Add Rocky ci-wheel Java packaging build and packaged-JAR tests

.github/workflows/build.yaml

@@ -505,46 +505,39 @@ jobs:
       push: true
       cuda: '["12.9", "13.3"]'
   # Build the cuDF Java JAR for every Maven classifier.
-  java-build:
+  java-build-matrix:
     needs: [telemetry-setup]
-    strategy:
-      fail-fast: false
-      matrix:
-        include:
-          - { cuda: "12.9", cuda_major: "12", arch: "x86_64", runner: "linux-amd64-cpu16" }
-          - { cuda: "13.3", cuda_major: "13", arch: "x86_64", runner: "linux-amd64-cpu16" }
-          - { cuda: "12.9", cuda_major: "12", arch: "aarch64", runner: "linux-arm64-cpu16" }
-          - { cuda: "13.3", cuda_major: "13", arch: "aarch64", runner: "linux-arm64-cpu16" }
-    runs-on: ${{ matrix.runner }}
     permissions:
       contents: read
-    steps:
-      - name: Checkout code repo
-        uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
-        with:
-          ref: ${{ inputs.sha }}
-          fetch-depth: 0
-          persist-credentials: false
-      - name: Build static libcudf
-        run: |
-          ./java/ci/build_static_libcudf.sh \
-            --output-dir "${RUNNER_TEMP}/libcudf" \
-            --cuda-version "${{ matrix.cuda }}"
-      - name: Build cuDF Java JAR
-        run: |
-          ./java/ci/build_cudf_java_jar.sh \
-            --libcudf-dir "${RUNNER_TEMP}/libcudf" \
-            --output-dir "${RUNNER_TEMP}/jars" \
-            --cuda-version "${{ matrix.cuda }}"
-      - name: Upload per-entry JAR artifact
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
-        with:
-          name: cudf_java_cuda${{ matrix.cuda_major }}_${{ matrix.arch }}
-          # Ship only the JARs and POMs. Exclude the per-classifier Maven build scratch dir.
-          path: |
-            ${{ runner.temp }}/jars
-            !${{ runner.temp }}/jars/.mvn-temp-target
-          if-no-files-found: error
+    uses: rapidsai/shared-workflows/.github/workflows/compute-matrix.yaml@main
+    with:
+      build_type: ${{ inputs.build_type || 'branch' }}
+      matrix_name: conda-cpp-build
+  java-build:
+    needs: [build-details, telemetry-setup, java-build-matrix]
+    permissions:
+      actions: read
+      contents: read
+      id-token: write
+      packages: read
+      pull-requests: read
+    secrets: inherit # zizmor: ignore[secrets-inherit]
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJSON(needs.java-build-matrix.outputs.matrix) }}
+    with:
+      build_type: ${{ inputs.build_type || 'branch' }}
+      branch: ${{ inputs.branch }}
+      date: ${{ inputs.date }}
+      sha: ${{ inputs.sha }}
+      build-datetime: ${{ needs.build-details.outputs.build-datetime }}
+      arch: ${{ matrix.ARCH }}
+      node_type: cpu16
+      container_image: "rapidsai/ci-wheel:26.10-cuda${{ matrix.CUDA_VER }}-${{ matrix.LINUX_VER }}-py${{ matrix.PY_VER }}"
+      script: "env RAPIDS_CUDA_VERSION=${{ matrix.CUDA_VER }} ci/build_java.sh"
+      file_to_upload: output_jars
+      artifact-name: java_pkg_${{ matrix.CUDA_VER }}_${{ matrix.ARCH }}
   # Assemble the per-classifier JARs into one Maven-repository layout.
   java-gather:
     needs: [java-build]
@@ -560,7 +553,7 @@ jobs:
       - name: Download per-entry JAR artifacts
         uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
-          pattern: cudf_java_cuda*
+          pattern: java_pkg_*
           path: ${{ runner.temp }}/jars
           merge-multiple: true
       - name: Assemble Maven repository layout

.github/workflows/pr.yaml

@@ -25,6 +25,9 @@ jobs:
       - conda-python-cudf-tests
       - conda-python-other-tests
       - conda-java-tests
+      - java-build-matrix
+      - java-build
+      - java-tests
       - conda-notebook-tests
       - docs-build
       - publish-api-docs
@@ -526,6 +529,75 @@ jobs:
       arch: "amd64"
       container_image: "rapidsai/ci-conda:26.10-latest"
       script: "ci/test_java.sh"
+  # Build and test the cuDF Java JAR for every Maven classifier.
+  java-build-matrix:
+    needs: [checks]
+    permissions:
+      contents: read
+    uses: rapidsai/shared-workflows/.github/workflows/compute-matrix.yaml@main
+    with:
+      build_type: pull-request
+      matrix_name: conda-cpp-build
+  java-build:
+    needs: [build-details, java-build-matrix, changed-files]
+    permissions:
+      actions: read
+      contents: read
+      id-token: write
+      packages: read
+      pull-requests: read
+    secrets: inherit # zizmor: ignore[secrets-inherit]
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_java
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJSON(needs.java-build-matrix.outputs.matrix) }}
+    with:
+      build_type: pull-request
+      build-datetime: ${{ needs.build-details.outputs.build-datetime }}
+      arch: ${{ matrix.ARCH }}
+      node_type: cpu16
+      container_image: "rapidsai/ci-wheel:26.10-cuda${{ matrix.CUDA_VER }}-${{ matrix.LINUX_VER }}-py${{ matrix.PY_VER }}"
+      script: "env RAPIDS_CUDA_VERSION=${{ matrix.CUDA_VER }} ci/build_java.sh"
+      file_to_upload: output_jars
+      artifact-name: java_pkg_${{ matrix.CUDA_VER }}_${{ matrix.ARCH }}
+  java-tests:
+    needs: [java-build, java-build-matrix]
+    if: ${{ !cancelled() && needs.java-build.result == 'success' }}
+    permissions:
+      actions: read
+      contents: read
+      packages: read
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJSON(needs.java-build-matrix.outputs.matrix) }}
+    runs-on: linux-${{ matrix.ARCH }}-gpu-l4-latest-1
+    container:
+      image: rapidsai/ci-wheel:26.10-cuda${{ matrix.CUDA_VER }}-${{ matrix.LINUX_VER }}-py${{ matrix.PY_VER }} # zizmor: ignore[unpinned-images]
+      env:
+        NVIDIA_VISIBLE_DEVICES: ${{ env.NVIDIA_VISIBLE_DEVICES }}
+    defaults:
+      run:
+        shell: bash
+    steps:
+      - name: Checkout
+        uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
+        with:
+          persist-credentials: false
+      - name: Download java-build artifact
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        with:
+          name: java_pkg_${{ matrix.CUDA_VER }}_${{ matrix.ARCH }}
+          path: java_pkg
+      - name: Run Java tests
+        env:
+          RAPIDS_CUDA_VERSION: ${{ matrix.CUDA_VER }}
+          LIBCUDF_LARGE_STRINGS_ENABLED: "0"
+        run: |
+          . java/ci/java_classifier.sh
+          JAVA_JAR="$(cudf_java_resolve_artifact_jar java_pkg)"
+          export JAVA_JAR
+          ci/test_packaged_java.sh
   conda-notebook-tests:
     needs: [conda-python-build, conda-python-build-noarch, changed-files]
     permissions:

ci/build_java.sh

@@ -0,0 +1,60 @@
+#!/bin/bash
+# SPDX-FileCopyrightText: Copyright (c) 2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+
+# CI entrypoint: static libcudf + classifier JAR for one matrix cell.
+#
+# Builds the static libcudf install tree and packages the matching classifier
+# JAR, writing ./output_jars/<classifier>/ for the custom-job upload step.
+#
+# Inputs (environment variables):
+#   RAPIDS_CUDA_VERSION        CUDA version, e.g. 12.9.2 (required).
+#   PARALLEL_LEVEL             Build parallelism (default: nproc).
+#   CMAKE_CUDA_ARCHITECTURES   Optional override for -DCMAKE_CUDA_ARCHITECTURES.
+#   JAVA_WORK_DIR              Optional scratch dir (default: <repo>/.java-work).
+
+set -euo pipefail
+
+REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${REPO_ROOT}"
+
+# shellcheck disable=SC1091
+. "${REPO_ROOT}/java/ci/ci_wheel_image.sh"
+# shellcheck disable=SC1091
+. "${REPO_ROOT}/java/ci/java_classifier.sh"
+
+if [[ -z ${RAPIDS_CUDA_VERSION:-} ]]; then
+  echo "Error: RAPIDS_CUDA_VERSION must be set" >&2
+  exit 1
+fi
+
+RAPIDS_CUDA_VERSION="$(cudf_java_normalize_cuda_version "${RAPIDS_CUDA_VERSION}")"
+export RAPIDS_CUDA_VERSION
+CLASSIFIER="$(cudf_java_maven_classifier "${RAPIDS_CUDA_VERSION}")"
+
+WORK_DIR="${JAVA_WORK_DIR:-${REPO_ROOT}/.java-work}"
+LIBCUDF_DIR="${WORK_DIR}/libcudf"
+CLASSIFIER_OUT="${REPO_ROOT}/output_jars/${CLASSIFIER}"
+
+cleanup_scratch() {
+  rm -rf "${WORK_DIR}"
+}
+trap cleanup_scratch EXIT
+
+rm -rf "${LIBCUDF_DIR}" "${CLASSIFIER_OUT}"
+mkdir -p "${LIBCUDF_DIR}" "${CLASSIFIER_OUT}"
+
+export REPO_ROOT
+export INSTALL_PREFIX="${LIBCUDF_DIR}"
+export BUILD_DIR="${WORK_DIR}/libcudf-build"
+export CUDF_INSTALL_DIR="${LIBCUDF_DIR}"
+export OUTPUT_DIR="${CLASSIFIER_OUT}"
+
+rapids-logger "Building static libcudf (${CLASSIFIER})"
+bash "${REPO_ROOT}/java/ci/build_static_libcudf_in_container.sh"
+
+rapids-logger "Packaging cuDF Java JAR (${CLASSIFIER})"
+bash "${REPO_ROOT}/java/ci/build_cudf_java_jar_in_container.sh"
+
+cudf_java_assert_classifier_artifacts "${CLASSIFIER_OUT}" "${CLASSIFIER}"
+ls -la "${CLASSIFIER_OUT}"

ci/test_packaged_java.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+# SPDX-FileCopyrightText: Copyright (c) 2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+
+# Run the Java tests against an already-packaged classifier JAR.
+#
+# Activates -Ppackaged-jar-tests so the surefire classpath uses the JAR at
+# JAVA_JAR instead of a locally compiled target/classes tree. Caller places
+# the JAR (e.g. via actions/download-artifact) before invoking this script.
+#
+# Inputs (environment variables):
+#   JAVA_JAR                       Absolute path to the classifier JAR (required).
+#   LIBCUDF_LARGE_STRINGS_ENABLED  Optional; defaults to 0 (same as ci/test_java.sh).
+
+set -euo pipefail
+
+REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "${REPO_ROOT}"
+
+if [[ -z ${JAVA_JAR:-} || ! -f ${JAVA_JAR} ]]; then
+  echo "Error: JAVA_JAR must point to an existing classifier JAR" >&2
+  exit 1
+fi
+
+if ! command -v mvn >/dev/null 2>&1 || ! command -v java >/dev/null 2>&1; then
+  # shellcheck disable=SC1091
+  . "${REPO_ROOT}/java/ci/setup_java_env.sh"
+fi
+
+# Match the existing conda Java test entrypoint in ci/test_java.sh.
+export LIBCUDF_LARGE_STRINGS_ENABLED="${LIBCUDF_LARGE_STRINGS_ENABLED:-0}"
+
+PRODUCT_JAR="$(cd "$(dirname "${JAVA_JAR}")" && pwd)/$(basename "${JAVA_JAR}")"
+rapids-logger "Product JAR: ${PRODUCT_JAR}"
+
+rm -rf "${REPO_ROOT}/java/target"
+
+pushd "${REPO_ROOT}/java" >/dev/null
+set +e
+timeout 30m mvn -B test -Ppackaged-jar-tests \
+  "-Dcudf.jar.path=${PRODUCT_JAR}" \
+  "-DCUDF_JNI_ENABLE_PROFILING=OFF"
+EXITCODE=$?
+set -e
+popd >/dev/null
+rapids-logger "Java tests exit=${EXITCODE}"
+exit "${EXITCODE}"

java/ci/README.md

@@ -4,14 +4,14 @@
 
 The scripts under `java/ci/` build the cuDF Java JAR for every Maven classifier the
 same way locally and in CI (GitHub Actions is only a thin wrapper that adds
-artifact upload/download). Each script pulls the RAPIDS `ci-conda` build image,
+artifact upload/download). Each script pulls the RAPIDS `ci-wheel` build image,
 runs the build in a throwaway container, and writes its output to a host
 directory. No local `docker build` is required, and no GPU is required to build.
 
 ### Prerequisites
 
 1. Docker is installed and the current user can run `docker`.
-2. Network access to pull `rapidsai/ci-conda:<rapids_version>-latest`.
+2. Network access to pull `rapidsai/ci-wheel:<rapids>-cuda<ver>-rockylinux8-py3.11`.
 
 ### Local one-command shortcut
 
@@ -90,6 +90,17 @@ per-entry artifact. The separate `java-gather` job downloads them (with
 `merge-multiple: true`, so all subdirs land in a single parent dir), runs
 Step 3, and uploads the combined `cudf_java_maven_repo` artifact.
 
+### Packaging-aware tests (local)
+
+Plain `cd java && mvn test` does not exercise the classifier JAR. Use
+`java/ci/test_packaged_java_local.sh` (or CI entrypoint
+`ci/test_packaged_java.sh`) to run the existing Java tests against a packaged
+JAR. Needs Docker + GPU.
+
+```bash
+./java/ci/test_packaged_java_local.sh --work-dir /tmp/java-build-test
+```
+
 ## Legacy: manual Dockerfile.rocky build (obsolete)
 
 > The `java/ci/Dockerfile.rocky` + `java/ci/build-in-docker.sh` flow below is the

java/ci/build-in-docker.sh

@@ -69,7 +69,7 @@ BUILD_ARG=(
   "-DCUDF_USE_PER_THREAD_DEFAULT_STREAM=$ENABLE_PTDS"
   "-DCUDF_JNI_LIBCUDF_STATIC=ON"
   "-DUSE_GDS=$ENABLE_GDS"
-  "-Dtest=*,!CuFileTest,!CudaFatalTest,!ColumnViewNonEmptyNullsTest,!NativeDepsLoaderTest"
+  "-Dtest=*,!CuFileTest,!CudaFatalTest,!ColumnViewNonEmptyNullsTest,!NativeDepsLoaderTest,!PackagedJarOriginCheck"
 )
 
 if [ "$SIGN_FILE" == true ]; then