Add environment-agnostic scripts for running ctests and pytests

ci/run_cudf_benchmark_smoketests.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# Ensure that benchmarks are runnable
+if [ -d "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/benchmarks/libcudf/" ]; then
+    # Support customizing the ctests' install location
+    cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/benchmarks/libcudf/";
+    # Run a small Google benchmark
+    ./MERGE_BENCH --benchmark_filter=/2/
+    # Run a small nvbench benchmark
+    ./STRINGS_NVBENCH --run-once --benchmark 0 --devices 0
+fi

ci/run_cudf_ctests.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+if [ -d "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf/" ]; then
+    # Support customizing the ctests' install location
+    cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf/";
+    ctest --output-on-failure "$@"
+fi
+
+# Ensure that benchmarks are runnable
+if [ -d "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/benchmarks/libcudf/" ]; then
+    # Support customizing the ctests' install location
+    cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/benchmarks/libcudf/";
+    # Run a small Google benchmark
+    ./MERGE_BENCH --benchmark_filter=/2/
+    # Run a small nvbench benchmark
+    ./STRINGS_NVBENCH --run-once --benchmark 0 --devices 0
+fi

ci/run_cudf_kafka_ctests.sh

@@ -0,0 +1,10 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+if [ -d "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf_kafka/" ]; then
+    # Support customizing the ctests' install location
+    cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf_kafka/";
+    ctest --output-on-failure "$@"
+fi

ci/run_cudf_memcheck_ctests.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -uo pipefail
+
+EXITCODE=0
+trap "EXITCODE=1" ERR
+
+if [ -d "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf/" ]; then
+    # Support customizing the ctests' install location
+    cd "${INSTALL_PREFIX:-${CONDA_PREFIX:-/usr}}/bin/gtests/libcudf/";
+    export GTEST_CUDF_RMM_MODE=cuda
+    for gt in ./*_TEST ; do
+        test_name=$(basename ${gt})
+        # Run gtests with compute-sanitizer
+        if [[ "$test_name" == "ERROR_TEST" ]] || [[ "$test_name" == "STREAM_IDENTIFICATION_TEST" ]]; then
+            continue
+        fi
+        echo "Running compute-sanitizer on $test_name"
+        compute-sanitizer --tool memcheck ${gt} "$@"
+    done
+    unset GTEST_CUDF_RMM_MODE
+fi
+
+exit ${EXITCODE}

ci/run_cudf_pandas_pytest_benchmarks.sh

@@ -0,0 +1,13 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# It is essential to cd into python/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
+
+# Support invoking run_cudf_pandas_pytest_benchmarks.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../python/cudf/
+
+CUDF_BENCHMARKS_USE_PANDAS=ON \
+CUDF_BENCHMARKS_DEBUG_ONLY=ON \
+pytest --cache-clear "$@" benchmarks

ci/run_cudf_pytest_benchmarks.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# It is essential to cd into python/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
+
+# Support invoking run_cudf_pytest_benchmarks.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../python/cudf/
+
+CUDF_BENCHMARKS_DEBUG_ONLY=ON \
+pytest --cache-clear "$@" benchmarks

ci/run_cudf_pytests.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# It is essential to cd into python/cudf/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
+
+# Support invoking run_cudf_pytests.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../python/cudf/cudf/
+
+pytest --cache-clear --ignore="benchmarks" "$@" tests

ci/run_custreamz_pytests.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# It is essential to cd into python/cudf/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
+
+# Support invoking run_custreamz_pytests.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../python/custreamz/custreamz/
+
+pytest --cache-clear "$@" tests

ci/run_dask_cudf_pytests.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+set -euo pipefail
+
+# It is essential to cd into python/cudf/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
+
+# Support invoking run_dask_cudf_pytests.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../python/dask_cudf/dask_cudf/
+
+pytest --cache-clear "$@" .

ci/test_cpp.sh

@@ -1,7 +1,10 @@
 #!/bin/bash
-# Copyright (c) 2022-2023, NVIDIA CORPORATION.
+# Copyright (c) 2022-2024, NVIDIA CORPORATION.
 
-source "$(dirname "$0")/test_cpp_common.sh"
+# Support invoking test_cpp.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../
+
+source ./ci/test_cpp_common.sh
 
 EXITCODE=0
 trap "EXITCODE=1" ERR
@@ -10,36 +13,23 @@ set +e
 # Run libcudf and libcudf_kafka gtests from libcudf-tests package
 export GTEST_OUTPUT=xml:${RAPIDS_TESTS_DIR}/
 
-pushd $CONDA_PREFIX/bin/gtests/libcudf/
 rapids-logger "Run libcudf gtests"
-ctest -j20 --output-on-failure
+./ci/run_cudf_ctests.sh -j20
 SUITEERROR=$?
-popd
 
 if (( ${SUITEERROR} == 0 )); then
-    pushd $CONDA_PREFIX/bin/gtests/libcudf_kafka/
     rapids-logger "Run libcudf_kafka gtests"
-    ctest -j20 --output-on-failure
+    ./ci/run_cudf_kafka_ctests.sh -j20
     SUITEERROR=$?
-    popd
 fi
 
 # Ensure that benchmarks are runnable
-pushd $CONDA_PREFIX/bin/benchmarks/libcudf/
 rapids-logger "Run tests of libcudf benchmarks"
 
 if (( ${SUITEERROR} == 0 )); then
-    # Run a small Google benchmark
-    ./MERGE_BENCH --benchmark_filter=/2/
-    SUITEERROR=$?
-fi
-
-if (( ${SUITEERROR} == 0 )); then
-    # Run a small nvbench benchmark
-    ./STRINGS_NVBENCH --run-once --benchmark 0 --devices 0
+    ./ci/run_cudf_benchmark_smoketests.sh
     SUITEERROR=$?
 fi
-popd
 
 rapids-logger "Test script exiting with value: $EXITCODE"
 exit ${EXITCODE}

ci/test_cpp_memcheck.sh

@@ -1,25 +1,16 @@
 #!/bin/bash
-# Copyright (c) 2023, NVIDIA CORPORATION.
+# Copyright (c) 2023-2024, NVIDIA CORPORATION.
 
-source "$(dirname "$0")/test_cpp_common.sh"
+# Support invoking test_cpp.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../
 
-EXITCODE=0
-trap "EXITCODE=1" ERR
-set +e
+source ./ci/test_cpp_common.sh
 
-# Run gtests with compute-sanitizer
 rapids-logger "Memcheck gtests with rmm_mode=cuda"
-export GTEST_CUDF_RMM_MODE=cuda
-COMPUTE_SANITIZER_CMD="compute-sanitizer --tool memcheck"
-for gt in "$CONDA_PREFIX"/bin/gtests/libcudf/*_TEST ; do
-    test_name=$(basename ${gt})
-    if [[ "$test_name" == "ERROR_TEST" ]] || [[ "$test_name" == "STREAM_IDENTIFICATION_TEST" ]]; then
-        continue
-    fi
-    echo "Running compute-sanitizer on $test_name"
-    ${COMPUTE_SANITIZER_CMD} ${gt} --gtest_output=xml:"${RAPIDS_TESTS_DIR}${test_name}.xml"
-done
-unset GTEST_CUDF_RMM_MODE
+
+./ci/run_cudf_memcheck_ctests.sh \
+    --gtest_output=xml:"${RAPIDS_TESTS_DIR}${test_name}.xml" \
+ && EXITCODE=$? || EXITCODE=$?;
 
 rapids-logger "Test script exiting with value: $EXITCODE"
 exit ${EXITCODE}

ci/test_python_cudf.sh

@@ -1,8 +1,11 @@
 #!/bin/bash
-# Copyright (c) 2022-2023, NVIDIA CORPORATION.
+# Copyright (c) 2022-2024, NVIDIA CORPORATION.
+
+# Support invoking test_python_cudf.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../;
 
 # Common setup steps shared by Python test jobs
-source "$(dirname "$0")/test_python_common.sh"
+source ./ci/test_python_common.sh
 
 rapids-logger "Check GPU usage"
 nvidia-smi
@@ -12,51 +15,37 @@ trap "EXITCODE=1" ERR
 set +e
 
 rapids-logger "pytest cudf"
-pushd python/cudf/cudf
-# It is essential to cd into python/cudf/cudf as `pytest-xdist` + `coverage` seem to work only at this directory level.
-pytest \
-  --cache-clear \
-  --ignore="benchmarks" \
+./ci/run_cudf_pytests.sh \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf.xml" \
   --numprocesses=8 \
   --dist=loadscope \
   --cov-config=../.coveragerc \
   --cov=cudf \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cudf-coverage.xml" \
-  --cov-report=term \
-  tests
-popd
+  --cov-report=term
 
 # Run benchmarks with both cudf and pandas to ensure compatibility is maintained.
 # Benchmarks are run in DEBUG_ONLY mode, meaning that only small data sizes are used.
 # Therefore, these runs only verify that benchmarks are valid.
 # They do not generate meaningful performance measurements.
-pushd python/cudf
+
 rapids-logger "pytest for cudf benchmarks"
-CUDF_BENCHMARKS_DEBUG_ONLY=ON \
-pytest \
-  --cache-clear \
+./ci/run_cudf_pytest_benchmarks.sh \
   --numprocesses=8 \
   --dist=loadscope \
   --cov-config=.coveragerc \
   --cov=cudf \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cudf-benchmark-coverage.xml" \
-  --cov-report=term \
-  benchmarks
+  --cov-report=term
 
 rapids-logger "pytest for cudf benchmarks using pandas"
-CUDF_BENCHMARKS_USE_PANDAS=ON \
-CUDF_BENCHMARKS_DEBUG_ONLY=ON \
-pytest \
-  --cache-clear \
+./ci/run_cudf_pandas_pytest_benchmarks.sh \
   --numprocesses=8 \
   --dist=loadscope \
   --cov-config=.coveragerc \
   --cov=cudf \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cudf-benchmark-pandas-coverage.xml" \
-  --cov-report=term \
-  benchmarks
-popd
+  --cov-report=term
 
 rapids-logger "Test script exiting with value: $EXITCODE"
 exit ${EXITCODE}

ci/test_python_other.sh

@@ -1,8 +1,11 @@
 #!/bin/bash
-# Copyright (c) 2022-2023, NVIDIA CORPORATION.
+# Copyright (c) 2022-2024, NVIDIA CORPORATION.
+
+# Support invoking test_python_cudf.sh outside the script directory
+cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../
 
 # Common setup steps shared by Python test jobs
-source "$(dirname "$0")/test_python_common.sh"
+source ./ci/test_python_common.sh
 
 rapids-mamba-retry install \
   --channel "${CPP_CHANNEL}" \
@@ -17,32 +20,24 @@ trap "EXITCODE=1" ERR
 set +e
 
 rapids-logger "pytest dask_cudf"
-pushd python/dask_cudf/dask_cudf
-pytest \
-  --cache-clear \
+./ci/run_dask_cudf_pytests.sh \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-dask-cudf.xml" \
   --numprocesses=8 \
   --dist=loadscope \
   --cov-config=../.coveragerc \
   --cov=dask_cudf \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/dask-cudf-coverage.xml" \
-  --cov-report=term \
-  .
-popd
+  --cov-report=term
 
 rapids-logger "pytest custreamz"
-pushd python/custreamz/custreamz
-pytest \
-  --cache-clear \
+./ci/run_custreamz_pytests.sh \
   --junitxml="${RAPIDS_TESTS_DIR}/junit-custreamz.xml" \
   --numprocesses=8 \
   --dist=loadscope \
   --cov-config=../.coveragerc \
   --cov=custreamz \
   --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/custreamz-coverage.xml" \
-  --cov-report=term \
-  tests
-popd
+  --cov-report=term
 
 rapids-logger "Test script exiting with value: $EXITCODE"
 exit ${EXITCODE}