dev/test-dependencies.sh - master - githubmirror/spark

dev/test-dependencies.sh
159 строк6 KB
Cheng Pan
[MINOR][BUILD] Mark spark-deps manifest as auto-generated
10 авг 2026, 06:49
10 авг 2026, 06:497b88cea
#!/usr/bin/env bash

#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements.  See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License.  You may obtain a copy of the License at
#
#    http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#

set -ex

FWDIR="$(cd "`dirname $0`"/..; pwd)"
cd "$FWDIR"

# Explicitly set locale in order to make `sort` output consistent across machines.
# See https://stackoverflow.com/questions/28881 for more details.
export LC_ALL=C

# TODO: This would be much nicer to do in SBT, once SBT supports Maven-style resolution.

# NOTE: These should match those in the release publishing script, and be kept in sync with
#   dev/create-release/release-build.sh
HADOOP_MODULE_PROFILES="-Phive-thriftserver -Pkubernetes -Pyarn -Phive \
    -Pspark-ganglia-lgpl -Pkinesis-asl -Pcredential-aws -Phadoop-cloud -Pjvm-profiler"
MVN="build/mvn"
HADOOP_HIVE_PROFILES=(
    hadoop-3-hive-2.3
)

MVN_EXEC_PLUGIN_VERSION=$(build/mvn help:evaluate \
    -Dexpression=exec-maven-plugin.version -q -DforceStdout | grep -E "[0-9]+\.[0-9]+\.[0-9]+")

# We'll switch the version to a temp. one, publish POMs using that new version, then switch back to
# the old version. We need to do this because the `dependency:build-classpath` task needs to
# resolve Spark's internal submodule dependencies.

# From http://stackoverflow.com/a/26514030
set +e
OLD_VERSION=$($MVN -q \
    -Dexec.executable="echo" \
    -Dexec.args='${project.version}' \
    --non-recursive \
    org.codehaus.mojo:exec-maven-plugin:${MVN_EXEC_PLUGIN_VERSION}:exec | grep -E '[0-9]+\.[0-9]+\.[0-9]+')
# dependency:get for guava and jetty-io are workaround for SPARK-37302.
GUAVA_VERSION=$(build/mvn help:evaluate -Dexpression=guava.version -q -DforceStdout | grep -E "^[0-9\.]+")
build/mvn dependency:get -Dartifact=com.google.guava:guava:${GUAVA_VERSION} -q
JETTY_VERSION=$(build/mvn help:evaluate -Dexpression=jetty.version -q -DforceStdout | grep -E "[0-9]+\.[0-9]+\.[0-9]+")
build/mvn dependency:get -Dartifact=org.eclipse.jetty:jetty-io:${JETTY_VERSION} -q
if [ $? != 0 ]; then
    echo -e "Error while getting version string from Maven:\n$OLD_VERSION"
    exit 1
fi
SCALA_BINARY_VERSION=$($MVN -q \
    -Dexec.executable="echo" \
    -Dexec.args='${scala.binary.version}' \
    --non-recursive \
    org.codehaus.mojo:exec-maven-plugin:${MVN_EXEC_PLUGIN_VERSION}:exec | grep -E '[0-9]+\.[0-9]+')
if [[ "$SCALA_BINARY_VERSION" != "2.13" ]]; then
  echo "Skip dependency testing on $SCALA_BINARY_VERSION"
  exit 0
fi
set -e
TEMP_VERSION="spark-$(python3 -S -c "import random; print(random.randrange(100000, 999999))")"

function reset_version {
  # Delete the temporary POMs that we wrote to the local Maven repo:
  find "$HOME/.m2/" | grep "$TEMP_VERSION" | xargs rm -rf

  # Restore the original version number:
  $MVN -q versions:set -DnewVersion=$OLD_VERSION -DgenerateBackupPoms=false > /dev/null
}
trap reset_version EXIT

$MVN -q versions:set -DnewVersion=$TEMP_VERSION -DgenerateBackupPoms=false > /dev/null

# Generate manifests for each Hadoop profile:
for HADOOP_HIVE_PROFILE in "${HADOOP_HIVE_PROFILES[@]}"; do
  if [[ $HADOOP_HIVE_PROFILE == **hadoop-3-hive-2.3** ]]; then
    HADOOP_PROFILE=hadoop-3
  fi
  echo "Performing Maven install for $HADOOP_HIVE_PROFILE"
  $MVN $HADOOP_MODULE_PROFILES -P$HADOOP_PROFILE jar:jar jar:test-jar install:install clean -q

  echo "Performing Maven validate for $HADOOP_HIVE_PROFILE"
  $MVN $HADOOP_MODULE_PROFILES -P$HADOOP_PROFILE validate -q

  echo "Generating dependency manifest for $HADOOP_HIVE_PROFILE"
  mkdir -p dev/pr-deps
  $MVN $HADOOP_MODULE_PROFILES -P$HADOOP_PROFILE dependency:build-classpath -pl assembly -am \
    | grep "Dependencies classpath:" -A 1 \
    | tail -n 1 | tr ":" "\n" | awk -F '/' '{
      # For each dependency classpath, we fetch the last three parts split by "/": artifact id, version, and jar name.
      # Since classifier, if exists, always sits between "artifact_id-version-" and ".jar" suffix in the jar name,
      # we extract classifier and put it right before the jar name explicitly.
      # For example, `orc-core/1.5.5/nohive/orc-core-1.5.5-nohive.jar`
      #                              ^^^^^^
      #                              extracted classifier
      #               `okio/1.15.0//okio-1.15.0.jar`
      #                           ^
      #                           empty for dependencies without classifier
      artifact_id=$(NF-2);
      version=$(NF-1);
      jar_name=$NF;
      classifier_start_index=length(artifact_id"-"version"-") + 1;
      classifier_end_index=index(jar_name, ".jar") - 1;
      classifier=substr(jar_name, classifier_start_index, classifier_end_index - classifier_start_index + 1);
      print artifact_id"/"version"/"classifier"/"jar_name
    }' | sort | grep -v spark > dev/pr-deps/spark-deps-$HADOOP_HIVE_PROFILE.tmp
  {
    printf '%s\n' \
      '#' \
      '# Generated by `dev/test-dependencies.sh --replace-manifest`.' \
      '# Do not edit manually.' \
      '#'
    cat dev/pr-deps/spark-deps-$HADOOP_HIVE_PROFILE.tmp
  } > dev/pr-deps/spark-deps-$HADOOP_HIVE_PROFILE
  rm dev/pr-deps/spark-deps-$HADOOP_HIVE_PROFILE.tmp
done

if [[ $@ == **replace-manifest** ]]; then
  echo "Replacing manifests and creating new files at dev/deps"
  rm -rf dev/deps
  mv dev/pr-deps dev/deps
  exit 0
fi

for HADOOP_HIVE_PROFILE in "${HADOOP_HIVE_PROFILES[@]}"; do
  set +e
  dep_diff="$(
    git diff \
    --no-index \
    dev/deps/spark-deps-$HADOOP_HIVE_PROFILE \
    dev/pr-deps/spark-deps-$HADOOP_HIVE_PROFILE \
  )"
  set -e
  if [ "$dep_diff" != "" ]; then
    echo "Spark's published dependencies DO NOT MATCH the manifest file (dev/spark-deps)."
    echo "To update the manifest file, run './dev/test-dependencies.sh --replace-manifest'."
    echo "$dep_diff"
    rm -rf dev/pr-deps
    exit 1
  fi
done

if [[ -d "$FWDIR/dev/pr-deps" ]]; then
  rm -rf "$FWDIR/dev/pr-deps"
fi

exit 0