ci: Add GitHub action for tessdata

.github/workflows/ci.yml

@@ -66,8 +66,32 @@ jobs:
           sudo apt-get install -y python3-poetry
           python3 ./install/common/build-image.py
 
+  download-tessdata:
+    name: Download and cache Tesseract data
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+    - name: Cache Tessdata
+      id: cache-tessdata
+      uses: actions/cache@v4
+      with:
+        path: share/tessdata/
+        key: v1-tessdata-${{ hashFiles('./install/common/download-tessdata.py') }}
+        enableCrossOsArchive: true
+    - uses: actions/setup-python@v4
+      with:
+        python-version: '3.11'
+    - name: Download Tessdata
+      run: |-
+          if [ -f "share/tessdata" ]; then
+            echo "Already cached, skipping"
+          else
+            python3 ./install/common/download-tessdata.py
+          fi
+
   windows:
     runs-on: windows-latest
+    needs: download-tessdata
     env:
       DUMMY_CONVERSION: 1
     steps:
@@ -77,6 +101,13 @@ jobs:
           python-version: "3.12"
       - run: pip install poetry
       - run: poetry install
+      - name: Restore cached tessdata
+        uses: actions/cache/restore@v4
+        with:
+          path: share/tessdata/
+          enableCrossOsArchive: true
+          fail-on-cache-miss: true
+          key: v1-tessdata-${{ hashFiles('./install/common/download-tessdata.py') }}
       - name: Run CLI tests
         run: poetry run make test
       # Taken from: https://github.com/orgs/community/discussions/27149#discussioncomment-3254829
@@ -90,6 +121,7 @@ jobs:
   macOS:
     name: "macOS (${{ matrix.arch }})"
     runs-on: ${{ matrix.runner }}
+    needs: download-tessdata
     strategy:
       matrix:
         include:
@@ -104,6 +136,13 @@ jobs:
       - uses: actions/setup-python@v5
         with:
           python-version: "3.12"
+      - name: Restore cached tessdata
+        uses: actions/cache/restore@v4
+        with:
+          path: share/tessdata/
+          enableCrossOsArchive: true
+          fail-on-cache-miss: true
+          key: v1-tessdata-${{ hashFiles('./install/common/download-tessdata.py') }}
       - run: pip install poetry
       - run: poetry install
       - name: Run CLI tests
@@ -301,7 +340,9 @@ jobs:
   run-tests:
     name: "run tests (${{ matrix.distro }} ${{ matrix.version }})"
     runs-on: ubuntu-latest
-    needs: build-container-image
+    needs:
+      - build-container-image
+      - download-tessdata
     strategy:
       matrix:
         include:
@@ -360,6 +401,22 @@ jobs:
             share/image-id.txt
           fail-on-cache-miss: true
 
+      - name: Restore cached tessdata
+        uses: actions/cache/restore@v4
+        with:
+          path: share/tessdata/
+          enableCrossOsArchive: true
+          fail-on-cache-miss: true
+          key: v1-tessdata-${{ hashFiles('./install/common/download-tessdata.py') }}
+
+      - name: Setup xvfb (Linux)
+        run: |
+          # Stuff copied wildly from several stackoverflow posts
+          sudo apt-get install -y xvfb libxkbcommon-x11-0 libxcb-icccm4 libxcb-image0 libxcb-keysyms1 libxcb-randr0 libxcb-render-util0 libxcb-xinerama0 libxcb-xinput0 libxcb-xfixes0 libxcb-shape0 libglib2.0-0 libgl1-mesa-dev '^libxcb.*-dev' libx11-xcb-dev libglu1-mesa-dev libxrender-dev libxi-dev libxkbcommon-dev libxkbcommon-x11-dev
+
+          # start xvfb in the background
+          sudo /usr/bin/Xvfb $DISPLAY -screen 0 1280x1024x24 &
+
       - name: Run CI tests
         run: |-
           # Pass the -ac Xserver flag, to disable host-based access controls.