[2.x] Merge shell tests into the Java CI pipeline

.github/workflows/maven.yml

@@ -57,3 +57,186 @@ jobs:
       run: mvn -V clean test install --show-version --batch-mode --no-transfer-progress -Pjacoco -Pci
     - name: Jacoco
       run: mvn jacoco:report
+    # The binary distribution is pure-Java and platform-independent, so we build it
+    # once here (Linux/JDK 17) and reuse it for the shell tests below instead of
+    # rebuilding it on every OS.
+    - name: Upload OpenNLP distribution
+      if: matrix.os == 'ubuntu-latest' && matrix.java == 17
+      uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+      with:
+        name: opennlp-distribution
+        path: |
+          opennlp-distr/target/*.tar.gz
+          !opennlp-distr/target/*-src*.tar.gz
+        if-no-files-found: error
+        retention-days: 1
+
+  # Shell tests run against the binary distribution produced by the "build" job
+  # above, so they no longer rebuild OpenNLP themselves.
+  test-unix-shell-ubuntu:
+    name: Shell Tests on Ubuntu
+    needs: build
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+
+      # Note: bats-core/bats-action@3.0.0 is not allowed to be used (needs an INFRA issue)
+      - name: Install Bats (Testing Framework)
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y bats
+
+      - name: Set up JDK 17
+        uses: actions/setup-java@1bcf9fb12cf4aa7d266a90ae39939e61372fe520 # v5.4.0
+        with:
+          distribution: temurin
+          java-version: 17
+
+      - name: Download OpenNLP distribution
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        with:
+          name: opennlp-distribution
+          path: dist
+
+      - name: Find and Extract OpenNLP Distribution
+        run: |
+          # Find the first non-src .tar.gz file in the downloaded artifact
+          TAR_FILE=$(find dist -maxdepth 1 -type f -name "*.tar.gz" ! -name "*-src*.tar.gz" | head -n 1)
+
+          # Ensure we found a file
+          if [ -z "$TAR_FILE" ]; then
+            echo "Error: No matching tar.gz file found in dist"
+            exit 1
+          fi
+
+          # Extract the tar.gz file
+          tar -xzf "$TAR_FILE" -C $HOME
+
+          # Get the directory name of the extracted content
+          EXTRACTED_DIR=$(tar -tf "$TAR_FILE" | head -n 1 | cut -f1 -d"/")
+
+          # Set OPENNLP_HOME dynamically
+          echo "OPENNLP_HOME=$HOME/$EXTRACTED_DIR" >> $GITHUB_ENV
+          echo "$HOME/$EXTRACTED_DIR/bin" >> $GITHUB_PATH
+
+      - name: Verify Extraction
+        run: |
+          echo "OPENNLP_HOME: $OPENNLP_HOME"
+          ls -l $OPENNLP_HOME/bin
+
+      - name: Run Bats Tests
+        run: |
+          bats ./opennlp-distr/src/test/sh
+        env:
+          JAVA_HOME: ${{ env.JAVA_HOME }}
+          OPENNLP_HOME: ${{ env.OPENNLP_HOME }}
+
+  test-unix-shell-macos:
+    name: Shell Tests on macOS
+    needs: build
+    runs-on: macos-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+
+      - name: Install Bats (Testing Framework)
+        run: |
+          brew update
+          brew install bats-core
+
+      - name: Set up JDK 17
+        uses: actions/setup-java@1bcf9fb12cf4aa7d266a90ae39939e61372fe520 # v5.4.0
+        with:
+          distribution: temurin
+          java-version: 17
+
+      - name: Download OpenNLP distribution
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        with:
+          name: opennlp-distribution
+          path: dist
+
+      - name: Find and Extract OpenNLP Distribution
+        run: |
+          TAR_FILE=$(find dist -maxdepth 1 -type f -name "*.tar.gz" ! -name "*-src*.tar.gz" | head -n 1)
+          if [ -z "$TAR_FILE" ]; then
+            echo "Error: No matching tar.gz file found in dist"
+            exit 1
+          fi
+          tar -xzf "$TAR_FILE" -C $HOME
+          EXTRACTED_DIR=$(tar -tf "$TAR_FILE" | head -n 1 | cut -f1 -d"/")
+          echo "OPENNLP_HOME=$HOME/$EXTRACTED_DIR" >> $GITHUB_ENV
+          echo "$HOME/$EXTRACTED_DIR/bin" >> $GITHUB_PATH
+
+      - name: Verify Extraction
+        run: |
+          echo "OPENNLP_HOME: $OPENNLP_HOME"
+          ls -l $OPENNLP_HOME/bin
+
+      - name: Run Bats Tests
+        run: |
+          bats ./opennlp-distr/src/test/sh
+        env:
+          JAVA_HOME: ${{ env.JAVA_HOME }}
+          OPENNLP_HOME: ${{ env.OPENNLP_HOME }}
+
+  test-windows-shell:
+    name: Shell Tests on Windows
+    needs: build
+    runs-on: windows-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+
+      - name: Install Pester
+        run: |
+          Install-Module -Name Pester -Force -Scope CurrentUser
+          Import-Module Pester
+        shell: pwsh
+
+      - name: Set up JDK 17
+        uses: actions/setup-java@1bcf9fb12cf4aa7d266a90ae39939e61372fe520 # v5.4.0
+        with:
+          distribution: temurin
+          java-version: 17
+
+      - name: Download OpenNLP distribution
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        with:
+          name: opennlp-distribution
+          path: dist
+
+      - name: Run Pester Tests # (one step to avoid environment issues on Windows)
+        run: |
+          # Find the first non-src .tar.gz file in the downloaded artifact
+          $TAR_FILE = Get-ChildItem -Path dist -Filter "*.tar.gz" | Where-Object { $_.Name -notlike "*-src*" } | Select-Object -First 1
+
+          # Ensure we found a file
+          if (-not $TAR_FILE) {
+            Write-Error "Error: No matching tar.gz file found in dist"
+            exit 1
+          }
+
+          # Extract the tar.gz file to the current directory
+          $Destination = "$(pwd)"
+          tar -xzf $TAR_FILE.FullName -C $Destination
+
+          # Get the directory name of the extracted content (excluding the tar path)
+          $EXTRACTED_DIR = (tar -tf $TAR_FILE.FullName | Select-Object -First 1).Split('/')[0]
+
+          # Set OPENNLP_HOME dynamically in the environment
+          $OPENNLP_HOME_PATH = "$Destination\$EXTRACTED_DIR"
+          Write-Host "OPENNLP_HOME=$OPENNLP_HOME_PATH"  # Debugging
+
+          # Ensure OPENNLP_HOME is recognized in the current session
+          $env:OPENNLP_HOME = $OPENNLP_HOME_PATH
+          $env:PATH = "$env:OPENNLP_HOME\bin;$env:PATH"
+
+          Invoke-Pester -Script "./opennlp-distr/src/test/ps/test_opennlp.Tests.ps1" -Output Detailed
+        shell: pwsh
+        env:
+          JAVA_HOME: ${{ env.JAVA_HOME }}