Build Playwright node components from source

Dockerfile

@@ -19,12 +19,15 @@ RUN \
     libc-dev \
     libffi-dev \
     libxslt-dev \
+    npm \
     openssl-dev \
     python3-dev \
+    zip \
     zlib-dev && \
   apk add --update --no-cache \
     libjpeg \
     libxslt \
+    nodejs \
     poppler-utils \
     python3 && \
   echo "**** install changedetection.io ****" && \
@@ -45,16 +48,35 @@ RUN \
     pip \
     wheel && \
   pip install -U --no-cache-dir --find-links https://wheel-index.linuxserver.io/alpine-3.19/ -r /app/changedetection/requirements.txt && \
-  PLAYWRIGHT_RELEASE=$(curl -sX GET "https://api.github.com/repos/microsoft/playwright-python/releases/latest" \
+  PLAYWRIGHT_PY_RELEASE=$(curl -sX GET "https://api.github.com/repos/microsoft/playwright-python/releases/latest" \
     | awk '/tag_name/{print $4;exit}' FS='[""]'); \
-  git clone --depth 1 --branch "${PLAYWRIGHT_RELEASE}" https://github.com/microsoft/playwright-python /tmp/playwright-python && \
+  git clone --depth 1 --branch "${PLAYWRIGHT_PY_RELEASE}" https://github.com/microsoft/playwright-python /tmp/playwright-python && \
   cd /tmp/playwright-python && \
   pip install -U --no-cache-dir . && \
+  git clone --depth 1 --branch "${PLAYWRIGHT_PY_RELEASE}" https://github.com/microsoft/playwright /tmp/playwright && \
+  cd /tmp/playwright && \
+  npm ci && \
+  npm run build && \
+  # Don't build for other platforms
+  sed -i '/-darwin-x64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-darwin-arm64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-linux-arm64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-win-x64/d' ./utils/build/build-playwright-driver.sh && \
+  # Don't download a static node binary, use the OS install
+  sed -i '/curl ${NODE_URL}/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/elif \[\[ "${ARCHIVE}" == "tar.gz" \]\]; then/,/else/d'  ./utils/build/build-playwright-driver.sh && \
+  sed -i '/cp .\/output\/${NODE_DIR}\/LICENSE .\/output\/playwright-${SUFFIX}\//d' ./utils/build/build-playwright-driver.sh && \
+  sed -i 's/"..\/..\/${NODE_DIR}\/${NPM_PATH}"/\/usr\/lib\/node_modules\/npm\/bin\/npm-cli.js/' ./utils/build/build-playwright-driver.sh && \
+  ./utils/build/build-playwright-driver.sh && \
+  rm -rf /lsiopy/lib/python3.11/site-packages/playwright/driver/* && \
+  cp -R ./utils/build/output/playwright-linux/* /lsiopy/lib/python3.11/site-packages/playwright/driver && \
+  ln -s /usr/bin/node /lsiopy/lib/python3.11/site-packages/playwright/driver/node && \
   apk del --purge \
     build-dependencies && \
   rm -rf \
     /tmp/* \
-    $HOME/.cache
+    $HOME/.cache \
+    $HOME/.npm
 
 COPY root/ /
 

Dockerfile.aarch64

@@ -19,12 +19,15 @@ RUN \
     libc-dev \
     libffi-dev \
     libxslt-dev \
+    npm \
     openssl-dev \
     python3-dev \
+    zip \
     zlib-dev && \
   apk add --update --no-cache \
     libjpeg \
     libxslt \
+    nodejs \
     poppler-utils \
     python3 && \
   echo "**** install changedetection.io ****" && \
@@ -45,16 +48,36 @@ RUN \
     pip \
     wheel && \
   pip install -U --no-cache-dir --find-links https://wheel-index.linuxserver.io/alpine-3.19/ -r /app/changedetection/requirements.txt && \
-  PLAYWRIGHT_RELEASE=$(curl -sX GET "https://api.github.com/repos/microsoft/playwright-python/releases/latest" \
+  echo "**** install playwright ****" && \
+  PLAYWRIGHT_PY_RELEASE=$(curl -sX GET "https://api.github.com/repos/microsoft/playwright-python/releases/latest" \
     | awk '/tag_name/{print $4;exit}' FS='[""]'); \
-  git clone --depth 1 --branch "${PLAYWRIGHT_RELEASE}" https://github.com/microsoft/playwright-python /tmp/playwright-python && \
+  git clone --depth 1 --branch "${PLAYWRIGHT_PY_RELEASE}" https://github.com/microsoft/playwright-python /tmp/playwright-python && \
   cd /tmp/playwright-python && \
   pip install -U --no-cache-dir . && \
+  git clone --depth 1 --branch "${PLAYWRIGHT_PY_RELEASE}" https://github.com/microsoft/playwright /tmp/playwright && \
+  cd /tmp/playwright && \
+  npm ci && \
+  npm run build && \
+  # Don't build for other platforms
+  sed -i '/-darwin-x64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-darwin-arm64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-linux-x64/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/-win-x64/d' ./utils/build/build-playwright-driver.sh && \
+  # Don't download a static node binary, use the OS install
+  sed -i '/curl ${NODE_URL}/d' ./utils/build/build-playwright-driver.sh && \
+  sed -i '/elif \[\[ "${ARCHIVE}" == "tar.gz" \]\]; then/,/else/d'  ./utils/build/build-playwright-driver.sh && \
+  sed -i '/cp .\/output\/${NODE_DIR}\/LICENSE .\/output\/playwright-${SUFFIX}\//d' ./utils/build/build-playwright-driver.sh && \
+  sed -i 's/"..\/..\/${NODE_DIR}\/${NPM_PATH}"/\/usr\/lib\/node_modules\/npm\/bin\/npm-cli.js/' ./utils/build/build-playwright-driver.sh && \
+  ./utils/build/build-playwright-driver.sh && \
+  rm -rf /lsiopy/lib/python3.11/site-packages/playwright/driver/* && \
+  cp -R ./utils/build/output/playwright-linux-arm64/* /lsiopy/lib/python3.11/site-packages/playwright/driver && \
+  ln -s /usr/bin/node /lsiopy/lib/python3.11/site-packages/playwright/driver/node && \
   apk del --purge \
     build-dependencies && \
   rm -rf \
     /tmp/* \
-    $HOME/.cache
+    $HOME/.cache \
+    $HOME/.npm
 
 COPY root/ /
 

README.md

@@ -61,7 +61,7 @@ The architectures supported by this image are:
 
 Webui is accessible at http://SERVERIP:PORT
 
-Please note that as of version 0.45.15 this image contains the [Playwright content fetcher](https://github.com/dgtlmoon/changedetection.io/wiki/Playwright-content-fetcher#docker-compose-based).
+Please note that as of version 0.45.16 this image contains the Playwright content fetcher.
 
 For more info read [the wiki](https://github.com/dgtlmoon/changedetection.io/wiki).
 
@@ -73,18 +73,18 @@ To help you get started creating a container from this image you can either use
 
 ```yaml
 ---
-version: "2.1"
 services:
-  changedetection:
+  changedetection.io:
     image: lscr.io/linuxserver/changedetection.io:latest
     container_name: changedetection
     environment:
       - PUID=1000
       - PGID=1000
-      - TZ=Europe/London
+      - TZ=Etc/UTC
       - BASE_URL= #optional
+      - PLAYWRIGHT_DRIVER_URL= #optional
     volumes:
-      - /path/to/appdata/config:/config
+      - /path/to/changedetection/config:/config
     ports:
       - 5000:5000
     restart: unless-stopped
@@ -99,6 +99,7 @@ docker run -d \
   -e PGID=1000 \
   -e TZ=Etc/UTC \
   -e BASE_URL= `#optional` \
+  -e PLAYWRIGHT_DRIVER_URL= `#optional` \
   -p 5000:5000 \
   -v /path/to/changedetection.io/config:/config \
   --restart unless-stopped \
@@ -116,6 +117,7 @@ Containers are configured using parameters passed at runtime (such as those abov
 | `-e PGID=1000` | for GroupID - see below for explanation |
 | `-e TZ=Etc/UTC` | specify a timezone to use, see this [list](https://en.wikipedia.org/wiki/List_of_tz_database_time_zones#List). |
 | `-e BASE_URL=` | Specify the full URL (including protocol) when running behind a reverse proxy |
+| `-e PLAYWRIGHT_DRIVER_URL=` | Specify the full URL to your chrome driver instance. See the [wiki](https://github.com/dgtlmoon/changedetection.io/wiki/Playwright-content-fetcher) for details. |
 | `-v /config` | Persistent config files |
 
 ## Environment variables from files (Docker secrets)
@@ -279,7 +281,8 @@ Once registered you can define the dockerfile to use with `-f Dockerfile.aarch64
 
 ## Versions
 
-* **08.03.24:** - Build Playwright from source, add libjpeg.
+* **09.03.24:** - Build Playwright from source because Microsoft's build and packaging process is awful.
+* **08.03.24:** - Build Playwright-python from source, add libjpeg.
 * **23.12.23:** - Rebase to Alpine 3.19.
 * **10.08.23:** - Add poppler-utils for pdf conversion tools.
 * **11.06.23:** - Rebase to Alpine 3.18, deprecate armhf.

readme-vars.yml

@@ -33,36 +33,39 @@ param_env_vars:
 opt_param_usage_include_env: true
 opt_param_env_vars:
   - { env_var: "BASE_URL", env_value: "", desc: "Specify the full URL (including protocol) when running behind a reverse proxy" }
+  - { env_var: "PLAYWRIGHT_DRIVER_URL", env_value: "", desc: "Specify the full URL to your chrome driver instance. See the [wiki](https://github.com/dgtlmoon/changedetection.io/wiki/Playwright-content-fetcher) for details." }
 
 custom_compose: |
   ---
-  version: "2.1"
   services:
-    changedetection:
+    changedetection.io:
       image: lscr.io/linuxserver/changedetection.io:latest
       container_name: changedetection
       environment:
         - PUID=1000
         - PGID=1000
-        - TZ=Europe/London
+        - TZ=Etc/UTC
         - BASE_URL= #optional
+        - PLAYWRIGHT_DRIVER_URL= #optional
       volumes:
-        - /path/to/appdata/config:/config
+        - /path/to/changedetection/config:/config
       ports:
         - 5000:5000
       restart: unless-stopped
+
 # application setup block
 app_setup_block_enabled: true
 app_setup_block: |
   Webui is accessible at http://SERVERIP:PORT
 
-  Please note that as of version 0.45.15 this image contains the [Playwright content fetcher](https://github.com/dgtlmoon/changedetection.io/wiki/Playwright-content-fetcher#docker-compose-based).
+  Please note that as of version 0.45.16 this image contains the Playwright content fetcher.
 
   For more info read [the wiki](https://github.com/dgtlmoon/changedetection.io/wiki).
 
 # changelog
 changelogs:
-  - { date: "08.03.24:", desc: "Build Playwright from source, add libjpeg."}
+  - { date: "09.03.24:", desc: "Build Playwright from source because Microsoft's build and packaging process is awful."}
+  - { date: "08.03.24:", desc: "Build Playwright-python from source, add libjpeg."}
   - { date: "23.12.23:", desc: "Rebase to Alpine 3.19."}
   - { date: "10.08.23:", desc: "Add poppler-utils for pdf conversion tools." }
   - { date: "11.06.23:", desc: "Rebase to Alpine 3.18, deprecate armhf." }