history_snapshot_max_length')
+        """
+        if not self._datastore:
+            return None
+
+        try:
+            value = self._datastore['settings']
+            for key in path:
+                value = value[key]
+            return value
+        except (KeyError, TypeError):
+            return None
+
+    def _get_commit_data(self):
+        """
+        Prepare data for commit (can be overridden by subclasses).
+
+        Returns:
+            dict: Data to serialize (filtered as needed by subclass)
+        """
+        import copy
+
+        # Acquire datastore lock to prevent concurrent modifications during copy
+        lock = self._datastore.lock if self._datastore and hasattr(self._datastore, 'lock') else None
+
+        if lock:
+            with lock:
+                snapshot = dict(self)
+        else:
+            snapshot = dict(self)
+
+        # Deep copy snapshot (slower, but done outside lock to minimize contention)
+        # Subclasses can override to filter keys (e.g., Watch excludes processor_config_*)
+        return {k: copy.deepcopy(v) for k, v in snapshot.items()}
+
+    def _save_to_disk(self, data_dict, uuid):
+        """
+        Save data to disk (must be implemented by subclasses).
+
+        Args:
+            data_dict: Dictionary to save
+            uuid: UUID for logging
+
+        Raises:
+            NotImplementedError: If subclass doesn't implement
+        """
+        raise NotImplementedError("Subclass must implement _save_to_disk()")
+
+    def commit(self):
+        """
+        Save this watch/tag immediately to disk using atomic write.
+
+        Common commit logic for Watch and Tag objects.
+        Subclasses override _get_commit_data() and _save_to_disk() for specifics.
+
+        Fire-and-forget: Logs errors but does not raise exceptions.
+        Data remains in memory even if save fails, so next commit will retry.
+        """
+        from loguru import logger
+
+        if not self.data_dir:
+            entity_type = self.__class__.__name__
+            logger.error(f"Cannot commit {entity_type} {self.get('uuid')} without datastore_path")
+            return
+
+        uuid = self.get('uuid')
+        if not uuid:
+            entity_type = self.__class__.__name__
+            logger.error(f"Cannot commit {entity_type} without UUID")
+            return
+
+        # Get data from subclass (may filter keys)
+        try:
+            data_dict = self._get_commit_data()
+        except Exception as e:
+            logger.error(f"Failed to prepare commit data for {uuid}: {e}")
+            return
+
+        # Save to disk via subclass implementation
+        try:
+            # Determine entity type from module name (Watch.py -> watch, Tag.py -> tag)
+            entity_type = _determine_entity_type(self.__class__)
+            filename = f"{entity_type}.json"
+            self._save_to_disk(data_dict, uuid)
+            logger.debug(f"Committed {entity_type} {uuid} to {uuid}/{filename}")
+        except Exception as e:
+            logger.error(f"Failed to commit {uuid}: {e}")

changedetectionio/model/persistence.py

@@ -0,0 +1,84 @@
+"""
+Entity persistence mixin for Watch and Tag models.
+
+Provides file-based persistence using atomic writes.
+"""
+
+import functools
+import inspect
+
+
+@functools.lru_cache(maxsize=None)
+def _determine_entity_type(cls):
+    """
+    Determine entity type from class hierarchy (cached at class level).
+
+    Args:
+        cls: The class to inspect
+
+    Returns:
+        str: Entity type ('watch', 'tag', etc.)
+
+    Raises:
+        ValueError: If entity type cannot be determined
+    """
+    for base_class in inspect.getmro(cls):
+        module_name = base_class.__module__
+        if module_name.startswith('changedetectionio.model.'):
+            # Get last part after dot: "changedetectionio.model.Watch" -> "watch"
+            return module_name.split('.')[-1].lower()
+
+    raise ValueError(
+        f"Cannot determine entity type for {cls.__module__}.{cls.__name__}. "
+        f"Entity must inherit from a class in changedetectionio.model (Watch or Tag)."
+    )
+
+
+class EntityPersistenceMixin:
+    """
+    Mixin providing file persistence for watch_base subclasses (Watch, Tag, etc.).
+
+    This mixin provides the _save_to_disk() method required by watch_base.commit().
+    It automatically determines the correct filename and size limits based on class hierarchy.
+
+    Usage:
+        class model(EntityPersistenceMixin, watch_base):  # in Watch.py
+            pass
+
+        class model(EntityPersistenceMixin, watch_base):  # in Tag.py
+            pass
+    """
+
+    def _save_to_disk(self, data_dict, uuid):
+        """
+        Save entity to disk using atomic write.
+
+        Implements the abstract method required by watch_base.commit().
+        Automatically determines filename and size limits from class hierarchy.
+
+        Args:
+            data_dict: Dictionary to save
+            uuid: UUID for logging
+
+        Raises:
+            ValueError: If entity type cannot be determined from class hierarchy
+        """
+        # Import here to avoid circular dependency
+        from changedetectionio.store.file_saving_datastore import save_entity_atomic
+
+        # Determine entity type (cached at class level, not instance level)
+        entity_type = _determine_entity_type(self.__class__)
+
+        # Set filename and size limits based on entity type
+        filename = f'{entity_type}.json'
+        max_size_mb = 10 if entity_type == 'watch' else 1
+
+        # Save using generic function
+        save_entity_atomic(
+            self.data_dir,
+            uuid,
+            data_dict,
+            filename=filename,
+            entity_type=entity_type,
+            max_size_mb=max_size_mb
+        )

changedetectionio/model/schema_utils.py

@@ -0,0 +1,92 @@
+"""
+Schema utilities for Watch and Tag models.
+
+Provides functions to extract readonly fields and properties from OpenAPI spec.
+Shared by both the model layer and API layer to avoid circular dependencies.
+"""
+
+import functools
+
+
+@functools.cache
+def get_openapi_schema_dict():
+    """
+    Get the raw OpenAPI spec dictionary for schema access.
+
+    Returns the YAML dict directly (not the OpenAPI object).
+    """
+    import os
+    import yaml
+
+    spec_path = os.path.join(os.path.dirname(__file__), '../../docs/api-spec.yaml')
+    if not os.path.exists(spec_path):
+        spec_path = os.path.join(os.path.dirname(__file__), '../docs/api-spec.yaml')
+
+    with open(spec_path, 'r', encoding='utf-8') as f:
+        return yaml.safe_load(f)
+
+
+@functools.cache
+def _resolve_readonly_fields(schema_name):
+    """
+    Generic helper to resolve readOnly fields, including allOf inheritance.
+
+    Args:
+        schema_name: Name of the schema (e.g., 'Watch', 'Tag')
+
+    Returns:
+        frozenset: All readOnly field names including inherited ones
+    """
+    spec_dict = get_openapi_schema_dict()
+    schema = spec_dict['components']['schemas'].get(schema_name, {})
+
+    readonly_fields = set()
+
+    # Handle allOf (schema inheritance)
+    if 'allOf' in schema:
+        for item in schema['allOf']:
+            # Resolve $ref to parent schema
+            if '$ref' in item:
+                ref_path = item['$ref'].split('/')[-1]
+                ref_schema = spec_dict['components']['schemas'].get(ref_path, {})
+                if 'properties' in ref_schema:
+                    for field_name, field_def in ref_schema['properties'].items():
+                        if field_def.get('readOnly') is True:
+                            readonly_fields.add(field_name)
+            # Check schema-specific properties
+            if 'properties' in item:
+                for field_name, field_def in item['properties'].items():
+                    if field_def.get('readOnly') is True:
+                        readonly_fields.add(field_name)
+    else:
+        # Direct properties (no inheritance)
+        if 'properties' in schema:
+            for field_name, field_def in schema['properties'].items():
+                if field_def.get('readOnly') is True:
+                    readonly_fields.add(field_name)
+
+    return frozenset(readonly_fields)
+
+
+@functools.cache
+def get_readonly_watch_fields():
+    """
+    Extract readOnly field names from Watch schema in OpenAPI spec.
+
+    Returns readOnly fields from WatchBase (uuid, date_created) + Watch-specific readOnly fields.
+
+    Used by:
+    - model/watch_base.py: Track when writable fields are edited
+    - api/Watch.py: Filter readonly fields from PUT requests
+    """
+    return _resolve_readonly_fields('Watch')
+
+
+@functools.cache
+def get_readonly_tag_fields():
+    """
+    Extract readOnly field names from Tag schema in OpenAPI spec.
+
+    Returns readOnly fields from WatchBase (uuid, date_created) + Tag-specific readOnly fields.
+    """
+    return _resolve_readonly_fields('Tag')

changedetectionio/notification_service.py

@@ -54,34 +54,128 @@ def _check_cascading_vars(datastore, var_name, watch):
     return None
 
 
+class FormattableTimestamp(str):
+    """
+    A str subclass representing a formatted datetime. As a plain string it renders
+    with the default format, but can also be called with a custom format argument
+    in Jinja2 templates:
+
+        {{ change_datetime }}                        → '2024-01-15 10:30:00 UTC'
+        {{ change_datetime(format='%Y') }}           → '2024'
+        {{ change_datetime(format='%A') }}           → 'Monday'
+        {{ change_datetime(format='%Y-%m-%d') }}     → '2024-01-15'
+
+    Being a str subclass means it is natively JSON serializable.
+    """
+    _DEFAULT_FORMAT = '%Y-%m-%d %H:%M:%S %Z'
+
+    def __new__(cls, timestamp):
+        dt = datetime.datetime.fromtimestamp(int(timestamp), tz=pytz.UTC)
+        local_tz = datetime.datetime.now().astimezone().tzinfo
+        dt_local = dt.astimezone(local_tz)
+        try:
+            formatted = dt_local.strftime(cls._DEFAULT_FORMAT)
+        except Exception:
+            formatted = dt_local.isoformat()
+        instance = super().__new__(cls, formatted)
+        instance._dt = dt_local
+        return instance
+
+    def __call__(self, format=_DEFAULT_FORMAT):
+        try:
+            return self._dt.strftime(format)
+        except Exception:
+            return self._dt.isoformat()
+
+
+class FormattableDiff(str):
+    """
+    A str subclass representing a rendered diff. As a plain string it renders
+    with the default options for that variant, but can be called with custom
+    arguments in Jinja2 templates:
+
+        {{ diff }}                                    → default diff output
+        {{ diff(lines=5) }}                           → truncate to 5 lines
+        {{ diff(added_only=true) }}                   → only show added lines
+        {{ diff(removed_only=true) }}                 → only show removed lines
+        {{ diff(context=3) }}                         → 3 lines of context around changes
+        {{ diff(word_diff=false) }}                   → line-level diff instead of word-level
+        {{ diff(lines=10, added_only=true) }}         → combine args
+        {{ diff_added(lines=5) }}                     → works on any diff_* variant too
+
+    Being a str subclass means it is natively JSON serializable.
+    """
+    def __new__(cls, prev_snapshot, current_snapshot, **base_kwargs):
+        if prev_snapshot or current_snapshot:
+            from changedetectionio import diff as diff_module
+            rendered = diff_module.render_diff(prev_snapshot, current_snapshot, **base_kwargs)
+        else:
+            rendered = ''
+        instance = super().__new__(cls, rendered)
+        instance._prev = prev_snapshot
+        instance._current = current_snapshot
+        instance._base_kwargs = base_kwargs
+        return instance
+
+    def __call__(self, lines=None, added_only=False, removed_only=False, context=0,
+                 word_diff=None, case_insensitive=False, ignore_junk=False):
+        from changedetectionio import diff as diff_module
+        kwargs = dict(self._base_kwargs)
+
+        if added_only:
+            kwargs['include_removed'] = False
+        if removed_only:
+            kwargs['include_added'] = False
+        if context:
+            kwargs['context_lines'] = int(context)
+        if word_diff is not None:
+            kwargs['word_diff'] = bool(word_diff)
+        if case_insensitive:
+            kwargs['case_insensitive'] = True
+        if ignore_junk:
+            kwargs['ignore_junk'] = True
+
+        result = diff_module.render_diff(self._prev or '', self._current or '', **kwargs)
+
+        if lines is not None:
+            result = '\n'.join(result.splitlines()[:int(lines)])
+
+        return result
+
+
+
 # What is passed around as notification context, also used as the complete list of valid {{ tokens }}
 class NotificationContextData(dict):
     def __init__(self, initial_data=None, **kwargs):
+        # ValidateJinja2Template() validates against the keynames of this dict to check for valid tokens in the body (user submission)
         super().__init__({
             'base_url': None,
+            'change_datetime': FormattableTimestamp(time.time()),
             'current_snapshot': None,
-            'diff': None,
-            'diff_clean': None,
-            'diff_added': None,
-            'diff_added_clean': None,
-            'diff_full': None,
-            'diff_full_clean': None,
-            'diff_patch': None,
-            'diff_removed': None,
-            'diff_removed_clean': None,
+            'diff': FormattableDiff('', ''),
+            'diff_clean': FormattableDiff('', '', include_change_type_prefix=False),
+            'diff_added': FormattableDiff('', '', include_removed=False),
+            'diff_added_clean': FormattableDiff('', '', include_removed=False, include_change_type_prefix=False),
+            'diff_full': FormattableDiff('', '', include_equal=True),
+            'diff_full_clean': FormattableDiff('', '', include_equal=True, include_change_type_prefix=False),
+            'diff_patch': FormattableDiff('', '', patch_format=True),
+            'diff_removed': FormattableDiff('', '', include_added=False),
+            'diff_removed_clean': FormattableDiff('', '', include_added=False, include_change_type_prefix=False),
             'diff_url': None,
             'markup_text_links_to_html_links': False, # If automatic conversion of plaintext to HTML should happen
             'notification_timestamp': time.time(),
+            'prev_snapshot': None,
             'preview_url': None,
             'screenshot': None,
-            'triggered_text': None,
             'timestamp_from': None,
             'timestamp_to': None,
+            'triggered_text': None,
             'uuid': 'XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX',  # Converted to 'watch_uuid' in create_notification_parameters
             'watch_mime_type': None,
             'watch_tag': None,
             'watch_title': None,
             'watch_url': 'https://WATCH-PLACE-HOLDER/',
+            'watch_uuid': 'XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX',  # Converted to 'watch_uuid' in create_notification_parameters
         })
 
         # Apply any initial data passed in
@@ -103,7 +197,7 @@ class NotificationContextData(dict):
         So we can test the output in the notification body
         """
         for key in self.keys():
-            if key in ['uuid', 'time', 'watch_uuid']:
+            if key in ['uuid', 'time', 'watch_uuid', 'change_datetime'] or key.startswith('diff'):
                 continue
             rand_str = 'RANDOM-PLACEHOLDER-'+''.join(random.choices(string.ascii_letters + string.digits, k=12))
             self[key] = rand_str
@@ -115,24 +209,6 @@ class NotificationContextData(dict):
 
         super().__setitem__(key, value)
 
-def timestamp_to_localtime(timestamp):
-    # Format the date using locale-aware formatting with timezone
-    dt = datetime.datetime.fromtimestamp(int(timestamp))
-    dt = dt.replace(tzinfo=pytz.UTC)
-
-    # Get local timezone-aware datetime
-    local_tz = datetime.datetime.now().astimezone().tzinfo
-    local_dt = dt.astimezone(local_tz)
-
-    # Format date with timezone - using strftime for locale awareness
-    try:
-        formatted_date = local_dt.strftime('%Y-%m-%d %H:%M:%S %Z')
-    except:
-        # Fallback if locale issues
-        formatted_date = local_dt.isoformat()
-
-    return formatted_date
-
 def add_rendered_diff_to_notification_vars(notification_scan_text:str, prev_snapshot:str, current_snapshot:str, word_diff:bool):
     """
     Efficiently renders only the diff placeholders that are actually used in the notification text.
@@ -150,13 +226,12 @@ def add_rendered_diff_to_notification_vars(notification_scan_text:str, prev_snap
     Returns:
         dict: Only the diff placeholders that were found in notification_scan_text, with rendered content
     """
-    from changedetectionio import diff
     import re
-    from functools import lru_cache
 
     now = time.time()
 
-    # Define specifications for each diff variant
+    # Define base kwargs for each diff variant — these become the stored defaults
+    # on the FormattableDiff object, so {{ diff(lines=5) }} overrides on top of them
     diff_specs = {
         'diff': {'word_diff': word_diff},
         'diff_clean': {'word_diff': word_diff, 'include_change_type_prefix': False},
@@ -169,22 +244,15 @@ def add_rendered_diff_to_notification_vars(notification_scan_text:str, prev_snap
         'diff_removed_clean': {'word_diff': word_diff, 'include_added': False, 'include_change_type_prefix': False},
     }
 
-    # Memoize render_diff to avoid duplicate renders with same kwargs
-    @lru_cache(maxsize=4)
-    def cached_render(kwargs_tuple):
-        return diff.render_diff(prev_snapshot, current_snapshot, **dict(kwargs_tuple))
-
     ret = {}
     rendered_count = 0
-    # Only check and render diff keys that exist in NotificationContextData
+    # Only create FormattableDiff objects for diff keys actually used in the notification text
     for key in NotificationContextData().keys():
         if key.startswith('diff') and key in diff_specs:
             # Check if this placeholder is actually used in the notification text
             pattern = rf"(?<![A-Za-z0-9_]){re.escape(key)}(?![A-Za-z0-9_])"
             if re.search(pattern, notification_scan_text, re.IGNORECASE):
-                kwargs = diff_specs[key]
-                # Convert dict to sorted tuple for cache key (handles duplicate kwarg combinations)
-                ret[key] = cached_render(tuple(sorted(kwargs.items())))
+                ret[key] = FormattableDiff(prev_snapshot, current_snapshot, **diff_specs[key])
                 rendered_count += 1
 
     if rendered_count:
@@ -198,7 +266,7 @@ def set_basic_notification_vars(current_snapshot, prev_snapshot, watch, triggere
         'current_snapshot': current_snapshot,
         'prev_snapshot': prev_snapshot,
         'screenshot': watch.get_screenshot() if watch and watch.get('notification_screenshot') else None,
-        'change_datetime': timestamp_to_localtime(timestamp_changed) if timestamp_changed else None,
+        'change_datetime': FormattableTimestamp(timestamp_changed) if timestamp_changed else None,
         'triggered_text': triggered_text,
         'uuid': watch.get('uuid') if watch else None,
         'watch_url': watch.get('url') if watch else None,

changedetectionio/pluggy_interface.py

@@ -105,6 +105,75 @@ class ChangeDetectionSpec:
         """
         pass
 
+    @hookspec
+    def register_processor(self):
+        """Register an external processor plugin.
+
+        External packages can implement this hook to register custom processors
+        that will be discovered alongside built-in processors.
+
+        Returns:
+            dict or None: Dictionary with processor information:
+                {
+                    'processor_name': str,      # Machine name (e.g., 'osint_recon')
+                    'processor_module': module, # Module containing processor.py
+                    'processor_class': class,   # The perform_site_check class
+                    'metadata': {               # Optional metadata
+                        'name': str,            # Display name
+                        'description': str,     # Description
+                        'processor_weight': int,# Sort weight (lower = higher priority)
+                        'list_badge_text': str, # Badge text for UI
+                    }
+                }
+                Return None if this plugin doesn't provide a processor
+        """
+        pass
+
+    @hookspec
+    def update_handler_alter(update_handler, watch, datastore):
+        """Modify or wrap the update_handler before it processes a watch.
+
+        This hook is called after the update_handler (perform_site_check instance) is created
+        but before it calls call_browser() and run_changedetection(). Plugins can use this to:
+        - Wrap the handler to add logging/metrics
+        - Modify handler configuration
+        - Add custom preprocessing logic
+
+        Args:
+            update_handler: The perform_site_check instance that will process the watch
+            watch: The watch dict being processed
+            datastore: The application datastore
+
+        Returns:
+            object or None: Return a modified/wrapped handler, or None to keep the original.
+                           If multiple plugins return handlers, they are chained in registration order.
+        """
+        pass
+
+    @hookspec
+    def update_finalize(update_handler, watch, datastore, processing_exception):
+        """Called after watch processing completes (success or failure).
+
+        This hook is called in the finally block after all processing is complete,
+        allowing plugins to perform cleanup, update metrics, or log final status.
+
+        The plugin can access update_handler.last_logging_insert_id if it was stored
+        during update_handler_alter, and use processing_exception to determine if
+        the processing succeeded or failed.
+
+        Args:
+            update_handler: The perform_site_check instance (may be None if creation failed)
+            watch: The watch dict that was processed (may be None if not loaded)
+            datastore: The application datastore
+            processing_exception: The exception from the main processing block, or None if successful.
+                                 This does NOT include cleanup exceptions - only exceptions from
+                                 the actual watch processing (fetch, diff, etc).
+
+        Returns:
+            None: This hook doesn't return a value
+        """
+        pass
+
 
 # Set up Plugin Manager
 plugin_manager = pluggy.PluginManager(PLUGIN_NAMESPACE)
@@ -475,4 +544,66 @@ def get_plugin_template_paths():
                 template_paths.append(templates_dir)
                 logger.debug(f"Added plugin template path: {templates_dir}")
 
-    return template_paths
+    return template_paths
+
+
+def apply_update_handler_alter(update_handler, watch, datastore):
+    """Apply update_handler_alter hooks from all plugins.
+
+    Allows plugins to wrap or modify the update_handler before it processes a watch.
+    Multiple plugins can chain modifications - each plugin receives the result from
+    the previous plugin.
+
+    Args:
+        update_handler: The perform_site_check instance to potentially modify
+        watch: The watch dict being processed
+        datastore: The application datastore
+
+    Returns:
+        object: The (potentially modified/wrapped) update_handler
+    """
+    # Get all plugins that implement the update_handler_alter hook
+    results = plugin_manager.hook.update_handler_alter(
+        update_handler=update_handler,
+        watch=watch,
+        datastore=datastore
+    )
+
+    # Chain results - each plugin gets the result from the previous one
+    current_handler = update_handler
+    if results:
+        for result in results:
+            if result is not None:
+                logger.debug(f"Plugin modified update_handler for watch {watch.get('uuid')}")
+                current_handler = result
+
+    return current_handler
+
+
+def apply_update_finalize(update_handler, watch, datastore, processing_exception):
+    """Apply update_finalize hooks from all plugins.
+
+    Called in the finally block after watch processing completes, allowing plugins
+    to perform cleanup, update metrics, or log final status.
+
+    Args:
+        update_handler: The perform_site_check instance (may be None)
+        watch: The watch dict that was processed (may be None)
+        datastore: The application datastore
+        processing_exception: The exception from processing, or None if successful
+
+    Returns:
+        None
+    """
+    try:
+        # Call all plugins that implement the update_finalize hook
+        plugin_manager.hook.update_finalize(
+            update_handler=update_handler,
+            watch=watch,
+            datastore=datastore,
+            processing_exception=processing_exception
+        )
+    except Exception as e:
+        # Don't let plugin errors crash the worker
+        logger.error(f"Error in update_finalize hook: {e}")
+        logger.exception(f"update_finalize hook exception details:")

changedetectionio/processors/README.md

@@ -9,6 +9,15 @@ Some suggestions for the future
 
 - `graphical` 
 
+## API schema extension (`api.yaml`)
+
+A processor can extend the Watch/Tag API schema by placing an `api.yaml` alongside its `__init__.py`.
+Define a `components.schemas.processor_config_<name>` entry and it will be merged into `WatchBase` at startup,
+making `processor_config_<name>` a valid field on all watch create/update API calls.
+The fully merged spec is served live at `/api/v1/full-spec`.
+
+See `restock_diff/api.yaml` for a working example.
+
 ## Todo
 
 - Make each processor return a extra list of sub-processed (so you could configure a single processor in different ways)

changedetectionio/processors/__init__.py

@@ -1,6 +1,6 @@
 from functools import lru_cache
 from loguru import logger
-from flask_babel import gettext
+from flask_babel import gettext, get_locale
 import importlib
 import inspect
 import os
@@ -17,9 +17,11 @@ def find_sub_packages(package_name):
     return [name for _, name, is_pkg in pkgutil.iter_modules(package.__path__) if is_pkg]
 
 
+@lru_cache(maxsize=1)
 def find_processors():
     """
     Find all subclasses of DifferenceDetectionProcessor in the specified package.
+    Results are cached to avoid repeated discovery.
 
     :param package_name: The name of the package to scan for processor modules.
     :return: A list of (module, class) tuples.
@@ -46,6 +48,23 @@ def find_processors():
         except (ModuleNotFoundError, ImportError) as e:
             logger.warning(f"Failed to import module {module_name}: {e} (find_processors())")
 
+    # Discover plugin processors via pluggy
+    try:
+        from changedetectionio.pluggy_interface import plugin_manager
+        plugin_results = plugin_manager.hook.register_processor()
+
+        for result in plugin_results:
+            if result and isinstance(result, dict):
+                processor_module = result.get('processor_module')
+                processor_name = result.get('processor_name')
+
+                if processor_module and processor_name:
+                    processors.append((processor_module, processor_name))
+                    plugin_path = getattr(processor_module, '__file__', 'unknown location')
+                    logger.info(f"Registered plugin processor: {processor_name} from {plugin_path}")
+    except Exception as e:
+        logger.warning(f"Error loading plugin processors: {e}")
+
     return processors
 
 
@@ -97,54 +116,138 @@ def find_processor_module(processor_name):
     return None
 
 
-def available_processors():
-    """
-    Get a list of processors by name and description for the UI elements.
-    Can be filtered via ALLOWED_PROCESSORS environment variable (comma-separated list).
-    :return: A list :)
+def get_processor_module(processor_name):
     """
+    Get the actual processor module (with perform_site_check class) by name.
+    Works for both built-in and plugin processors.
 
+    Args:
+        processor_name: Processor machine name (e.g., 'text_json_diff', 'osint_recon')
+
+    Returns:
+        module: The processor module containing perform_site_check, or None if not found
+    """
+    processor_classes = find_processors()
+    processor_tuple = next((tpl for tpl in processor_classes if tpl[1] == processor_name), None)
+
+    if processor_tuple:
+        # Return the actual processor module (first element of tuple)
+        return processor_tuple[0]
+
+    return None
+
+
+def get_processor_submodule(processor_name, submodule_name):
+    """
+    Get an optional submodule from a processor (e.g., 'difference', 'extract', 'preview').
+    Works for both built-in and plugin processors.
+
+    Args:
+        processor_name: Processor machine name (e.g., 'text_json_diff', 'osint_recon')
+        submodule_name: Name of the submodule (e.g., 'difference', 'extract', 'preview')
+
+    Returns:
+        module: The submodule if it exists, or None if not found
+    """
+    processor_classes = find_processors()
+    processor_tuple = next((tpl for tpl in processor_classes if tpl[1] == processor_name), None)
+
+    if not processor_tuple:
+        return None
+
+    processor_module = processor_tuple[0]
+    parent_module = get_parent_module(processor_module)
+
+    if not parent_module:
+        return None
+
+    # Try to import the submodule
+    try:
+        # For built-in processors: changedetectionio.processors.text_json_diff.difference
+        # For plugin processors: changedetectionio_osint.difference
+        parent_module_name = parent_module.__name__
+        submodule_full_name = f"{parent_module_name}.{submodule_name}"
+        return importlib.import_module(submodule_full_name)
+    except (ModuleNotFoundError, ImportError):
+        return None
+
+
+@lru_cache(maxsize=1)
+def get_plugin_processor_metadata():
+    """Get metadata from plugin processors."""
+    metadata = {}
+    try:
+        from changedetectionio.pluggy_interface import plugin_manager
+        plugin_results = plugin_manager.hook.register_processor()
+
+        for result in plugin_results:
+            if result and isinstance(result, dict):
+                processor_name = result.get('processor_name')
+                meta = result.get('metadata', {})
+                if processor_name:
+                    metadata[processor_name] = meta
+    except Exception as e:
+        logger.warning(f"Error getting plugin processor metadata: {e}")
+    return metadata
+
+@lru_cache(maxsize=32)
+def _available_processors_cached(locale_str):
+    """
+    Internal cached function that includes locale in cache key.
+    This ensures translations are cached per-language instead of globally.
+
+    :param locale_str: The locale string (e.g., 'en', 'it', 'zh')
+    :return: A list of tuples (processor_name, translated_description, weight)
+    """
     processor_classes = find_processors()
 
-    # Check if ALLOWED_PROCESSORS env var is set
-    # For now we disable it, need to make a deploy with lots of new code and this will be an overload
-    allowed_processors_env = os.getenv('ALLOWED_PROCESSORS', 'text_json_diff, restock_diff').strip()
-    allowed_processors = None
-    if allowed_processors_env:
+    # Check if DISABLED_PROCESSORS env var is set
+    disabled_processors_env = os.getenv('DISABLED_PROCESSORS', 'image_ssim_diff').strip()
+    disabled_processors = []
+    if disabled_processors_env:
         # Parse comma-separated list and strip whitespace
-        allowed_processors = [p.strip() for p in allowed_processors_env.split(',') if p.strip()]
-        logger.info(f"ALLOWED_PROCESSORS set, filtering to: {allowed_processors}")
+        disabled_processors = [p.strip() for p in disabled_processors_env.split(',') if p.strip()]
+        logger.info(f"DISABLED_PROCESSORS set, disabling: {disabled_processors}")
 
     available = []
+    plugin_metadata = get_plugin_processor_metadata()
+
     for module, sub_package_name in processor_classes:
-        # Filter by allowed processors if set
-        if allowed_processors and sub_package_name not in allowed_processors:
-            logger.debug(f"Skipping processor '{sub_package_name}' (not in ALLOWED_PROCESSORS)")
+        # Skip disabled processors
+        if sub_package_name in disabled_processors:
+            logger.debug(f"Skipping processor '{sub_package_name}' (in DISABLED_PROCESSORS)")
             continue
 
-        # Try to get the 'name' attribute from the processor module first
-        if hasattr(module, 'name'):
-            description = gettext(module.name)
+        # Check if this is a plugin processor
+        if sub_package_name in plugin_metadata:
+            meta = plugin_metadata[sub_package_name]
+            description = gettext(meta.get('name', sub_package_name))
+            # Plugin processors start from weight 10 to separate them from built-in processors
+            weight = 100 + meta.get('processor_weight', 0)
         else:
-            # Fall back to processor_description from parent module's __init__.py
-            parent_module = get_parent_module(module)
-            if parent_module and hasattr(parent_module, 'processor_description'):
-                description = gettext(parent_module.processor_description)
+            # Try to get the 'name' attribute from the processor module first
+            if hasattr(module, 'name'):
+                description = gettext(module.name)
             else:
-                # Final fallback to a readable name
-                description = sub_package_name.replace('_', ' ').title()
+                # Fall back to processor_description from parent module's __init__.py
+                parent_module = get_parent_module(module)
+                if parent_module and hasattr(parent_module, 'processor_description'):
+                    description = gettext(parent_module.processor_description)
+                else:
+                    # Final fallback to a readable name
+                    description = sub_package_name.replace('_', ' ').title()
 
-        # Get weight for sorting (lower weight = higher in list)
-        weight = 0  # Default weight for processors without explicit weight
+            # Get weight for sorting (lower weight = higher in list)
+            weight = 0  # Default weight for processors without explicit weight
 
-        # Check processor module itself first
-        if hasattr(module, 'processor_weight'):
-            weight = module.processor_weight
-        else:
-            # Fall back to parent module (package __init__.py)
-            parent_module = get_parent_module(module)
-            if parent_module and hasattr(parent_module, 'processor_weight'):
-                weight = parent_module.processor_weight
+            # Check processor module itself first
+            if hasattr(module, 'processor_weight'):
+                weight = module.processor_weight
+            else:
+                # Fall back to parent module (package __init__.py)
+                parent_module = get_parent_module(module)
+                if parent_module and hasattr(parent_module, 'processor_weight'):
+                    weight = parent_module.processor_weight
 
         available.append((sub_package_name, description, weight))
 
@@ -154,6 +257,36 @@ def available_processors():
     # Return as tuples without weight (for backwards compatibility)
     return [(name, desc) for name, desc, weight in available]
 
+def available_processors():
+    """
+    Get a list of processors by name and description for the UI elements.
+    Can be filtered via DISABLED_PROCESSORS environment variable (comma-separated list).
+
+    This function delegates to a locale-aware cached version to ensure translations
+    are cached per-language instead of globally.
+
+    :return: A list of tuples (processor_name, translated_description)
+    """
+    # Get current locale and use it as cache key
+    # Convert Babel Locale object to string for use as cache key
+    locale = get_locale()
+    locale_str = str(locale) if locale else 'en'
+    return _available_processors_cached(locale_str)
+
+
+def get_default_processor():
+    """
+    Get the default processor to use when none is specified.
+    Returns the first available processor based on weight (lowest weight = highest priority).
+    This ensures forms auto-select a valid processor even when DISABLED_PROCESSORS filters the list.
+
+    :return: The processor name string (e.g., 'text_json_diff')
+    """
+    available = available_processors()
+    if available:
+        return available[0][0]  # Return the processor name from first tuple
+    return 'text_json_diff'  # Fallback if somehow no processors are available
+
 
 def get_processor_badge_texts():
     """
@@ -279,3 +412,76 @@ def get_processor_badge_css():
 
     return '\n\n'.join(css_rules)
 
+
+def save_processor_config(datastore, watch_uuid, config_data):
+    """
+    Save processor-specific configuration to JSON file.
+
+    This is a shared helper function used by both the UI edit form and API endpoints
+    to consistently handle processor configuration storage.
+
+    Args:
+        datastore: The application datastore instance
+        watch_uuid: UUID of the watch
+        config_data: Dictionary of configuration data to save (with processor_config_* prefix removed)
+
+    Returns:
+        bool: True if saved successfully, False otherwise
+    """
+    if not config_data:
+        return True
+
+    try:
+        from changedetectionio.processors.base import difference_detection_processor
+
+        # Get processor name from watch
+        watch = datastore.data['watching'].get(watch_uuid)
+        if not watch:
+            logger.error(f"Cannot save processor config: watch {watch_uuid} not found")
+            return False
+
+        processor_name = watch.get('processor', 'text_json_diff')
+
+        # Create a processor instance to access config methods
+        processor_instance = difference_detection_processor(datastore, watch_uuid)
+
+        # Use processor name as filename so each processor keeps its own config
+        config_filename = f'{processor_name}.json'
+        processor_instance.update_extra_watch_config(config_filename, config_data)
+
+        logger.debug(f"Saved processor config to {config_filename}: {config_data}")
+        return True
+
+    except Exception as e:
+        logger.error(f"Failed to save processor config: {e}")
+        return False
+
+
+def extract_processor_config_from_form_data(form_data):
+    """
+    Extract processor_config_* fields from form data and return separate dicts.
+
+    This is a shared helper function used by both the UI edit form and API endpoints
+    to consistently handle processor configuration extraction.
+
+    IMPORTANT: This function modifies form_data in-place by removing processor_config_* fields.
+
+    Args:
+        form_data: Dictionary of form data (will be modified in-place)
+
+    Returns:
+        dict: Dictionary of processor config data (with processor_config_* prefix removed)
+    """
+    processor_config_data = {}
+
+    # Use list() to create a copy of keys since we're modifying the dict
+    for field_name in list(form_data.keys()):
+        if field_name.startswith('processor_config_'):
+            config_key = field_name.replace('processor_config_', '')
+            # Save all values (including empty strings) to allow explicit clearing of settings
+            processor_config_data[config_key] = form_data[field_name]
+            # Remove from form_data to prevent it from reaching datastore
+            del form_data[field_name]
+
+    return processor_config_data
+

changedetectionio/processors/base.py

@@ -1,10 +1,15 @@
+import asyncio
 import re
 import hashlib
+
+from changedetectionio.browser_steps.browser_steps import browser_steps_get_valid_steps
 from changedetectionio.content_fetchers.base import Fetcher
 from changedetectionio.strtobool import strtobool
+from changedetectionio.validate_url import is_private_hostname
 from copy import deepcopy
 from abc import abstractmethod
 import os
+from urllib.parse import urlparse
 from loguru import logger
 
 SCREENSHOT_FORMAT_JPEG = 'JPEG'
@@ -19,14 +24,97 @@ class difference_detection_processor():
     xpath_data = None
     preferred_proxy = None
     screenshot_format = SCREENSHOT_FORMAT_JPEG
+    last_raw_content_checksum = None
 
     def __init__(self, datastore, watch_uuid):
         self.datastore = datastore
         self.watch_uuid = watch_uuid
+
+        # Create a stable snapshot of the watch for processing
+        # Why deepcopy?
+        # 1. Prevents "dict changed during iteration" errors if watch is modified during processing
+        # 2. Preserves Watch object with properties (.link, .is_pdf, etc.) - can't use dict()
+        # 3. Safe now: Watch.__deepcopy__() shares datastore ref (no memory leak) but copies dict data
         self.watch = deepcopy(self.datastore.data['watching'].get(watch_uuid))
+
         # Generic fetcher that should be extended (requests, playwright etc)
         self.fetcher = Fetcher()
 
+        # Load the last raw content checksum from file
+        self.read_last_raw_content_checksum()
+
+    def update_last_raw_content_checksum(self, checksum):
+        """
+        Save the raw content MD5 checksum to file.
+        This is used for skip logic - avoid reprocessing if raw HTML unchanged.
+        """
+        if not checksum:
+            return
+
+        watch = self.datastore.data['watching'].get(self.watch_uuid)
+        if not watch:
+            return
+
+        data_dir = watch.data_dir
+        if not data_dir:
+            return
+
+        watch.ensure_data_dir_exists()
+        checksum_file = os.path.join(data_dir, 'last-checksum.txt')
+
+        try:
+            with open(checksum_file, 'w', encoding='utf-8') as f:
+                f.write(checksum)
+            self.last_raw_content_checksum = checksum
+        except IOError as e:
+            logger.warning(f"Failed to write checksum file for {self.watch_uuid}: {e}")
+
+    def read_last_raw_content_checksum(self):
+        """
+        Read the last raw content MD5 checksum from file.
+        Returns None if file doesn't exist (first run) or can't be read.
+        """
+        watch = self.datastore.data['watching'].get(self.watch_uuid)
+        if not watch:
+            self.last_raw_content_checksum = None
+            return
+
+        data_dir = watch.data_dir
+        if not data_dir:
+            self.last_raw_content_checksum = None
+            return
+
+        checksum_file = os.path.join(data_dir, 'last-checksum.txt')
+
+        if not os.path.isfile(checksum_file):
+            self.last_raw_content_checksum = None
+            return
+
+        try:
+            with open(checksum_file, 'r', encoding='utf-8') as f:
+                self.last_raw_content_checksum = f.read().strip()
+        except IOError as e:
+            logger.warning(f"Failed to read checksum file for {self.watch_uuid}: {e}")
+            self.last_raw_content_checksum = None
+
+
+    async def validate_iana_url(self):
+        """Pre-flight SSRF check — runs DNS lookup in executor to avoid blocking the event loop.
+        Covers all fetchers (requests, playwright, puppeteer, plugins) since every fetch goes
+        through call_browser().
+        """
+        if strtobool(os.getenv('ALLOW_IANA_RESTRICTED_ADDRESSES', 'false')):
+            return
+        parsed = urlparse(self.watch.link)
+        if not parsed.hostname:
+            return
+        loop = asyncio.get_running_loop()
+        if await loop.run_in_executor(None, is_private_hostname, parsed.hostname):
+            raise Exception(
+                f"Fetch blocked: '{self.watch.link}' resolves to a private/reserved IP address. "
+                f"Set ALLOW_IANA_RESTRICTED_ADDRESSES=true to allow."
+            )
+
     async def call_browser(self, preferred_proxy_id=None):
 
         from requests.structures import CaseInsensitiveDict
@@ -40,6 +128,8 @@ class difference_detection_processor():
                     "file:// type access is denied for security reasons."
                 )
 
+        await self.validate_iana_url()
+
         # Requests, playwright, other browser via wss:// etc, fetch_extra_something
         prefer_fetch_backend = self.watch.get('fetch_backend', 'system')
 
@@ -103,7 +193,7 @@ class difference_detection_processor():
                                    )
 
         if self.watch.has_browser_steps:
-            self.fetcher.browser_steps = self.watch.get('browser_steps', [])
+            self.fetcher.browser_steps = browser_steps_get_valid_steps(self.watch.get('browser_steps', []))
             self.fetcher.browser_steps_screenshot_path = os.path.join(self.datastore.datastore_path, self.watch.get('uuid'))
 
         # Tweak the base config with the per-watch ones
@@ -170,6 +260,16 @@ class difference_detection_processor():
         # @todo .quit here could go on close object, so we can run JS if change-detected
         await self.fetcher.quit(watch=self.watch)
 
+        # Sanitize lone surrogates - these can appear when servers return malformed/mixed-encoding
+        # content that gets decoded into surrogate characters (e.g. \udcad). Without this,
+        # encode('utf-8') raises UnicodeEncodeError downstream in checksums, diffs, file writes, etc.
+        # Covers all fetchers (requests, playwright, puppeteer, selenium) in one place.
+        # Also note: By this point we SHOULD know the original encoding so it can safely convert to utf-8 for the rest of the app.
+        # See: https://github.com/dgtlmoon/changedetection.io/issues/3952
+
+        if self.fetcher.content and isinstance(self.fetcher.content, str):
+            self.fetcher.content = self.fetcher.content.encode('utf-8', errors='replace').decode('utf-8')
+
         # After init, call run_changedetection() which will do the actual change-detection
 
     def get_extra_watch_config(self, filename):
@@ -186,12 +286,12 @@ class difference_detection_processor():
         import os
 
         watch = self.datastore.data['watching'].get(self.watch_uuid)
-        watch_data_dir = watch.watch_data_dir
+        data_dir = watch.data_dir
 
-        if not watch_data_dir:
+        if not data_dir:
             return {}
 
-        filepath = os.path.join(watch_data_dir, filename)
+        filepath = os.path.join(data_dir, filename)
 
         if not os.path.isfile(filepath):
             return {}
@@ -216,16 +316,16 @@ class difference_detection_processor():
         import os
 
         watch = self.datastore.data['watching'].get(self.watch_uuid)
-        watch_data_dir = watch.watch_data_dir
+        data_dir = watch.data_dir
 
-        if not watch_data_dir:
-            logger.warning(f"Cannot save extra watch config {filename}: no watch_data_dir")
+        if not data_dir:
+            logger.warning(f"Cannot save extra watch config {filename}: no data_dir")
             return
 
         # Ensure directory exists
         watch.ensure_data_dir_exists()
 
-        filepath = os.path.join(watch_data_dir, filename)
+        filepath = os.path.join(data_dir, filename)
 
         try:
             # If merge is enabled, read existing data first
@@ -250,8 +350,16 @@ class difference_detection_processor():
         except IOError as e:
             logger.error(f"Failed to write extra watch config {filename}: {e}")
 
+    def get_raw_document_checksum(self):
+        checksum = None
+
+        if self.fetcher.content:
+            checksum = hashlib.md5(self.fetcher.content.encode('utf-8')).hexdigest()
+
+        return checksum
+
     @abstractmethod
-    def run_changedetection(self, watch):
+    def run_changedetection(self, watch, force_reprocess=False):
         update_obj = {'last_notification_error': False, 'last_error': False}
         some_data = 'xxxxx'
         update_obj["previous_md5"] = hashlib.md5(some_data.encode('utf-8')).hexdigest()

changedetectionio/processors/image_ssim_diff/__init__.py

@@ -12,6 +12,13 @@ processor_description = "Visual/Screenshot change detection (Fast)"
 processor_name = "image_ssim_diff"
 processor_weight = 2  # Lower weight = appears at top, heavier weight = appears lower (bottom)
 
+# Processor capabilities
+supports_visual_selector = True
+supports_browser_steps = True
+supports_text_filters_and_triggers = False
+supports_text_filters_and_triggers_elements = False
+supports_request_type = True
+
 PROCESSOR_CONFIG_NAME = f"{Path(__file__).parent.name}.json"
 
 # Subprocess timeout settings

changedetectionio/processors/image_ssim_diff/difference.py

@@ -130,7 +130,7 @@ def get_asset(asset_name, watch, datastore, request):
                 except Exception as e:
                     exception_container[0] = e
 
-            thread = threading.Thread(target=thread_target)
+            thread = threading.Thread(target=thread_target, daemon=True, name="ImageDiff-Asset")
             thread.start()
             thread.join(timeout=60)
 
@@ -284,7 +284,7 @@ def _draw_bounding_box_if_configured(img_bytes, watch, datastore):
             except Exception as e:
                 exception_container[0] = e
 
-        thread = threading.Thread(target=thread_target)
+        thread = threading.Thread(target=thread_target, daemon=True, name="ImageDiff-BoundingBox")
         thread.start()
         thread.join(timeout=15)
 
@@ -393,7 +393,7 @@ def render(watch, datastore, request, url_for, render_template, flash, redirect)
             except Exception as e:
                 exception_container[0] = e
 
-        thread = threading.Thread(target=thread_target)
+        thread = threading.Thread(target=thread_target, daemon=True, name="ImageDiff-ChangePercentage")
         thread.start()
         thread.join(timeout=60)
 
@@ -414,7 +414,7 @@ def render(watch, datastore, request, url_for, render_template, flash, redirect)
 
     # Load historical data if available (for charts/visualization)
     comparison_data = {}
-    comparison_config_path = os.path.join(watch.watch_data_dir, "visual_comparison_data.json")
+    comparison_config_path = os.path.join(watch.data_dir, "visual_comparison_data.json")
     if os.path.isfile(comparison_config_path):
         try:
             with open(comparison_config_path, 'r') as f:

changedetectionio/processors/image_ssim_diff/edit_hook.py

@@ -90,7 +90,7 @@ def on_config_save(watch, processor_config, datastore):
             processor_config['auto_track_region'] = False
 
             # Remove old template file if exists
-            template_path = os.path.join(watch.watch_data_dir, CROPPED_IMAGE_TEMPLATE_FILENAME)
+            template_path = os.path.join(watch.data_dir, CROPPED_IMAGE_TEMPLATE_FILENAME)
             if os.path.exists(template_path):
                 os.remove(template_path)
                 logger.debug(f"Removed old template file: {template_path}")

changedetectionio/processors/image_ssim_diff/image_handler/isolated_libvips.py

@@ -13,14 +13,9 @@ Research: https://github.com/libvips/pyvips/issues/234
 
 import multiprocessing
 
-# CRITICAL: Use 'spawn' instead of 'fork' to avoid inheriting parent's
+# CRITICAL: Use 'spawn' context instead of 'fork' to avoid inheriting parent's
 # LibVIPS threading state which can cause hangs in gaussblur operations
 # https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods
-try:
-    multiprocessing.set_start_method('spawn', force=False)
-except RuntimeError:
-    # Already set, ignore
-    pass
 
 
 def _worker_generate_diff(conn, img_bytes_from, img_bytes_to, threshold, blur_sigma, max_width, max_height):
@@ -95,9 +90,10 @@ def generate_diff_isolated(img_bytes_from, img_bytes_to, threshold, blur_sigma,
     Returns:
         bytes: JPEG diff image or None on failure
     """
-    parent_conn, child_conn = multiprocessing.Pipe()
+    ctx = multiprocessing.get_context('spawn')
+    parent_conn, child_conn = ctx.Pipe()
 
-    p = multiprocessing.Process(
+    p = ctx.Process(
         target=_worker_generate_diff,
         args=(child_conn, img_bytes_from, img_bytes_to, threshold, blur_sigma, max_width, max_height)
     )
@@ -140,7 +136,8 @@ def calculate_change_percentage_isolated(img_bytes_from, img_bytes_to, threshold
     Returns:
         float: Change percentage
     """
-    parent_conn, child_conn = multiprocessing.Pipe()
+    ctx = multiprocessing.get_context('spawn')
+    parent_conn, child_conn = ctx.Pipe()
 
     def _worker_calculate(conn):
         try:
@@ -185,7 +182,7 @@ def calculate_change_percentage_isolated(img_bytes_from, img_bytes_to, threshold
         finally:
             conn.close()
 
-    p = multiprocessing.Process(target=_worker_calculate, args=(child_conn,))
+    p = ctx.Process(target=_worker_calculate, args=(child_conn,))
     p.start()
 
     result = 0.0
@@ -233,7 +230,8 @@ def compare_images_isolated(img_bytes_from, img_bytes_to, threshold, blur_sigma,
         tuple: (changed_detected, change_percentage)
     """
     print(f"[Parent] Starting compare_images_isolated subprocess", flush=True)
-    parent_conn, child_conn = multiprocessing.Pipe()
+    ctx = multiprocessing.get_context('spawn')
+    parent_conn, child_conn = ctx.Pipe()
 
     def _worker_compare(conn):
         try:
@@ -301,7 +299,7 @@ def compare_images_isolated(img_bytes_from, img_bytes_to, threshold, blur_sigma,
         finally:
             conn.close()
 
-    p = multiprocessing.Process(target=_worker_compare, args=(child_conn,))
+    p = ctx.Process(target=_worker_compare, args=(child_conn,))
     print(f"[Parent] Starting subprocess (pid will be assigned)", flush=True)
     p.start()
     print(f"[Parent] Subprocess started (pid={p.pid}), waiting for result (30s timeout)", flush=True)

changedetectionio/processors/image_ssim_diff/processor.py

@@ -30,7 +30,7 @@ class perform_site_check(difference_detection_processor):
     # Override to use PNG format for better image comparison (JPEG compression creates noise)
     screenshot_format = SCREENSHOT_FORMAT_PNG
 
-    def run_changedetection(self, watch):
+    def run_changedetection(self, watch, force_reprocess=False):
         """
         Perform screenshot comparison using OpenCV subprocess handler.
 
@@ -204,7 +204,7 @@ class perform_site_check(difference_detection_processor):
                 except Exception as e:
                     exception_container[0] = e
 
-            thread = threading.Thread(target=thread_target)
+            thread = threading.Thread(target=thread_target, daemon=True, name="ImageDiff-Processor")
             thread.start()
             thread.join(timeout=60)
 

changedetectionio/processors/restock_diff/__init__.py

@@ -4,6 +4,13 @@ from changedetectionio.model.Watch import model as BaseWatch
 from typing import Union
 import re
 
+# Processor capabilities
+supports_visual_selector = True
+supports_browser_steps = True
+supports_text_filters_and_triggers = True
+supports_text_filters_and_triggers_elements = True
+supports_request_type = True
+
 class Restock(dict):
 
     def parse_currency(self, raw_value: str) -> Union[float, None]:
@@ -24,6 +31,7 @@ class Restock(dict):
 
         if standardized_value:
             # Convert to float
+            # @todo locale needs to be the locale of the webpage
             return float(parse_decimal(standardized_value, locale='en'))
 
         return None
@@ -60,10 +68,6 @@ class Watch(BaseWatch):
         super().__init__(*arg, **kw)
         self['restock'] = Restock(kw['default']['restock']) if kw.get('default') and kw['default'].get('restock') else Restock()
 
-        self['restock_settings'] = kw['default']['restock_settings'] if kw.get('default',{}).get('restock_settings') else {
-            'follow_price_changes': True,
-            'in_stock_processing' : 'in_stock_only'
-        } #@todo update
 
     def clear_watch(self):
         super().clear_watch()

changedetectionio/processors/restock_diff/api.yaml

@@ -0,0 +1,149 @@
+components:
+  schemas:
+    processor_config_restock_diff:
+      type: object
+      description: Configuration for the restock_diff processor (restock and price tracking)
+      properties:
+        in_stock_processing:
+          type: string
+          enum: [in_stock_only, all_changes, 'off']
+          default: in_stock_only
+          description: |
+            When to trigger on stock changes:
+            - `in_stock_only`: Only trigger on Out Of Stock -> In Stock transitions
+            - `all_changes`: Trigger on any availability change
+            - `off`: Disable stock/availability tracking
+        follow_price_changes:
+          type: boolean
+          default: true
+          description: Monitor and track price changes
+        price_change_min:
+          type: [number, 'null']
+          description: Trigger a notification when the price drops below this value
+        price_change_max:
+          type: [number, 'null']
+          description: Trigger a notification when the price rises above this value
+        price_change_threshold_percent:
+          type: [number, 'null']
+          minimum: 0
+          maximum: 100
+          description: Minimum price change percentage since the original price to trigger a notification
+
+paths:
+  /watch:
+    post:
+      x-code-samples:
+        - lang: 'curl'
+          label: 'Restock & price tracking'
+          source: |
+            curl -X POST "http://localhost:5000/api/v1/watch" \
+              -H "x-api-key: YOUR_API_KEY" \
+              -H "Content-Type: application/json" \
+              -d '{
+                "url": "https://example.com/product",
+                "processor": "restock_diff",
+                "processor_config_restock_diff": {
+                  "in_stock_processing": "in_stock_only",
+                  "follow_price_changes": true,
+                  "price_change_threshold_percent": 5
+                }
+              }'
+        - lang: 'Python'
+          label: 'Restock & price tracking'
+          source: |
+            import requests
+
+            headers = {
+                'x-api-key': 'YOUR_API_KEY',
+                'Content-Type': 'application/json'
+            }
+            data = {
+                'url': 'https://example.com/product',
+                'processor': 'restock_diff',
+                'processor_config_restock_diff': {
+                    'in_stock_processing': 'in_stock_only',
+                    'follow_price_changes': True,
+                    'price_change_threshold_percent': 5,
+                }
+            }
+            response = requests.post('http://localhost:5000/api/v1/watch',
+                                     headers=headers, json=data)
+            print(response.json())
+
+  /watch/{uuid}:
+    put:
+      x-code-samples:
+        - lang: 'curl'
+          label: 'Update restock config'
+          source: |
+            curl -X PUT "http://localhost:5000/api/v1/watch/YOUR-UUID" \
+              -H "x-api-key: YOUR_API_KEY" \
+              -H "Content-Type: application/json" \
+              -d '{
+                "processor_config_restock_diff": {
+                  "in_stock_processing": "all_changes",
+                  "follow_price_changes": true,
+                  "price_change_min": 10.00,
+                  "price_change_max": 500.00
+                }
+              }'
+        - lang: 'Python'
+          label: 'Update restock config'
+          source: |
+            import requests
+
+            headers = {
+                'x-api-key': 'YOUR_API_KEY',
+                'Content-Type': 'application/json'
+            }
+            uuid = 'YOUR-UUID'
+            data = {
+                'processor_config_restock_diff': {
+                    'in_stock_processing': 'all_changes',
+                    'follow_price_changes': True,
+                    'price_change_min': 10.00,
+                    'price_change_max': 500.00,
+                }
+            }
+            response = requests.put(f'http://localhost:5000/api/v1/watch/{uuid}',
+                                    headers=headers, json=data)
+            print(response.text)
+
+  /tag/{uuid}:
+    put:
+      x-code-samples:
+        - lang: 'curl'
+          label: 'Set restock config on group/tag'
+          source: |
+            curl -X PUT "http://localhost:5000/api/v1/tag/YOUR-TAG-UUID" \
+              -H "x-api-key: YOUR_API_KEY" \
+              -H "Content-Type: application/json" \
+              -d '{
+                "overrides_watch": true,
+                "processor_config_restock_diff": {
+                  "in_stock_processing": "in_stock_only",
+                  "follow_price_changes": true,
+                  "price_change_threshold_percent": 10
+                }
+              }'
+        - lang: 'Python'
+          label: 'Set restock config on group/tag'
+          source: |
+            import requests
+
+            headers = {
+                'x-api-key': 'YOUR_API_KEY',
+                'Content-Type': 'application/json'
+            }
+            tag_uuid = 'YOUR-TAG-UUID'
+            data = {
+                'overrides_watch': True,
+                'processor_config_restock_diff': {
+                    'in_stock_processing': 'in_stock_only',
+                    'follow_price_changes': True,
+                    'price_change_threshold_percent': 10,
+                }
+            }
+            response = requests.put(f'http://localhost:5000/api/v1/tag/{tag_uuid}',
+                                    headers=headers, json=data)
+            print(response.text)

changedetectionio/processors/restock_diff/forms.py

@@ -31,7 +31,7 @@ class RestockSettingsForm(Form):
     follow_price_changes = BooleanField(_l('Follow price changes'), default=True)
 
 class processor_settings_form(processor_text_json_diff_form):
-    restock_settings = FormField(RestockSettingsForm)
+    processor_config_restock_diff = FormField(RestockSettingsForm)
 
     def extra_tab_content(self):
         return _l('Restock & Price Detection')
@@ -48,34 +48,34 @@ class processor_settings_form(processor_text_json_diff_form):
 
         output += """
         {% from '_helpers.html' import render_field, render_checkbox_field, render_button %}
-        <script>        
+        <script>
             $(document).ready(function () {
-                toggleOpacity('#restock_settings-follow_price_changes', '.price-change-minmax', true);
+                toggleOpacity('#processor_config_restock_diff-follow_price_changes', '.price-change-minmax', true);
             });
         </script>
 
         <fieldset id="restock-fieldset-price-group">
             <div class="pure-control-group">
                 <fieldset class="pure-group inline-radio">
-                    {{ render_field(form.restock_settings.in_stock_processing) }}
+                    {{ render_field(form.processor_config_restock_diff.in_stock_processing) }}
                 </fieldset>
                 <fieldset class="pure-group">
-                    {{ render_checkbox_field(form.restock_settings.follow_price_changes) }}
+                    {{ render_checkbox_field(form.processor_config_restock_diff.follow_price_changes) }}
                     <span class="pure-form-message-inline">Changes in price should trigger a notification</span>
                 </fieldset>
-                <fieldset class="pure-group price-change-minmax">               
-                    {{ render_field(form.restock_settings.price_change_min, placeholder=watch.get('restock', {}).get('price')) }}
+                <fieldset class="pure-group price-change-minmax">
+                    {{ render_field(form.processor_config_restock_diff.price_change_min, placeholder=watch.get('restock', {}).get('price')) }}
                     <span class="pure-form-message-inline">Minimum amount, Trigger a change/notification when the price drops <i>below</i> this value.</span>
                 </fieldset>
                 <fieldset class="pure-group price-change-minmax">
-                    {{ render_field(form.restock_settings.price_change_max, placeholder=watch.get('restock', {}).get('price')) }}
+                    {{ render_field(form.processor_config_restock_diff.price_change_max, placeholder=watch.get('restock', {}).get('price')) }}
                     <span class="pure-form-message-inline">Maximum amount, Trigger a change/notification when the price rises <i>above</i> this value.</span>
                 </fieldset>
                 <fieldset class="pure-group price-change-minmax">
-                    {{ render_field(form.restock_settings.price_change_threshold_percent) }}
+                    {{ render_field(form.processor_config_restock_diff.price_change_threshold_percent) }}
                     <span class="pure-form-message-inline">Price must change more than this % to trigger a change since the first check.</span><br>
                     <span class="pure-form-message-inline">For example, If the product is $1,000 USD originally, <strong>2%</strong> would mean it has to change more than $20 since the first check.</span><br>
-                </fieldset>                
+                </fieldset>
             </div>
         </fieldset>
         """

changedetectionio/processors/restock_diff/processor.py

@@ -2,6 +2,7 @@ from ..base import difference_detection_processor
 from ..exceptions import ProcessorException
 from . import Restock
 from loguru import logger
+from changedetectionio.content_fetchers.exceptions import checksumFromPreviousCheckWasTheSame
 
 import urllib3
 import time
@@ -56,6 +57,259 @@ def _deduplicate_prices(data):
     return list(unique_data)
 
 
+# =============================================================================
+# MEMORY MANAGEMENT: Why We Use Multiprocessing (Linux Only)
+# =============================================================================
+#
+# The get_itemprop_availability() function uses 'extruct' to parse HTML metadata
+# (JSON-LD, microdata, OpenGraph, etc). Extruct internally uses lxml, which wraps
+# libxml2 - a C library that allocates memory at the C level.
+#
+# Memory Leak Problem:
+# --------------------
+# 1. lxml's document_fromstring() creates thousands of Python objects backed by
+#    C-level allocations (nodes, attributes, text content)
+# 2. Python's garbage collector can mark these objects as collectible, but
+#    cannot force the OS to reclaim the actual C-level memory
+# 3. malloc/free typically doesn't return memory to OS - it just marks it as
+#    "free in the process address space"
+# 4. With repeated parsing of large HTML (5MB+ pages), memory accumulates even
+#    after Python GC runs
+#
+# Why Multiprocessing Fixes This:
+# --------------------------------
+# When a subprocess exits, the OS forcibly reclaims ALL memory including C-level
+# allocations that Python GC couldn't release. This ensures clean memory state
+# after each extraction.
+#
+# Performance Impact:
+# -------------------
+# - Memray analysis showed 1.2M document_fromstring allocations per page
+# - Without subprocess: memory grows by ~50-500MB per parse and lingers
+# - With subprocess: ~35MB overhead but forces full cleanup after each run
+# - Trade-off: 35MB resource_tracker vs 500MB+ accumulated leak = much better at scale
+#
+# References:
+# -----------
+# - lxml memory issues: https://medium.com/devopss-hole/python-lxml-memory-leak-b8d0b1000dc7
+# - libxml2 caching behavior: https://www.mail-archive.com/lxml@python.org/msg00026.html
+# - GC limitations with C extensions: https://benbernardblog.com/tracking-down-a-freaky-python-memory-leak-part-2/
+#
+# Additional Context:
+# -------------------
+# - jsonpath_ng (used to query the parsed data) is pure Python and doesn't leak
+# - The leak is specifically from lxml's document parsing, not the JSONPath queries
+# - Linux-only because multiprocessing spawn is well-tested there; other platforms
+#   use direct call as fallback
+#
+# Alternative Solution (Future Optimization):
+# -------------------------------------------
+# This entire problem could be avoided by using regex to extract just the machine
+# data blocks (JSON-LD, microdata, OpenGraph tags) BEFORE parsing with lxml:
+#
+#   1. Use regex to extract <script type="application/ld+json">...</script> blocks
+#   2. Use regex to extract <meta property="og:*"> tags
+#   3. Use regex to find itemprop/itemtype attributes and their containing elements
+#   4. Parse ONLY those extracted snippets instead of the entire HTML document
+#
+# Benefits:
+#   - Avoids parsing 5MB of HTML when we only need a few KB of metadata
+#   - Eliminates the lxml memory leak entirely
+#   - Faster extraction (regex is much faster than DOM parsing)
+#   - No subprocess overhead needed
+#
+# Trade-offs:
+#   - Regex for HTML is brittle (comments, CDATA, edge cases)
+#   - Microdata extraction would be complex (need to track element boundaries)
+#   - Would need extensive testing to ensure we don't miss valid data
+#   - extruct is battle-tested; regex solution would need similar maturity
+#
+# For now, the subprocess approach is safer and leverages existing extruct code.
+# =============================================================================
+
+
+def _extract_itemprop_availability_worker(pipe_conn):
+    """
+    Subprocess worker for itemprop extraction (Linux memory management).
+
+    Uses spawn multiprocessing to isolate extruct/lxml memory allocations.
+    When the subprocess exits, the OS reclaims ALL memory including lxml's
+    C-level allocations that Python's GC cannot release.
+
+    Args:
+        pipe_conn: Pipe connection to receive HTML and send result
+    """
+    import json
+    import gc
+
+    html_content = None
+    result_data = None
+
+    try:
+        # Receive HTML as raw bytes (no pickle)
+        html_bytes = pipe_conn.recv_bytes()
+        html_content = html_bytes.decode('utf-8')
+
+        # Explicitly delete html_bytes to free memory
+        del html_bytes
+        gc.collect()
+
+        # Perform extraction in subprocess (uses extruct/lxml)
+        result_data = get_itemprop_availability(html_content)
+
+        # Convert Restock object to dict for JSON serialization
+        result = {
+            'success': True,
+            'data': dict(result_data) if result_data else {}
+        }
+        pipe_conn.send_bytes(json.dumps(result).encode('utf-8'))
+
+        # Clean up before exit
+        del result_data, html_content, result
+        gc.collect()
+
+    except MoreThanOnePriceFound:
+        # Serialize the specific exception type
+        result = {
+            'success': False,
+            'exception_type': 'MoreThanOnePriceFound'
+        }
+        pipe_conn.send_bytes(json.dumps(result).encode('utf-8'))
+
+    except Exception as e:
+        # Serialize other exceptions
+        result = {
+            'success': False,
+            'exception_type': type(e).__name__,
+            'exception_message': str(e)
+        }
+        pipe_conn.send_bytes(json.dumps(result).encode('utf-8'))
+
+    finally:
+        # Final cleanup before subprocess exits
+        # Variables may already be deleted in try block, so use try/except
+        try:
+            del html_content
+        except (NameError, UnboundLocalError):
+            pass
+        try:
+            del result_data
+        except (NameError, UnboundLocalError):
+            pass
+        gc.collect()
+        pipe_conn.close()
+
+
+def extract_itemprop_availability_safe(html_content) -> Restock:
+    """
+    Extract itemprop availability with hybrid approach for memory efficiency.
+
+    Strategy (fastest to slowest, least to most memory):
+    1. Try pure Python extraction (JSON-LD, OpenGraph, microdata) - covers 80%+ of cases
+    2. Fall back to extruct with subprocess isolation on Linux for complex cases
+
+    Args:
+        html_content: HTML string to parse
+
+    Returns:
+        Restock: Extracted availability data
+
+    Raises:
+        MoreThanOnePriceFound: When multiple prices detected
+        Other exceptions: From extruct/parsing
+    """
+    import platform
+
+    # Step 1: Try pure Python extraction first (fast, no lxml, no memory leak)
+    try:
+        from .pure_python_extractor import extract_metadata_pure_python, query_price_availability
+
+        logger.trace("Attempting pure Python metadata extraction (no lxml)")
+        extracted_data = extract_metadata_pure_python(html_content)
+        price_data = query_price_availability(extracted_data)
+
+        # If we got price AND availability, we're done!
+        if price_data.get('price') and price_data.get('availability'):
+            result = Restock(price_data)
+            logger.debug(f"Pure Python extraction successful: {dict(result)}")
+            return result
+
+        # If we got some data but not everything, still try extruct for completeness
+        if price_data.get('price') or price_data.get('availability'):
+            logger.debug(f"Pure Python extraction partial: {price_data}, will try extruct for completeness")
+
+    except Exception as e:
+        logger.debug(f"Pure Python extraction failed: {e}, falling back to extruct")
+
+    # Step 2: Fall back to extruct (uses lxml, needs subprocess on Linux)
+    logger.trace("Falling back to extruct (lxml-based) with subprocess isolation")
+
+    # Only use subprocess isolation on Linux
+    # Other platforms may have issues with spawn or don't need the aggressive memory management
+    if platform.system() == 'Linux':
+        import multiprocessing
+        import json
+        import gc
+
+        try:
+            ctx = multiprocessing.get_context('spawn')
+            parent_conn, child_conn = ctx.Pipe()
+            p = ctx.Process(target=_extract_itemprop_availability_worker, args=(child_conn,))
+            p.start()
+
+            # Send HTML as raw bytes (no pickle)
+            html_bytes = html_content.encode('utf-8')
+            parent_conn.send_bytes(html_bytes)
+
+            # Explicitly delete html_bytes copy immediately after sending
+            del html_bytes
+            gc.collect()
+
+            # Receive result as JSON
+            result_bytes = parent_conn.recv_bytes()
+            result = json.loads(result_bytes.decode('utf-8'))
+
+            # Wait for subprocess to complete
+            p.join()
+
+            # Close pipes
+            parent_conn.close()
+            child_conn.close()
+
+            # Clean up all subprocess-related objects
+            del p, parent_conn, child_conn, result_bytes
+            gc.collect()
+
+            # Handle result or re-raise exception
+            if result['success']:
+                # Reconstruct Restock object from dict
+                restock_obj = Restock(result['data'])
+                # Clean up result dict
+                del result
+                gc.collect()
+                return restock_obj
+            else:
+                # Re-raise the exception that occurred in subprocess
+                exception_type = result['exception_type']
+                exception_msg = result.get('exception_message', '')
+                del result
+                gc.collect()
+
+                if exception_type == 'MoreThanOnePriceFound':
+                    raise MoreThanOnePriceFound()
+                else:
+                    raise Exception(f"{exception_type}: {exception_msg}")
+
+        except Exception as e:
+            # If multiprocessing itself fails, log and fall back to direct call
+            logger.warning(f"Subprocess extraction failed: {e}, falling back to direct call")
+            gc.collect()
+            return get_itemprop_availability(html_content)
+    else:
+        # Non-Linux: direct call (no subprocess overhead needed)
+        return get_itemprop_availability(html_content)
+
+
 # should return Restock()
 # add casting?
 def get_itemprop_availability(html_content) -> Restock:
@@ -150,61 +404,82 @@ class perform_site_check(difference_detection_processor):
     screenshot = None
     xpath_data = None
 
-    def run_changedetection(self, watch):
+    def run_changedetection(self, watch, force_reprocess=False):
         import hashlib
 
         if not watch:
             raise Exception("Watch no longer exists.")
 
+        current_raw_document_checksum = self.get_raw_document_checksum()
+        # Skip processing only if BOTH conditions are true:
+        # 1. HTML content unchanged (checksum matches last saved checksum)
+        # 2. Watch configuration was not edited (including trigger_text, filters, etc.)
+        # The was_edited flag handles all watch configuration changes, so we don't need
+        # separate checks for trigger_text or other processing rules.
+        if (not force_reprocess and
+            not watch.was_edited and
+            self.last_raw_content_checksum and
+            self.last_raw_content_checksum == current_raw_document_checksum):
+            raise checksumFromPreviousCheckWasTheSame()
+
         # Unset any existing notification error
         update_obj = {'last_notification_error': False, 'last_error': False, 'restock':  Restock()}
 
         self.screenshot = self.fetcher.screenshot
         self.xpath_data = self.fetcher.xpath_data
 
-        # Track the content type
-        update_obj['content_type'] = self.fetcher.headers.get('Content-Type', '')
+        # Track the content type (readonly field, doesn't trigger was_edited)
+        update_obj['content-type'] = self.fetcher.headers.get('Content-Type', '')  # Use hyphen (matches OpenAPI spec)
         update_obj["last_check_status"] = self.fetcher.get_last_status_code()
 
+        # Save the raw content checksum to file (processor implementation detail, not watch config)
+        self.update_last_raw_content_checksum(current_raw_document_checksum)
+
         # Only try to process restock information (like scraping for keywords) if the page was actually rendered correctly.
         # Otherwise it will assume "in stock" because nothing suggesting the opposite was found
-        from ...html_tools import html_to_text
-        text = html_to_text(self.fetcher.content)
-        logger.debug(f"Length of text after conversion: {len(text)}")
-        if not len(text):
-            from ...content_fetchers.exceptions import ReplyWithContentButNoText
-            raise ReplyWithContentButNoText(url=watch.link,
-                                            status_code=self.fetcher.get_last_status_code(),
-                                            screenshot=self.fetcher.screenshot,
-                                            html_content=self.fetcher.content,
-                                            xpath_data=self.fetcher.xpath_data
-                                            )
+#useless
+#        from ...html_tools import html_to_text
+#        text = html_to_text(self.fetcher.content)
+#        logger.debug(f"Length of text after conversion: {len(text)}")
+#        if not len(text):
+#            from ...content_fetchers.exceptions import ReplyWithContentButNoText
+#            raise ReplyWithContentButNoText(url=watch.link,
+#                                            status_code=self.fetcher.get_last_status_code(),
+#                                            screenshot=self.fetcher.screenshot,
+#                                            html_content=self.fetcher.content,
+#                                            xpath_data=self.fetcher.xpath_data
+#                                            )
 
         # Which restock settings to compare against?
-        restock_settings = watch.get('restock_settings', {})
+        # Settings are stored in restock_diff.json (migrated from watch.json by update_30).
+        _extra_config = self.get_extra_watch_config('restock_diff.json')
+        restock_settings = _extra_config.get('restock_diff') or {
+            'follow_price_changes': True,
+            'in_stock_processing': 'in_stock_only',
+        }
 
         # See if any tags have 'activate for individual watches in this tag/group?' enabled and use the first we find
         for tag_uuid in watch.get('tags'):
             tag = self.datastore.data['settings']['application']['tags'].get(tag_uuid, {})
             if tag.get('overrides_watch'):
-                restock_settings = tag.get('restock_settings', {})
+                restock_settings = tag.get('processor_config_restock_diff') or {}
                 logger.info(f"Watch {watch.get('uuid')} - Tag '{tag.get('title')}' selected for restock settings override")
                 break
 
 
         itemprop_availability = {}
+        multiple_prices_found = False
 
         # Try built-in extraction first, this will scan metadata in the HTML
+        # On Linux, this runs in a subprocess to prevent lxml/extruct memory leaks
         try:
-            itemprop_availability = get_itemprop_availability(self.fetcher.content)
+            itemprop_availability = extract_itemprop_availability_safe(self.fetcher.content)
         except MoreThanOnePriceFound as e:
-            # Add the real data
-            raise ProcessorException(message="Cannot run, more than one price detected, this plugin is only for product pages with ONE product, try the content-change detection mode.",
-                                     url=watch.get('url'),
-                                     status_code=self.fetcher.get_last_status_code(),
-                                     screenshot=self.fetcher.screenshot,
-                                     xpath_data=self.fetcher.xpath_data
-                                     )
+            # Don't raise immediately - let plugins try to handle this case
+            # Plugins might be able to determine which price is correct
+            logger.warning(f"Built-in detection found multiple prices on {watch.get('url')}, will try plugin override")
+            multiple_prices_found = True
+            itemprop_availability = {}
 
         # If built-in extraction didn't get both price AND availability, try plugin override
         # Only check plugin if this watch is using a fetcher that might provide better data
@@ -216,9 +491,21 @@ class perform_site_check(difference_detection_processor):
             from changedetectionio.pluggy_interface import get_itemprop_availability_from_plugin
             fetcher_name = watch.get('fetch_backend', 'html_requests')
 
-            # Only try plugin override if not using system default (which might be anything)
-            if fetcher_name and fetcher_name != 'system':
-                logger.debug("Calling extra plugins for getting item price/availability")
+            # Resolve 'system' to the actual fetcher being used
+            # This allows plugins to work even when watch uses "system settings default"
+            if fetcher_name == 'system':
+                # Get the actual fetcher that was used (from self.fetcher)
+                # Fetcher class name gives us the actual backend (e.g., 'html_requests', 'html_webdriver')
+                actual_fetcher = type(self.fetcher).__name__
+                if 'html_requests' in actual_fetcher.lower():
+                    fetcher_name = 'html_requests'
+                elif 'webdriver' in actual_fetcher.lower() or 'playwright' in actual_fetcher.lower():
+                    fetcher_name = 'html_webdriver'
+                logger.debug(f"Resolved 'system' fetcher to actual fetcher: {fetcher_name}")
+
+            # Try plugin override - plugins can decide if they support this fetcher
+            if fetcher_name:
+                logger.debug(f"Calling extra plugins for getting item price/availability (fetcher: {fetcher_name})")
                 plugin_availability = get_itemprop_availability_from_plugin(self.fetcher.content, fetcher_name, self.fetcher, watch.link)
 
                 if plugin_availability:
@@ -233,6 +520,16 @@ class perform_site_check(difference_detection_processor):
                 if not plugin_availability:
                     logger.debug("No item price/availability from plugins")
 
+        # If we had multiple prices and plugins also failed, NOW raise the exception
+        if multiple_prices_found and not itemprop_availability.get('price'):
+            raise ProcessorException(
+                message="Cannot run, more than one price detected, this plugin is only for product pages with ONE product, try the content-change detection mode.",
+                url=watch.get('url'),
+                status_code=self.fetcher.get_last_status_code(),
+                screenshot=self.fetcher.screenshot,
+                xpath_data=self.fetcher.xpath_data
+            )
+
         # Something valid in get_itemprop_availability() by scraping metadata ?
         if itemprop_availability.get('price') or itemprop_availability.get('availability'):
             # Store for other usage

changedetectionio/processors/restock_diff/pure_python_extractor.py

@@ -0,0 +1,289 @@
+"""
+Pure Python metadata extractor - no lxml, no memory leaks.
+
+This module provides a fast, memory-efficient alternative to extruct for common
+e-commerce metadata extraction. It handles:
+- JSON-LD (covers 80%+ of modern sites)
+- OpenGraph meta tags
+- Basic microdata attributes
+
+Uses Python's built-in html.parser instead of lxml/libxml2, avoiding C-level
+memory allocation issues. For edge cases, the main processor can fall back to
+extruct (with subprocess isolation on Linux).
+"""
+
+from html.parser import HTMLParser
+import json
+import re
+from loguru import logger
+
+
+class JSONLDExtractor(HTMLParser):
+    """
+    Extract JSON-LD structured data from HTML.
+
+    Finds all <script type="application/ld+json"> tags and parses their content.
+    Handles multiple JSON-LD blocks on the same page.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.in_jsonld = False
+        self.data = []  # List of all parsed JSON-LD objects
+        self.current_script = []
+
+    def handle_starttag(self, tag, attrs):
+        if tag == 'script':
+            # Check if this is a JSON-LD script tag
+            for attr, value in attrs:
+                if attr == 'type' and value == 'application/ld+json':
+                    self.in_jsonld = True
+                    self.current_script = []
+                    break
+
+    def handle_data(self, data):
+        if self.in_jsonld:
+            self.current_script.append(data)
+
+    def handle_endtag(self, tag):
+        if tag == 'script' and self.in_jsonld:
+            # Parse the accumulated script content
+            script_content = ''.join(self.current_script)
+            if script_content.strip():
+                try:
+                    # Parse JSON (handles both objects and arrays)
+                    parsed = json.loads(script_content)
+                    if isinstance(parsed, list):
+                        self.data.extend(parsed)
+                    else:
+                        self.data.append(parsed)
+                except json.JSONDecodeError as e:
+                    logger.debug(f"Failed to parse JSON-LD: {e}")
+                    pass
+
+            self.in_jsonld = False
+            self.current_script = []
+
+
+class OpenGraphExtractor(HTMLParser):
+    """
+    Extract OpenGraph meta tags from HTML.
+
+    Finds <meta property="og:*"> tags commonly used for social media sharing.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.og_data = {}
+
+    def handle_starttag(self, tag, attrs):
+        if tag == 'meta':
+            attrs_dict = dict(attrs)
+            prop = attrs_dict.get('property', '')
+
+            # Extract OpenGraph properties
+            if prop.startswith('og:'):
+                content = attrs_dict.get('content', '')
+                if content:
+                    self.og_data[prop] = content
+
+
+class MicrodataExtractor(HTMLParser):
+    """
+    Extract basic microdata attributes from HTML.
+
+    Finds elements with itemprop attributes. This is a simplified extractor
+    that doesn't handle nested itemscope/itemtype hierarchies - for complex
+    cases, use extruct as fallback.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.microdata = {}
+        self.current_itemprop = None
+
+    def handle_starttag(self, tag, attrs):
+        attrs_dict = dict(attrs)
+
+        if 'itemprop' in attrs_dict:
+            itemprop = attrs_dict['itemprop']
+
+            # Price/currency/availability can be in content/href attributes
+            if itemprop == 'price':
+                if 'content' in attrs_dict:
+                    self.microdata['price'] = attrs_dict['content']
+                else:
+                    self.current_itemprop = 'price'
+
+            elif itemprop == 'priceCurrency':
+                if 'content' in attrs_dict:
+                    self.microdata['currency'] = attrs_dict['content']
+                else:
+                    self.current_itemprop = 'priceCurrency'
+
+            elif itemprop == 'availability':
+                # Can be in href (link) or content (meta)
+                if 'href' in attrs_dict:
+                    self.microdata['availability'] = attrs_dict['href']
+                elif 'content' in attrs_dict:
+                    self.microdata['availability'] = attrs_dict['content']
+                else:
+                    self.current_itemprop = 'availability'
+
+    def handle_data(self, data):
+        # Capture text content for itemprop elements
+        if self.current_itemprop == 'price':
+            # Try to extract numeric price from text
+            try:
+                price_text = re.sub(r'[^\d.]', '', data.strip())
+                if price_text:
+                    self.microdata['price'] = float(price_text)
+            except ValueError:
+                pass
+        elif self.current_itemprop == 'priceCurrency':
+            currency = data.strip()
+            if currency:
+                self.microdata['currency'] = currency
+        elif self.current_itemprop == 'availability':
+            availability = data.strip()
+            if availability:
+                self.microdata['availability'] = availability
+
+    def handle_endtag(self, tag):
+        # Reset current itemprop after closing tag
+        self.current_itemprop = None
+
+
+def extract_metadata_pure_python(html_content):
+    """
+    Extract structured metadata from HTML using pure Python parsers.
+
+    Returns a dict with three keys:
+    - 'json-ld': List of parsed JSON-LD objects
+    - 'opengraph': Dict of OpenGraph properties
+    - 'microdata': Dict of microdata properties
+
+    Args:
+        html_content: HTML string to parse
+
+    Returns:
+        dict: Extracted metadata in three formats
+    """
+    result = {
+        'json-ld': [],
+        'opengraph': {},
+        'microdata': {}
+    }
+
+    # Extract JSON-LD
+    try:
+        jsonld_extractor = JSONLDExtractor()
+        jsonld_extractor.feed(html_content)
+        result['json-ld'] = jsonld_extractor.data
+        logger.trace(f"Pure Python: Found {len(jsonld_extractor.data)} JSON-LD blocks")
+    except Exception as e:
+        logger.debug(f"JSON-LD extraction failed: {e}")
+
+    # Extract OpenGraph
+    try:
+        og_extractor = OpenGraphExtractor()
+        og_extractor.feed(html_content)
+        result['opengraph'] = og_extractor.og_data
+        if result['opengraph']:
+            logger.trace(f"Pure Python: Found {len(og_extractor.og_data)} OpenGraph tags")
+    except Exception as e:
+        logger.debug(f"OpenGraph extraction failed: {e}")
+
+    # Extract Microdata
+    try:
+        microdata_extractor = MicrodataExtractor()
+        microdata_extractor.feed(html_content)
+        result['microdata'] = microdata_extractor.microdata
+        if result['microdata']:
+            logger.trace(f"Pure Python: Found microdata: {result['microdata']}")
+    except Exception as e:
+        logger.debug(f"Microdata extraction failed: {e}")
+
+    return result
+
+
+def query_price_availability(extracted_data):
+    """
+    Query extracted metadata for price and availability information.
+
+    Uses jsonpath_ng to query JSON-LD data (same approach as extruct).
+    Falls back to OpenGraph and microdata if JSON-LD doesn't have the data.
+
+    Args:
+        extracted_data: Dict from extract_metadata_pure_python()
+
+    Returns:
+        dict: {'price': float, 'currency': str, 'availability': str}
+    """
+    from jsonpath_ng import parse
+
+    result = {}
+
+    # 1. Try JSON-LD first (most reliable and common)
+    for data in extracted_data.get('json-ld', []):
+        try:
+            # Use jsonpath to find price/availability anywhere in the structure
+            price_parse = parse('$..(price|Price)')
+            availability_parse = parse('$..(availability|Availability)')
+            currency_parse = parse('$..(priceCurrency|currency|priceCurrency)')
+
+            price_results = [m.value for m in price_parse.find(data)]
+            if price_results and not result.get('price'):
+                # Handle various price formats
+                price_val = price_results[0]
+                if isinstance(price_val, (int, float)):
+                    result['price'] = float(price_val)
+                elif isinstance(price_val, str):
+                    # Extract numeric value from string
+                    try:
+                        result['price'] = float(re.sub(r'[^\d.]', '', price_val))
+                    except ValueError:
+                        pass
+
+            avail_results = [m.value for m in availability_parse.find(data)]
+            if avail_results and not result.get('availability'):
+                result['availability'] = str(avail_results[0])
+
+            curr_results = [m.value for m in currency_parse.find(data)]
+            if curr_results and not result.get('currency'):
+                result['currency'] = str(curr_results[0])
+
+            # If we found price, this JSON-LD block is good
+            if result.get('price'):
+                logger.debug(f"Pure Python: Found price data in JSON-LD: {result}")
+                break
+
+        except Exception as e:
+            logger.debug(f"Error querying JSON-LD: {e}")
+            continue
+
+    # 2. Try OpenGraph if JSON-LD didn't provide everything
+    og_data = extracted_data.get('opengraph', {})
+    if not result.get('price') and 'og:price:amount' in og_data:
+        try:
+            result['price'] = float(og_data['og:price:amount'])
+        except ValueError:
+            pass
+    if not result.get('currency') and 'og:price:currency' in og_data:
+        result['currency'] = og_data['og:price:currency']
+    if not result.get('availability') and 'og:availability' in og_data:
+        result['availability'] = og_data['og:availability']
+
+    # 3. Use microdata as last resort
+    microdata = extracted_data.get('microdata', {})
+    if not result.get('price') and 'price' in microdata:
+        result['price'] = microdata['price']
+    if not result.get('currency') and 'currency' in microdata:
+        result['currency'] = microdata['currency']
+    if not result.get('availability') and 'availability' in microdata:
+        result['availability'] = microdata['availability']
+
+    # result['price'] could be float or str here, depending on the website, for example it might contain "1,00" commas, etc.
+    # using something like babel you need to know the locale of the website and even then it can be problematic
+    # we dont really do anything with the price data so far.. so just accept it the way it comes.
+    return result

changedetectionio/processors/text_json_diff/__init__.py

@@ -1,6 +1,12 @@
-
 from loguru import logger
 
+# Processor capabilities
+supports_visual_selector = True
+supports_browser_steps = True
+supports_text_filters_and_triggers = True
+supports_text_filters_and_triggers_elements = True
+supports_request_type = True
+
 
 
 def _task(watch, update_handler):
@@ -11,7 +17,8 @@ def _task(watch, update_handler):
 
     try:
         # The slow process (we run 2 of these in parallel)
-        changed_detected, update_obj, text_after_filter = update_handler.run_changedetection(watch=watch)
+        # Always force reprocess for preview - we want to show the filtered content regardless of checksums
+        changed_detected, update_obj, text_after_filter = update_handler.run_changedetection(watch=watch, force_reprocess=True)
     except FilterNotFoundInResponse as e:
         text_after_filter = f"Filter not found in HTML: {str(e)}"
     except ReplyWithContentButNoText as e:
@@ -29,7 +36,7 @@ def _task(watch, update_handler):
 
 
 def prepare_filter_prevew(datastore, watch_uuid, form_data):
-    '''Used by @app.route("/edit/<string:uuid>/preview-rendered", methods=['POST'])'''
+    '''Used by @app.route("/edit/<uuid_str:uuid>/preview-rendered", methods=['POST'])'''
     from changedetectionio import forms, html_tools
     from changedetectionio.model.Watch import model as watch_model
     from concurrent.futures import ThreadPoolExecutor
@@ -49,7 +56,7 @@ def prepare_filter_prevew(datastore, watch_uuid, form_data):
 
     tmp_watch = deepcopy(datastore.data['watching'].get(watch_uuid))
 
-    if tmp_watch and tmp_watch.history and os.path.isdir(tmp_watch.watch_data_dir):
+    if tmp_watch and tmp_watch.history and os.path.isdir(tmp_watch.data_dir):
         # Splice in the temporary stuff from the form
         form = forms.processor_text_json_diff_form(formdata=form_data if request.method == 'POST' else None,
                                                    data=form_data
@@ -58,11 +65,11 @@ def prepare_filter_prevew(datastore, watch_uuid, form_data):
         # Only update vars that came in via the AJAX post
         p = {k: v for k, v in form.data.items() if k in form_data.keys()}
         tmp_watch.update(p)
-        blank_watch_no_filters = watch_model()
+        blank_watch_no_filters = watch_model(datastore_path=datastore.datastore_path, __datastore=datastore.data)
         blank_watch_no_filters['url'] = tmp_watch.get('url')
 
         latest_filename = next(reversed(tmp_watch.history))
-        html_fname = os.path.join(tmp_watch.watch_data_dir, f"{latest_filename}.html.br")
+        html_fname = os.path.join(tmp_watch.data_dir, f"{latest_filename}.html.br")
         with open(html_fname, 'rb') as f:
             decompressed_data = brotli.decompress(f.read()).decode('utf-8') if html_fname.endswith('.br') else f.read().decode('utf-8')
 

changedetectionio/processors/text_json_diff/processor.py

@@ -7,6 +7,7 @@ import re
 import urllib3
 
 from changedetectionio.conditions import execute_ruleset_against_all_plugins
+from changedetectionio.content_fetchers.exceptions import checksumFromPreviousCheckWasTheSame
 from ..base import difference_detection_processor
 from changedetectionio.html_tools import PERL_STYLE_REGEX, cdata_in_document_to_text, TRANSLATE_WHITESPACE_TABLE
 from changedetectionio import html_tools, content_fetchers
@@ -346,6 +347,7 @@ class ContentProcessor:
     def extract_text_from_html(self, html_content, stream_content_type):
         """Convert HTML to plain text."""
         do_anchor = self.datastore.data["settings"]["application"].get("render_anchor_tag_content", False)
+
         return html_tools.html_to_text(
             html_content=html_content,
             render_anchor_tag_content=do_anchor,
@@ -368,12 +370,24 @@ class ChecksumCalculator:
 # (set_proxy_from_list)
 class perform_site_check(difference_detection_processor):
 
-    def run_changedetection(self, watch):
+    def run_changedetection(self, watch, force_reprocess=False):
         changed_detected = False
 
         if not watch:
             raise Exception("Watch no longer exists.")
 
+        current_raw_document_checksum = self.get_raw_document_checksum()
+        # Skip processing only if BOTH conditions are true:
+        # 1. HTML content unchanged (checksum matches last saved checksum)
+        # 2. Watch configuration was not edited (including trigger_text, filters, etc.)
+        # The was_edited flag handles all watch configuration changes, so we don't need
+        # separate checks for trigger_text or other processing rules.
+        if (not force_reprocess and
+            not watch.was_edited and
+            self.last_raw_content_checksum and
+            self.last_raw_content_checksum == current_raw_document_checksum):
+            raise checksumFromPreviousCheckWasTheSame()
+
         # Initialize components
         filter_config = FilterConfig(watch, self.datastore)
         content_processor = ContentProcessor(self.fetcher, watch, filter_config, self.datastore)
@@ -391,9 +405,11 @@ class perform_site_check(difference_detection_processor):
         self.screenshot = self.fetcher.screenshot
         self.xpath_data = self.fetcher.xpath_data
 
-        # Track the content type and checksum before filters
-        update_obj['content_type'] = ctype_header
-        update_obj['previous_md5_before_filters'] = hashlib.md5(self.fetcher.content.encode('utf-8')).hexdigest()
+        # Track the content type (readonly field, doesn't trigger was_edited)
+        update_obj['content-type'] = ctype_header  # Use hyphen (matches OpenAPI spec and watch_base default)
+
+        # Save the raw content checksum to file (processor implementation detail, not watch config)
+        self.update_last_raw_content_checksum(current_raw_document_checksum)
 
         # === CONTENT PREPROCESSING ===
         # Avoid creating unnecessary intermediate string copies by reassigning only when needed

changedetectionio/queue_handlers.py

@@ -5,51 +5,57 @@ import heapq
 import queue
 import threading
 
-try:
-    import janus
-except ImportError:
-    logger.critical(f"CRITICAL: janus library is required. Install with: pip install janus")
-    raise
+# Janus is no longer required - we use pure threading.Queue for multi-loop support
+# try:
+#     import janus
+# except ImportError:
+#     pass  # Not needed anymore
 
 
 class RecheckPriorityQueue:
     """
-    Ultra-reliable priority queue using janus for async/sync bridging.
-    
-    CRITICAL DESIGN NOTE: Both sync_q and async_q are required because:
-    - sync_q: Used by Flask routes, ticker threads, and other synchronous code
-    - async_q: Used by async workers (the actual fetchers/processors) and coroutines
-    
-    DO NOT REMOVE EITHER INTERFACE - they bridge different execution contexts:
-    - Synchronous code (Flask, threads) cannot use async methods without blocking
-    - Async code cannot use sync methods without blocking the event loop
-    - janus provides the only safe bridge between these two worlds
-    
-    Attempting to unify to async-only would require:
-    - Converting all Flask routes to async (major breaking change)
-    - Using asyncio.run() in sync contexts (causes deadlocks)
-    - Thread-pool wrapping (adds complexity and overhead)
-    
-    Minimal implementation focused on reliability:
-    - Pure janus for sync/async bridge
-    - Thread-safe priority ordering  
-    - Bulletproof error handling with critical logging
+    Thread-safe priority queue supporting multiple async event loops.
+
+    ARCHITECTURE:
+    - Multiple async workers, each with its own event loop in its own thread
+    - Hybrid sync/async design for maximum scalability
+    - Sync interface for ticker thread (threading.Queue)
+    - Async interface for workers (asyncio.Event - NO executor threads!)
+
+    SCALABILITY:
+    - Scales to 100-200+ workers without executor thread exhaustion
+    - Async workers wait on asyncio.Event (pure coroutines, no threads)
+    - Sync callers use threading.Queue (backward compatible)
+
+    WHY NOT JANUS:
+    - Janus binds to ONE event loop at creation time
+    - Our architecture has 15+ workers, each with separate event loops
+    - Workers in different threads/loops cannot share janus async interface
+
+    WHY NOT RUN_IN_EXECUTOR:
+    - With 200 workers, run_in_executor() would block 200 threads
+    - Exhausts ThreadPoolExecutor, starves Flask HTTP handlers
+    - Pure async approach uses 0 threads while waiting
     """
-    
+
     def __init__(self, maxsize: int = 0):
         try:
-            self._janus_queue = janus.Queue(maxsize=maxsize)
-            # BOTH interfaces required - see class docstring for why
-            self.sync_q = self._janus_queue.sync_q   # Flask routes, ticker thread
-            self.async_q = self._janus_queue.async_q # Async workers
-            
+            import asyncio
+
+            # Sync interface: threading.Queue for ticker thread and Flask routes
+            self._notification_queue = queue.Queue(maxsize=maxsize if maxsize > 0 else 0)
+
             # Priority storage - thread-safe
             self._priority_items = []
             self._lock = threading.RLock()
-            
+
+            # No event signaling needed - pure polling approach
+            # Workers check queue every 50ms (latency acceptable: 0-500ms)
+            # Scales to 1000+ workers: each sleeping worker = ~4KB coroutine, not thread
+
             # Signals for UI updates
             self.queue_length_signal = signal('queue_length')
-            
+
             logger.debug("RecheckPriorityQueue initialized successfully")
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to initialize RecheckPriorityQueue: {str(e)}")
@@ -58,37 +64,48 @@ class RecheckPriorityQueue:
     # SYNC INTERFACE (for ticker thread)
     def put(self, item, block: bool = True, timeout: Optional[float] = None):
         """Thread-safe sync put with priority ordering"""
+        logger.trace(f"RecheckQueue.put() called for item: {self._get_item_uuid(item)}, block={block}, timeout={timeout}")
         try:
-            # Add to priority storage
+            # CRITICAL: Add to both priority storage AND notification queue atomically
+            # to prevent desynchronization where item exists but no notification
             with self._lock:
                 heapq.heappush(self._priority_items, item)
-            
-            # Notify via janus sync queue
-            self.sync_q.put(True, block=block, timeout=timeout)
-            
-            # Emit signals
-            self._emit_put_signals(item)
-            
-            logger.debug(f"Successfully queued item: {self._get_item_uuid(item)}")
-            return True
-            
-        except Exception as e:
-            logger.critical(f"CRITICAL: Failed to put item {self._get_item_uuid(item)}: {str(e)}")
-            # Remove from priority storage if janus put failed
+
+                # Add notification - use blocking with timeout for safety
+                # Notification queue is unlimited size, so should never block in practice
+                # but timeout ensures we detect any unexpected issues (deadlock, etc)
+                try:
+                    self._notification_queue.put(True, block=True, timeout=5.0)
+                except Exception as notif_e:
+                    # Notification failed - MUST remove from priority_items to keep in sync
+                    # This prevents "Priority queue inconsistency" errors in get()
+                    logger.critical(f"CRITICAL: Notification queue put failed, removing from priority_items: {notif_e}")
+                    self._priority_items.remove(item)
+                    heapq.heapify(self._priority_items)
+                    raise  # Re-raise to be caught by outer exception handler
+
+            # Signal emission after successful queue - log but don't fail the operation
+            # Item is already safely queued, so signal failure shouldn't affect queue state
             try:
-                with self._lock:
-                    if item in self._priority_items:
-                        self._priority_items.remove(item)
-                        heapq.heapify(self._priority_items)
-            except Exception as cleanup_e:
-                logger.critical(f"CRITICAL: Failed to cleanup after put failure: {str(e)}")
+                self._emit_put_signals(item)
+            except Exception as signal_e:
+                logger.error(f"Failed to emit put signals but item queued successfully: {signal_e}")
+
+            logger.trace(f"Successfully queued item: {self._get_item_uuid(item)}")
+            return True
+
+        except Exception as e:
+            logger.critical(f"CRITICAL: Failed to put item {self._get_item_uuid(item)}: {type(e).__name__}: {str(e)}")
+            # Item should have been cleaned up in the inner try/except if notification failed
             return False
     
     def get(self, block: bool = True, timeout: Optional[float] = None):
         """Thread-safe sync get with priority ordering"""
+        logger.trace(f"RecheckQueue.get() called, block={block}, timeout={timeout}")
+        import queue as queue_module
         try:
-            # Wait for notification
-            self.sync_q.get(block=block, timeout=timeout)
+            # Wait for notification (this doesn't return the actual item, just signals availability)
+            self._notification_queue.get(block=block, timeout=timeout)
 
             # Get highest priority item
             with self._lock:
@@ -97,66 +114,91 @@ class RecheckPriorityQueue:
                     raise Exception("Priority queue inconsistency")
                 item = heapq.heappop(self._priority_items)
 
-            # Emit signals
-            self._emit_get_signals()
+            # Signal emission after successful retrieval - log but don't lose the item
+            # Item is already retrieved, so signal failure shouldn't affect queue state
+            try:
+                self._emit_get_signals()
+            except Exception as signal_e:
+                logger.error(f"Failed to emit get signals but item retrieved successfully: {signal_e}")
 
-            logger.debug(f"Successfully retrieved item: {self._get_item_uuid(item)}")
+            logger.trace(f"RecheckQueue.get() successfully retrieved item: {self._get_item_uuid(item)}")
             return item
 
+        except queue_module.Empty:
+            # Queue is empty with timeout - expected behavior
+            logger.trace(f"RecheckQueue.get() timed out - queue is empty (timeout={timeout})")
+            raise  # noqa
         except Exception as e:
-            logger.critical(f"CRITICAL: Failed to get item from queue: {str(e)}")
+            # Re-raise without logging - caller (worker) will handle and log appropriately
+            logger.trace(f"RecheckQueue.get() failed with exception: {type(e).__name__}: {str(e)}")
             raise
     
     # ASYNC INTERFACE (for workers)
-    async def async_put(self, item):
-        """Pure async put with priority ordering"""
+    async def async_put(self, item, executor=None):
+        """Async put with priority ordering - uses thread pool to avoid blocking
+
+        Args:
+            item: Item to add to queue
+            executor: Optional ThreadPoolExecutor. If None, uses default pool.
+        """
+        logger.trace(f"RecheckQueue.async_put() called for item: {self._get_item_uuid(item)}, executor={executor}")
+        import asyncio
         try:
-            # Add to priority storage
-            with self._lock:
-                heapq.heappush(self._priority_items, item)
-            
-            # Notify via janus async queue
-            await self.async_q.put(True)
-            
-            # Emit signals
-            self._emit_put_signals(item)
-            
-            logger.debug(f"Successfully async queued item: {self._get_item_uuid(item)}")
-            return True
-            
+            # Use run_in_executor to call sync put without blocking event loop
+            loop = asyncio.get_event_loop()
+            result = await loop.run_in_executor(
+                executor,  # Use provided executor or default
+                lambda: self.put(item, block=True, timeout=5.0)
+            )
+
+            logger.trace(f"RecheckQueue.async_put() successfully queued item: {self._get_item_uuid(item)}")
+            return result
+
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to async put item {self._get_item_uuid(item)}: {str(e)}")
-            # Remove from priority storage if janus put failed
-            try:
-                with self._lock:
-                    if item in self._priority_items:
-                        self._priority_items.remove(item)
-                        heapq.heapify(self._priority_items)
-            except Exception as cleanup_e:
-                logger.critical(f"CRITICAL: Failed to cleanup after async put failure: {str(e)}")
             return False
-    
-    async def async_get(self):
-        """Pure async get with priority ordering"""
+
+    async def async_get(self, executor=None, timeout=1.0):
+        """
+        Efficient async get using executor for blocking call.
+
+        HYBRID APPROACH: Best of both worlds
+        - Uses run_in_executor for efficient blocking (no polling overhead)
+        - Single timeout (no double-timeout race condition)
+        - Scales well: executor sized to match worker count
+
+        With FETCH_WORKERS=10: 10 threads blocked max (acceptable)
+        With FETCH_WORKERS=200: Need executor with 200+ threads (see worker_pool.py)
+
+        Args:
+            executor: ThreadPoolExecutor (sized to match worker count)
+            timeout: Maximum time to wait in seconds
+
+        Returns:
+            Item from queue
+
+        Raises:
+            queue.Empty: If timeout expires with no item available
+        """
+        logger.trace(f"RecheckQueue.async_get() called, timeout={timeout}")
+        import asyncio
         try:
-            # Wait for notification
-            await self.async_q.get()
+            # Use run_in_executor to call sync get efficiently
+            # No outer asyncio.wait_for wrapper = no double timeout issue!
+            loop = asyncio.get_event_loop()
+            item = await loop.run_in_executor(
+                executor,
+                lambda: self.get(block=True, timeout=timeout)
+            )
 
-            # Get highest priority item
-            with self._lock:
-                if not self._priority_items:
-                    logger.critical(f"CRITICAL: Async queue notification received but no priority items available")
-                    raise Exception("Priority queue inconsistency")
-                item = heapq.heappop(self._priority_items)
-
-            # Emit signals
-            self._emit_get_signals()
-
-            logger.debug(f"Successfully async retrieved item: {self._get_item_uuid(item)}")
+            logger.trace(f"RecheckQueue.async_get() successfully retrieved item: {self._get_item_uuid(item)}")
             return item
 
+        except queue.Empty:
+            logger.trace(f"RecheckQueue.async_get() timed out - queue is empty")
+            raise
         except Exception as e:
-            logger.critical(f"CRITICAL: Failed to async get item from queue: {str(e)}")
+            logger.critical(f"CRITICAL: Failed to async get item from queue: {type(e).__name__}: {str(e)}")
             raise
     
     # UTILITY METHODS
@@ -172,11 +214,45 @@ class RecheckPriorityQueue:
     def empty(self) -> bool:
         """Check if queue is empty"""
         return self.qsize() == 0
-    
-    def close(self):
-        """Close the janus queue"""
+
+    def get_queued_uuids(self) -> list:
+        """Get list of all queued UUIDs efficiently with single lock"""
         try:
-            self._janus_queue.close()
+            with self._lock:
+                return [item.item['uuid'] for item in self._priority_items if hasattr(item, 'item') and 'uuid' in item.item]
+        except Exception as e:
+            logger.critical(f"CRITICAL: Failed to get queued UUIDs: {str(e)}")
+            return []
+
+    def clear(self):
+        """Clear all items from both priority storage and notification queue"""
+        try:
+            with self._lock:
+                # Clear priority items
+                self._priority_items.clear()
+
+                # Drain all notifications to prevent stale notifications
+                # This is critical for test cleanup to prevent queue desynchronization
+                drained = 0
+                while not self._notification_queue.empty():
+                    try:
+                        self._notification_queue.get_nowait()
+                        drained += 1
+                    except queue.Empty:
+                        break
+
+                if drained > 0:
+                    logger.debug(f"Cleared queue: removed {drained} notifications")
+
+            return True
+        except Exception as e:
+            logger.critical(f"CRITICAL: Failed to clear queue: {str(e)}")
+            return False
+
+    def close(self):
+        """Close the queue"""
+        try:
+            # Nothing to close for threading.Queue
             logger.debug("RecheckPriorityQueue closed successfully")
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to close RecheckPriorityQueue: {str(e)}")
@@ -308,7 +384,7 @@ class RecheckPriorityQueue:
         except Exception:
             pass
         return 'unknown'
-    
+
     def _emit_put_signals(self, item):
         """Emit signals when item is added"""
         try:
@@ -317,14 +393,14 @@ class RecheckPriorityQueue:
                 watch_check_update = signal('watch_check_update')
                 if watch_check_update:
                     watch_check_update.send(watch_uuid=item.item['uuid'])
-            
+
             # Queue length signal
             if self.queue_length_signal:
                 self.queue_length_signal.send(length=self.qsize())
-                
+
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to emit put signals: {str(e)}")
-    
+
     def _emit_get_signals(self):
         """Emit signals when item is removed"""
         try:
@@ -348,76 +424,115 @@ class NotificationQueue:
     Simple wrapper around janus with bulletproof error handling.
     """
     
-    def __init__(self, maxsize: int = 0):
+    def __init__(self, maxsize: int = 0, datastore=None):
         try:
-            self._janus_queue = janus.Queue(maxsize=maxsize)
-            # BOTH interfaces required - see class docstring for why
-            self.sync_q = self._janus_queue.sync_q   # Flask routes, threads
-            self.async_q = self._janus_queue.async_q # Async workers
+            # Use pure threading.Queue to avoid event loop binding issues
+            self._notification_queue = queue.Queue(maxsize=maxsize if maxsize > 0 else 0)
             self.notification_event_signal = signal('notification_event')
+            self.datastore = datastore  # For checking all_muted setting
+            self._lock = threading.RLock()
             logger.debug("NotificationQueue initialized successfully")
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to initialize NotificationQueue: {str(e)}")
             raise
+
+    def set_datastore(self, datastore):
+        """Set datastore reference after initialization (for circular dependency handling)"""
+        self.datastore = datastore
     
     def put(self, item: Dict[str, Any], block: bool = True, timeout: Optional[float] = None):
         """Thread-safe sync put with signal emission"""
+        logger.trace(f"NotificationQueue.put() called for item: {item.get('uuid', 'unknown')}, block={block}, timeout={timeout}")
         try:
-            self.sync_q.put(item, block=block, timeout=timeout)
+            # Check if all notifications are muted
+            if self.datastore and self.datastore.data['settings']['application'].get('all_muted', False):
+                logger.debug(f"Notification blocked - all notifications are muted: {item.get('uuid', 'unknown')}")
+                return False
+
+            with self._lock:
+                self._notification_queue.put(item, block=block, timeout=timeout)
             self._emit_notification_signal(item)
-            logger.debug(f"Successfully queued notification: {item.get('uuid', 'unknown')}")
+            logger.trace(f"NotificationQueue.put() successfully queued notification: {item.get('uuid', 'unknown')}")
             return True
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to put notification {item.get('uuid', 'unknown')}: {str(e)}")
             return False
     
-    async def async_put(self, item: Dict[str, Any]):
-        """Pure async put with signal emission"""
+    async def async_put(self, item: Dict[str, Any], executor=None):
+        """Async put with signal emission - uses thread pool
+
+        Args:
+            item: Notification item to queue
+            executor: Optional ThreadPoolExecutor
+        """
+        logger.trace(f"NotificationQueue.async_put() called for item: {item.get('uuid', 'unknown')}, executor={executor}")
+        import asyncio
         try:
-            await self.async_q.put(item)
-            self._emit_notification_signal(item)
-            logger.debug(f"Successfully async queued notification: {item.get('uuid', 'unknown')}")
+            # Check if all notifications are muted
+            if self.datastore and self.datastore.data['settings']['application'].get('all_muted', False):
+                logger.debug(f"Notification blocked - all notifications are muted: {item.get('uuid', 'unknown')}")
+                return False
+
+            loop = asyncio.get_event_loop()
+            await loop.run_in_executor(executor, lambda: self.put(item, block=True, timeout=5.0))
+            logger.trace(f"NotificationQueue.async_put() successfully queued notification: {item.get('uuid', 'unknown')}")
             return True
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to async put notification {item.get('uuid', 'unknown')}: {str(e)}")
             return False
-    
+
     def get(self, block: bool = True, timeout: Optional[float] = None):
         """Thread-safe sync get"""
+        logger.trace(f"NotificationQueue.get() called, block={block}, timeout={timeout}")
         try:
-            return self.sync_q.get(block=block, timeout=timeout)
+            with self._lock:
+                item = self._notification_queue.get(block=block, timeout=timeout)
+            logger.trace(f"NotificationQueue.get() retrieved item: {item.get('uuid', 'unknown') if isinstance(item, dict) else 'unknown'}")
+            return item
         except queue.Empty as e:
+            logger.trace(f"NotificationQueue.get() timed out - queue is empty (timeout={timeout})")
             raise e
         except Exception as e:
-            logger.critical(f"CRITICAL: Failed to get notification: {str(e)}")
+            logger.critical(f"CRITICAL: Failed to get notification: {type(e).__name__}: {str(e)}")
             raise e
-    
-    async def async_get(self):
-        """Pure async get"""
+
+    async def async_get(self, executor=None):
+        """Async get - uses thread pool
+
+        Args:
+            executor: Optional ThreadPoolExecutor
+        """
+        logger.trace(f"NotificationQueue.async_get() called, executor={executor}")
+        import asyncio
         try:
-            return await self.async_q.get()
+            loop = asyncio.get_event_loop()
+            item = await loop.run_in_executor(executor, lambda: self.get(block=True, timeout=1.0))
+            logger.trace(f"NotificationQueue.async_get() retrieved item: {item.get('uuid', 'unknown') if isinstance(item, dict) else 'unknown'}")
+            return item
         except queue.Empty as e:
+            logger.trace(f"NotificationQueue.async_get() timed out - queue is empty")
             raise e
         except Exception as e:
-            logger.critical(f"CRITICAL: Failed to async get notification: {str(e)}")
+            logger.critical(f"CRITICAL: Failed to async get notification: {type(e).__name__}: {str(e)}")
             raise e
     
     def qsize(self) -> int:
         """Get current queue size"""
         try:
-            return self.sync_q.qsize()
+            with self._lock:
+                return self._notification_queue.qsize()
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to get notification queue size: {str(e)}")
             return 0
-    
+
     def empty(self) -> bool:
         """Check if queue is empty"""
         return self.qsize() == 0
-    
+
     def close(self):
-        """Close the janus queue"""
+        """Close the queue"""
         try:
-            self._janus_queue.close()
+            # Nothing to close for threading.Queue
             logger.debug("NotificationQueue closed successfully")
         except Exception as e:
             logger.critical(f"CRITICAL: Failed to close NotificationQueue: {str(e)}")

changedetectionio/realtime/events.py

@@ -37,9 +37,9 @@ def register_watch_operation_handlers(socketio, datastore):
                 # Import here to avoid circular imports
                 from changedetectionio.flask_app import update_q
                 from changedetectionio import queuedWatchMetaData
-                from changedetectionio import worker_handler
+                from changedetectionio import worker_pool
                 
-                worker_handler.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
+                worker_pool.queue_item_async_safe(update_q, queuedWatchMetaData.PrioritizedItem(priority=1, item={'uuid': uuid}))
                 logger.info(f"Socket.IO: Queued recheck for watch {uuid}")
             else:
                 emit('operation_result', {'success': False, 'error': f'Unknown operation: {op}'})

changedetectionio/realtime/socket_server.py

@@ -145,16 +145,13 @@ def handle_watch_update(socketio, **kwargs):
         # Emit the watch update to all connected clients
         from changedetectionio.flask_app import update_q
         from changedetectionio.flask_app import _jinja2_filter_datetime
-        from changedetectionio import worker_handler
+        from changedetectionio import worker_pool
 
         # Get list of watches that are currently running
-        running_uuids = worker_handler.get_running_uuids()
+        running_uuids = worker_pool.get_running_uuids()
 
-        # Get list of watches in the queue
-        queue_list = []
-        for q_item in update_q.queue:
-            if hasattr(q_item, 'item') and 'uuid' in q_item.item:
-                queue_list.append(q_item.item['uuid'])
+        # Get list of watches in the queue (efficient single-lock method)
+        queue_list = update_q.get_queued_uuids()
 
         # Get the error texts from the watch
         error_texts = watch.compile_error_texts()
@@ -243,7 +240,10 @@ def init_socketio(app, datastore):
                         async_mode=async_mode,
                         cors_allowed_origins=cors_origins,  # None means same-origin only
                         logger=strtobool(os.getenv('SOCKETIO_LOGGING', 'False')),
-                        engineio_logger=strtobool(os.getenv('SOCKETIO_LOGGING', 'False')))
+                        engineio_logger=strtobool(os.getenv('SOCKETIO_LOGGING', 'False')),
+                        # Disable WebSocket compression to prevent memory accumulation
+                        # Flask-Compress already handles HTTP response compression
+                        engineio_options={'http_compression': False, 'compression_threshold': 0})
 
     # Set up event handlers
     logger.info("Socket.IO: Registering connect event handler")
@@ -252,23 +252,34 @@ def init_socketio(app, datastore):
     def event_checkbox_operations(data):
         from changedetectionio.blueprint.ui import _handle_operations
         from changedetectionio import queuedWatchMetaData
-        from changedetectionio import worker_handler
+        from changedetectionio import worker_pool
         from changedetectionio.flask_app import update_q, watch_check_update
+        import threading
+
         logger.trace(f"Got checkbox operations event: {data}")
 
         datastore = socketio.datastore
 
-        _handle_operations(
-            op=data.get('op'),
-            uuids=data.get('uuids'),
-            datastore=datastore,
-            extra_data=data.get('extra_data'),
-            worker_handler=worker_handler,
-            update_q=update_q,
-            queuedWatchMetaData=queuedWatchMetaData,
-            watch_check_update=watch_check_update,
-            emit_flash=False
-        )
+        def run_operation():
+            """Run the operation in a background thread to avoid blocking the socket.io event loop"""
+            try:
+                _handle_operations(
+                    op=data.get('op'),
+                    uuids=data.get('uuids'),
+                    datastore=datastore,
+                    extra_data=data.get('extra_data'),
+                    worker_pool=worker_pool,
+                    update_q=update_q,
+                    queuedWatchMetaData=queuedWatchMetaData,
+                    watch_check_update=watch_check_update,
+                    emit_flash=False
+                )
+            except Exception as e:
+                logger.error(f"Error in checkbox operation thread: {e}")
+
+        # Start operation in a disposable daemon thread
+        thread = threading.Thread(target=run_operation, daemon=True, name=f"checkbox-op-{data.get('op')}")
+        thread.start()
 
     @socketio.on('connect')
     def handle_connect():
@@ -334,4 +345,4 @@ def init_socketio(app, datastore):
 
     logger.info("Socket.IO initialized and attached to main Flask app")
     logger.info(f"Socket.IO: Registered event handlers: {socketio.handlers if hasattr(socketio, 'handlers') else 'No handlers found'}")
-    return socketio
+    return socketio

changedetectionio/run_basic_tests.sh

@@ -10,6 +10,7 @@
 set -e
 
 SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
+rm tests/logs/* -f
 
 # Since theres no curl installed lets roll with python3
 check_sanity() {
@@ -43,12 +44,12 @@ data_sanity_test () {
   cd ..
   TMPDIR=$(mktemp -d)
   PORT_N=$((5000 + RANDOM % (6501 - 5000)))
-  ./changedetection.py -p $PORT_N -d $TMPDIR -u "https://localhost?test-url-is-sanity=1" &
+  ALLOW_IANA_RESTRICTED_ADDRESSES=true ./changedetection.py -p $PORT_N -d $TMPDIR -u "https://localhost?test-url-is-sanity=1" &
   PID=$!
   sleep 5
   kill $PID
   sleep 2
-  ./changedetection.py -p $PORT_N -d $TMPDIR &
+  ALLOW_IANA_RESTRICTED_ADDRESSES=true ./changedetection.py -p $PORT_N -d $TMPDIR &
   PID=$!
   sleep 5
   # On a restart the URL should still be there
@@ -64,45 +65,43 @@ data_sanity_test
 echo "-------------------- Running rest of tests in parallel -------------------------------"
 
 # REMOVE_REQUESTS_OLD_SCREENSHOTS disabled so that we can write a screenshot and send it in test_notifications.py without a real browser
-REMOVE_REQUESTS_OLD_SCREENSHOTS=false \
+FETCH_WORKERS=2 REMOVE_REQUESTS_OLD_SCREENSHOTS=false \
 pytest tests/test_*.py \
-  -n 30 \
+  -n 8 \
   --dist=load \
   -vvv \
   -s \
   --capture=no \
+  -k "not test_queue_system" \
   --log-cli-level=DEBUG \
   --log-cli-format="%(asctime)s [%(process)d] [%(levelname)s] %(name)s: %(message)s"
 
 echo "---------------------------- DONE parallel test ---------------------------------------"
 
+FETCH_WORKERS=20 pytest -vvv -s tests/test_queue_handler.py
+
 echo "RUNNING WITH BASE_URL SET"
 
 # Now re-run some tests with BASE_URL enabled
 # Re #65 - Ability to include a link back to the installation, in the notification.
 export BASE_URL="https://really-unique-domain.io"
 
-REMOVE_REQUESTS_OLD_SCREENSHOTS=false pytest -vv -s --maxfail=1 tests/test_notification.py
-
-
 # Re-run with HIDE_REFERER set - could affect login
 export HIDE_REFERER=True
-pytest -vv -s --maxfail=1 tests/test_access_control.py
+REMOVE_REQUESTS_OLD_SCREENSHOTS=false pytest -vv -s --maxfail=1 tests/test_notification.py tests/test_access_control.py
+
 
 # Re-run a few tests that will trigger brotli based storage
-export SNAPSHOT_BROTLI_COMPRESSION_THRESHOLD=5
-pytest -vv -s --maxfail=1 tests/test_access_control.py
-REMOVE_REQUESTS_OLD_SCREENSHOTS=false pytest tests/test_notification.py
-pytest -vv -s --maxfail=1 tests/test_backend.py
-pytest -vv -s --maxfail=1 tests/test_rss.py
-pytest -vv -s --maxfail=1 tests/test_unique_lines.py
+# And again with brotli+screenshot attachment
+SNAPSHOT_BROTLI_COMPRESSION_THRESHOLD=5 REMOVE_REQUESTS_OLD_SCREENSHOTS=false pytest -vv -s --maxfail=1 --dist=load tests/test_backend.py tests/test_rss.py tests/test_unique_lines.py tests/test_notification.py  tests/test_access_control.py
 
-# Try high concurrency
-FETCH_WORKERS=130 pytest  tests/test_history_consistency.py -v -l
+# Try high concurrency with aggressive worker restarts
+FETCH_WORKERS=50 WORKER_MAX_RUNTIME=2 WORKER_MAX_JOBS=1 pytest  tests/test_history_consistency.py -vv -l -s
 
 # Check file:// will pickup a file when enabled
 echo "Hello world" > /tmp/test-file.txt
 ALLOW_FILE_URI=yes pytest -vv -s  tests/test_security.py
 
 
-
+# Run it again so that brotli kicks in
+TEST_WITH_BROTLI=1 SNAPSHOT_BROTLI_COMPRESSION_THRESHOLD=100 FETCH_WORKERS=20 pytest tests/test_history_consistency.py -vv -l -s

changedetectionio/static/favicons/site.webmanifest

@@ -1,19 +1,25 @@
 {
-    "name": "",
-    "short_name": "",
+    "name": "ChangeDetection.io",
+    "short_name": "ChangeDetect",
+    "description": "Self-hosted website change detection and monitoring",
     "icons": [
         {
             "src": "android-chrome-192x192.png",
             "sizes": "192x192",
-            "type": "image/png"
+            "type": "image/png",
+            "purpose": "any maskable"
         },
         {
             "src": "android-chrome-256x256.png",
             "sizes": "256x256",
-            "type": "image/png"
+            "type": "image/png",
+            "purpose": "any maskable"
         }
     ],
-    "theme_color": "#ffffff",
+    "start_url": "/",
+    "theme_color": "#5bbad5",
     "background_color": "#ffffff",
-    "display": "standalone"
+    "display": "standalone",
+    "categories": ["utilities", "productivity"],
+    "orientation": "any"
 }

changedetectionio/static/js/browser-steps.js

@@ -17,8 +17,6 @@ $(document).ready(function () {
         set_scale();
     });
     // Should always be disabled
-    $('#browser_steps-0-operation option[value="Goto site"]').prop("selected", "selected");
-    $('#browser_steps-0-operation').attr('disabled', 'disabled');
 
     $('#browsersteps-click-start').click(function () {
         $("#browsersteps-click-start").fadeOut();
@@ -45,12 +43,6 @@ $(document).ready(function () {
         browsersteps_session_id = false;
         apply_buttons_disabled = false;
         ctx.clearRect(0, 0, c.width, c.height);
-        set_first_gotosite_disabled();
-    }
-
-    function set_first_gotosite_disabled() {
-        $('#browser_steps >li:first-child select').val('Goto site').attr('disabled', 'disabled');
-        $('#browser_steps >li:first-child').css('opacity', '0.5');
     }
 
     // Show seconds remaining until the browser interface needs to restart the session
@@ -243,14 +235,54 @@ $(document).ready(function () {
         ctx.fill();
     }
 
+    // Reusable AJAX function for browser step operations
+    function executeBrowserStep(url, data = {}) {
+        $('#browser-steps-ui .loader .spinner').fadeIn();
+        apply_buttons_disabled = true;
+        $('ul#browser_steps li .control .apply').css('opacity', 0.5);
+        $("#browsersteps-img").css('opacity', 0.65);
+
+        return $.ajax({
+            method: "POST",
+            url: url,
+            data: data,
+            statusCode: {
+                400: function () {
+                    alert("There was a problem processing the request, please reload the page.");
+                    $("#loading-status-text").hide();
+                    $('#browser-steps-ui .loader .spinner').fadeOut();
+                },
+                401: function (data) {
+                    alert(data.responseText);
+                    $("#loading-status-text").hide();
+                    $('#browser-steps-ui .loader .spinner').fadeOut();
+                }
+            }
+        }).done(function (data) {
+            xpath_data = data.xpath_data;
+            $('#browsersteps-img').attr('src', data.screenshot);
+            $('#browser-steps-ui .loader .spinner').fadeOut();
+            apply_buttons_disabled = false;
+            $("#browsersteps-img").css('opacity', 1);
+            $('ul#browser_steps li .control .apply').css('opacity', 1);
+            $("#loading-status-text").hide();
+        }).fail(function (data) {
+            console.log(data);
+            if (data.responseText && data.responseText.includes("Browser session expired")) {
+                disable_browsersteps_ui();
+            }
+            apply_buttons_disabled = false;
+            $("#loading-status-text").hide();
+            $('ul#browser_steps li .control .apply').css('opacity', 1);
+            $("#browsersteps-img").css('opacity', 1);
+        });
+    }
+
     function start() {
         console.log("Starting browser-steps UI");
         browsersteps_session_id = false;
-        // @todo This setting of the first one should be done at the datalayer but wtforms doesnt wanna play nice
-        $('#browser_steps >li:first-child').removeClass('empty');
-        set_first_gotosite_disabled();
         $('#browser-steps-ui .loader .spinner').show();
-        $('.clear,.remove', $('#browser_steps >li:first-child')).hide();
+        // Request a new session
         $.ajax({
             type: "GET",
             url: browser_steps_start_url,
@@ -267,11 +299,12 @@ $(document).ready(function () {
         }).done(function (data) {
             $("#loading-status-text").fadeIn();
             browsersteps_session_id = data.browsersteps_session_id;
-            // This should trigger 'Goto site'
-            console.log("Got startup response, requesting Goto-Site (first) step fake click");
-            $('#browser_steps >li:first-child .apply').click();
             browser_interface_seconds_remaining = 500;
-            set_first_gotosite_disabled();
+            // Request goto_site operation
+            executeBrowserStep(
+                browser_steps_sync_url + "&browsersteps_session_id=" + browsersteps_session_id + "&goto_website_url_first_step=true"
+            );
+
         }).fail(function (data) {
             console.log(data);
             alert('There was an error communicating with the server.');
@@ -280,7 +313,6 @@ $(document).ready(function () {
     }
 
     function disable_browsersteps_ui() {
-        set_first_gotosite_disabled();
         $("#browser-steps-ui").css('opacity', '0.3');
         $('#browsersteps-selector-canvas').off("mousemove mousedown click");
     }
@@ -328,16 +360,13 @@ $(document).ready(function () {
     // Add the extra buttons to the steps
     $('ul#browser_steps li').each(function (i) {
             var s = '<div class="control">' + '<a data-step-index=' + i + ' class="pure-button button-secondary button-green button-xsmall apply" >Apply</a>&nbsp;';
-            if (i > 0) {
-                // The first step never gets these (Goto-site)
-                s += `<a data-step-index="${i}" class="pure-button button-secondary button-xsmall clear" >Clear</a>&nbsp;` +
-                    `<a data-step-index="${i}" class="pure-button button-secondary button-red button-xsmall remove" >Remove</a>`;
+            s += `<a data-step-index="${i}" class="pure-button button-secondary button-xsmall clear" >Clear</a>&nbsp;` +
+                `<a data-step-index="${i}" class="pure-button button-secondary button-red button-xsmall remove" >Remove</a>`;
 
-                // if a screenshot is available
-                if (browser_steps_available_screenshots.includes(i.toString())) {
-                    var d = (browser_steps_last_error_step === i+1) ? 'before' : 'after';
-                    s += `&nbsp;<a data-step-index="${i}" class="pure-button button-secondary button-xsmall show-screenshot" title="Show screenshot from last run" data-type="${d}">Pic</a>&nbsp;`;
-                }
+            // if a screenshot is available
+            if (browser_steps_available_screenshots.includes(i.toString())) {
+                var d = (browser_steps_last_error_step === i+1) ? 'before' : 'after';
+                s += `&nbsp;<a data-step-index="${i}" class="pure-button button-secondary button-xsmall show-screenshot" title="Show screenshot from last run" data-type="${d}">Pic</a>&nbsp;`;
             }
             s += '</div>';
             $(this).append(s)
@@ -376,80 +405,35 @@ $(document).ready(function () {
     });
 
     $('ul#browser_steps li .control .apply').click(function (event) {
-        // sequential requests @todo refactor
         if (apply_buttons_disabled) {
             return;
         }
 
         var current_data = $(event.currentTarget).closest('li');
-        $('#browser-steps-ui .loader .spinner').fadeIn();
-        apply_buttons_disabled = true;
-        $('ul#browser_steps li .control .apply').css('opacity', 0.5);
-        $("#browsersteps-img").css('opacity', 0.65);
-
-        var is_last_step = 0;
         var step_n = $(event.currentTarget).data('step-index');
 
-        // On the last step, we should also be getting data ready for the visual selector
+        // Determine if this is the last configured step
+        var is_last_step = 0;
         $('ul#browser_steps li select').each(function (i) {
             if ($(this).val() !== 'Choose one') {
                 is_last_step += 1;
             }
         });
-
-        if (is_last_step == (step_n + 1)) {
-            is_last_step = true;
-        } else {
-            is_last_step = false;
-        }
+        is_last_step = (is_last_step == (step_n + 1));
 
         console.log("Requesting step via POST " + $("select[id$='operation']", current_data).first().val());
-        // POST the currently clicked step form widget back and await response, redraw
-        $.ajax({
-            method: "POST",
-            url: browser_steps_sync_url + "&browsersteps_session_id=" + browsersteps_session_id,
-            data: {
+
+        // Execute the browser step
+        executeBrowserStep(
+            browser_steps_sync_url + "&browsersteps_session_id=" + browsersteps_session_id,
+            {
                 'operation': $("select[id$='operation']", current_data).first().val(),
                 'selector': $("input[id$='selector']", current_data).first().val(),
                 'optional_value': $("input[id$='optional_value']", current_data).first().val(),
                 'step_n': step_n,
                 'is_last_step': is_last_step
-            },
-            statusCode: {
-                400: function () {
-                    // More than likely the CSRF token was lost when the server restarted
-                    alert("There was a problem processing the request, please reload the page.");
-                    $("#loading-status-text").hide();
-                    $('#browser-steps-ui .loader .spinner').fadeOut();
-                },
-                401: function (data) {
-                    // More than likely the CSRF token was lost when the server restarted
-                    alert(data.responseText);
-                    $("#loading-status-text").hide();
-                    $('#browser-steps-ui .loader .spinner').fadeOut();
-                }
             }
-        }).done(function (data) {
-            // it should return the new state (selectors available and screenshot)
-            xpath_data = data.xpath_data;
-            $('#browsersteps-img').attr('src', data.screenshot);
-            $('#browser-steps-ui .loader .spinner').fadeOut();
-            apply_buttons_disabled = false;
-            $("#browsersteps-img").css('opacity', 1);
-            $('ul#browser_steps li .control .apply').css('opacity', 1);
-            $("#loading-status-text").hide();
-            set_first_gotosite_disabled();
-        }).fail(function (data) {
-            console.log(data);
-            if (data.responseText.includes("Browser session expired")) {
-                disable_browsersteps_ui();
-            }
-            apply_buttons_disabled = false;
-            $("#loading-status-text").hide();
-            $('ul#browser_steps li .control .apply').css('opacity', 1);
-            $("#browsersteps-img").css('opacity', 1);
-        });
-
+        );
     });
 
     $('ul#browser_steps li .control .show-screenshot').click(function (element) {

changedetectionio/static/js/modal.js

@@ -184,7 +184,8 @@ $(document).ready(function() {
         }
         // If it's a button in a form, submit the form
         else if ($element.is('button')) {
-          $element.closest('form').submit();
+          // Use requestSubmit() to include the button's name/value in the form data
+          $element.closest('form')[0].requestSubmit($element[0]);
         }
       }
     };

changedetectionio/static/js/realtime.js

@@ -76,7 +76,7 @@ $(document).ready(function () {
 
     // Cache DOM elements for performance
     const queueBubble = document.getElementById('queue-bubble');
-
+    const queueSizePagerInfoText = document.getElementById('queue-size-int');
     // Only try to connect if authentication isn't required or user is authenticated
     // The 'is_authenticated' variable will be set in the template
     if (typeof is_authenticated !== 'undefined' ? is_authenticated : true) {
@@ -118,6 +118,10 @@ $(document).ready(function () {
 
             socket.on('queue_size', function (data) {
                 console.log(`${data.event_timestamp} - Queue size update: ${data.q_length}`);
+                if(queueSizePagerInfoText) {
+                    queueSizePagerInfoText.textContent = parseInt(data.q_length).toLocaleString() || 'None';
+                }
+                document.body.classList.toggle('has-queue', parseInt(data.q_length) > 0);
 
                 // Update queue bubble in action sidebar
                 //if (queueBubble) {

changedetectionio/static/js/search-modal.js

@@ -69,6 +69,19 @@
       }
     });
 
+    // Handle Enter key in search input
+    if (searchInput) {
+      searchInput.addEventListener('keydown', function(e) {
+        if (e.key === 'Enter') {
+          e.preventDefault();
+          if (searchForm) {
+            // Trigger form submission programmatically
+            searchForm.dispatchEvent(new Event('submit'));
+          }
+        }
+      });
+    }
+
     // Handle form submission
     if (searchForm) {
       searchForm.addEventListener('submit', function(e) {
@@ -88,8 +101,10 @@
           params.append('tags', tags);
         }
 
-        // Navigate to search results
-        window.location.href = '?' + params.toString();
+        // Navigate to search results (always redirect to watchlist home)
+        // Use base_path if available (for sub-path deployments like /enlighten-richerx)
+        const basePath = typeof base_path !== 'undefined' ? base_path : '';
+        window.location.href = basePath + '/?' + params.toString();
       });
     }
   });

changedetectionio/static/styles/diff.css

@@ -1 +1 @@
-#diff-form{background:rgba(0,0,0,.05);padding:1em;border-radius:10px;margin-bottom:1em;color:#fff;font-size:.9rem;text-align:center}#diff-form label.from-to-label{width:4rem;text-decoration:none;padding:.5rem}#diff-form label.from-to-label#change-from{color:#b30000;background:#fadad7}#diff-form label.from-to-label#change-to{background:#eaf2c2;color:#406619}#diff-form #diff-style>span{display:inline-block;padding:.3em}#diff-form #diff-style>span label{font-weight:normal}#diff-form *{vertical-align:middle}body.difference-page section.content{padding-top:40px}#diff-ui{background:var(--color-background);padding:1rem;border-radius:5px}@media(min-width: 767px){#diff-ui{min-width:50%}}#diff-ui #text{font-size:11px}#diff-ui pre{white-space:break-spaces}#diff-ui h1{display:inline;font-size:100%}#diff-ui #result{white-space:pre-wrap;word-break:break-word;overflow-wrap:break-word}#diff-ui .source{position:absolute;right:1%;top:.2em}@-moz-document url-prefix(){#diff-ui body{height:99%}}#diff-ui td#diff-col div{text-align:justify;white-space:pre-wrap}#diff-ui .ignored{background-color:#ccc;opacity:.7}#diff-ui .triggered{background-color:#1b98f8}#diff-ui .ignored.triggered{background-color:red}#diff-ui .tab-pane-inner#screenshot{text-align:center}#diff-ui .tab-pane-inner#screenshot img{max-width:99%}#diff-ui .pure-form button.reset-margin{margin:0px}#diff-ui .diff-fieldset{display:flex;align-items:center;gap:4px;flex-wrap:wrap}#diff-ui ul#highlightSnippetActions{list-style-type:none;display:flex;align-items:center;justify-content:center;gap:1.5rem;flex-wrap:wrap;padding:0;margin:0}#diff-ui ul#highlightSnippetActions li{display:flex;flex-direction:column;align-items:center;text-align:center;padding:.5rem;gap:.3rem}#diff-ui ul#highlightSnippetActions li button,#diff-ui ul#highlightSnippetActions li a{white-space:nowrap}#diff-ui ul#highlightSnippetActions span{font-size:.8rem;color:var(--color-text-input-description)}#diff-ui #cell-diff-jump-visualiser{display:flex;flex-direction:row;gap:1px;background:var(--color-background);border-radius:3px;overflow-x:hidden;position:sticky;top:0;z-index:10;padding-top:1rem;padding-bottom:1rem;justify-content:center}#diff-ui #cell-diff-jump-visualiser>div{flex:1;min-width:1px;max-width:10px;height:10px;background:var(--color-background-button-cancel);opacity:.3;border-radius:1px;transition:opacity .2s;position:relative}#diff-ui #cell-diff-jump-visualiser>div.deletion{background:#b30000;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.insertion{background:#406619;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.note{background:#406619;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.mixed{background:linear-gradient(to right, #b30000 50%, #406619 50%);opacity:1}#diff-ui #cell-diff-jump-visualiser>div.current-position::after{content:"";position:absolute;bottom:-6px;left:50%;transform:translateX(-50%);width:0;height:0;border-left:4px solid rgba(0,0,0,0);border-right:4px solid rgba(0,0,0,0);border-bottom:4px solid var(--color-text)}#diff-ui #cell-diff-jump-visualiser>div:hover{opacity:.8;cursor:pointer}#text-diff-heading-area .snapshot-age{padding:4px;margin:.5rem 0;background-color:var(--color-background-snapshot-age);border-radius:3px;font-weight:bold;margin-bottom:4px}#text-diff-heading-area .snapshot-age.error{background-color:var(--color-error-background-snapshot-age);color:var(--color-error-text-snapshot-age)}#text-diff-heading-area .snapshot-age>*{padding-right:1rem}
+#diff-form{background:rgba(0,0,0,.05);padding:1em;border-radius:10px;margin-bottom:1em;color:#fff;font-size:.9rem;text-align:center}#diff-form label.from-to-label{width:4rem;text-decoration:none;padding:.5rem}#diff-form label.from-to-label#change-from{color:#b30000;background:#fadad7}#diff-form label.from-to-label#change-to{background:#eaf2c2;color:#406619}#diff-form #diff-style>span{display:inline-block;padding:.3em}#diff-form #diff-style>span label{font-weight:normal}#diff-form *{vertical-align:middle}body.difference-page section.content{padding-top:40px}#diff-ui{background:var(--color-background);padding:1rem;border-radius:5px}@media(min-width: 767px){#diff-ui{min-width:50%}}#diff-ui #text{font-size:11px}#diff-ui pre{white-space:break-spaces;overflow-wrap:anywhere}#diff-ui h1{display:inline;font-size:100%}#diff-ui #result{white-space:pre-wrap;word-break:break-word;overflow-wrap:break-word}#diff-ui .source{position:absolute;right:1%;top:.2em}@-moz-document url-prefix(){#diff-ui body{height:99%}}#diff-ui td#diff-col div{text-align:justify;white-space:pre-wrap}#diff-ui .ignored{background-color:#ccc;opacity:.7}#diff-ui .triggered{background-color:#1b98f8}#diff-ui .ignored.triggered{background-color:red}#diff-ui .tab-pane-inner#screenshot{text-align:center}#diff-ui .tab-pane-inner#screenshot img{max-width:99%}#diff-ui .pure-form button.reset-margin{margin:0px}#diff-ui .diff-fieldset{display:flex;align-items:center;gap:4px;flex-wrap:wrap}#diff-ui ul#highlightSnippetActions{list-style-type:none;display:flex;align-items:center;justify-content:center;gap:1.5rem;flex-wrap:wrap;padding:0;margin:0}#diff-ui ul#highlightSnippetActions li{display:flex;flex-direction:column;align-items:center;text-align:center;padding:.5rem;gap:.3rem}#diff-ui ul#highlightSnippetActions li button,#diff-ui ul#highlightSnippetActions li a{white-space:nowrap}#diff-ui ul#highlightSnippetActions span{font-size:.8rem;color:var(--color-text-input-description)}#diff-ui #cell-diff-jump-visualiser{display:flex;flex-direction:row;gap:1px;background:var(--color-background);border-radius:3px;overflow-x:hidden;position:sticky;top:0;z-index:10;padding-top:1rem;padding-bottom:1rem;justify-content:center}#diff-ui #cell-diff-jump-visualiser>div{flex:1;min-width:1px;max-width:10px;height:10px;background:var(--color-background-button-cancel);opacity:.3;border-radius:1px;transition:opacity .2s;position:relative}#diff-ui #cell-diff-jump-visualiser>div.deletion{background:#b30000;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.insertion{background:#406619;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.note{background:#406619;opacity:1}#diff-ui #cell-diff-jump-visualiser>div.mixed{background:linear-gradient(to right, #b30000 50%, #406619 50%);opacity:1}#diff-ui #cell-diff-jump-visualiser>div.current-position::after{content:"";position:absolute;bottom:-6px;left:50%;transform:translateX(-50%);width:0;height:0;border-left:4px solid rgba(0,0,0,0);border-right:4px solid rgba(0,0,0,0);border-bottom:4px solid var(--color-text)}#diff-ui #cell-diff-jump-visualiser>div:hover{opacity:.8;cursor:pointer}#text-diff-heading-area .snapshot-age{padding:4px;margin:.5rem 0;background-color:var(--color-background-snapshot-age);border-radius:3px;font-weight:bold;margin-bottom:4px}#text-diff-heading-area .snapshot-age.error{background-color:var(--color-error-background-snapshot-age);color:var(--color-error-text-snapshot-age)}#text-diff-heading-area .snapshot-age>*{padding-right:1rem}

changedetectionio/static/styles/scss/diff.scss

@@ -62,6 +62,7 @@ body.difference-page {
 
   pre {
     white-space: break-spaces;
+    overflow-wrap: anywhere;
   }
 
 

changedetectionio/static/styles/scss/parts/_hamburger_menu.scss

@@ -110,6 +110,9 @@
           background: var(--color-background-menu-link-hover);
         }
       }
+      &#menu-pause, &#menu-mute {
+        display: none;
+      }
     }
   }
 }

changedetectionio/static/styles/scss/parts/_menu.scss

@@ -2,6 +2,13 @@
   padding: 0.5rem 1em;
   line-height: 1.2rem;
 }
+#menu-mute, #menu-pause {
+  padding-left: 0.3rem;
+  padding-right: 0.3rem;
+  img {
+    height: 1.2rem;
+  }
+}
 
 .pure-menu-item {
   svg {

changedetectionio/static/styles/scss/parts/_pagination.scss

@@ -1,6 +1,4 @@
 .pagination-page-info {
-  color: #fff;
-  font-size: 0.85rem;
   text-transform: capitalize;
 }
 

changedetectionio/static/styles/scss/parts/_toast.scss

@@ -184,24 +184,24 @@ html[data-darkmode=true] {
 // Mobile adjustments
 @media only screen and (max-width: 768px) {
   .toast-container {
-    left: 10px !important;
-    right: 10px !important;
-    top: 10px !important;
-    transform: none !important;
-    align-items: stretch;
+    left: 50% !important;
+    right: auto !important;
+    top: 80px !important;
+    transform: translateX(-50%) !important;
+    align-items: center;
 
     &.toast-bottom-right,
     &.toast-bottom-center,
     &.toast-bottom-left {
       top: auto !important;
-      bottom: 10px !important;
+      bottom: 80px !important;
     }
   }
 
   .toast {
     min-width: auto;
     max-width: none;
-    width: 100%;
+    width: 80vw;
     transform: translateY(-100px);
 
     &.toast-show {

changedetectionio/static/styles/scss/parts/_watch_table-mobile.scss

@@ -125,6 +125,11 @@ $grid-gap: 0.5rem;
           border-bottom: none;
         }
 
+        // Empty state message - span full width on mobile
+        > td[colspan] {
+          grid-column: 1 / -1;
+        }
+
         > td.title-col {
           grid-column: 1 / -1;
           grid-row: 1;

changedetectionio/static/styles/scss/parts/_watch_table.scss

@@ -1,4 +1,32 @@
 /* table related */
+#stats_row {
+  display: flex;
+  align-items: center;
+  width: 100%;
+  color: #fff;
+  font-size: 0.85rem;
+  >* {
+    padding-bottom: 0.5rem;
+  }
+  .left {
+    text-align: left;
+  }
+
+  .right {
+    opacity: 0.5;
+    transition: opacity 0.6s ease;
+    margin-left: auto; /* pushes it to the far right */
+    text-align: right;
+  }
+}
+body.has-queue {
+  #stats_row {
+    .right {
+      opacity: 1.0;
+    }
+  }
+}
+
 .watch-table {
   width: 100%;
   font-size: 80%;