python/run-tests

#!/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.
#


# Figure out where the Spark framework is installed
FWDIR="$(cd "`dirname "$0"`"; cd ../; pwd)"

. "$FWDIR"/bin/load-spark-env.sh

# CD into the python directory to find things on the right path
cd "$FWDIR/python"

FAILED=0
LOG_FILE=unit-tests.log
START=$(date +"%s")

rm -f $LOG_FILE

# Remove the metastore and warehouse directory created by the HiveContext tests in Spark SQL
rm -rf metastore warehouse

function run_test() {
    echo -en "Running test: $1 ... " | tee -a $LOG_FILE
    start=$(date +"%s")
    SPARK_TESTING=1 time "$FWDIR"/bin/pyspark $1 > $LOG_FILE 2>&1

    FAILED=$((PIPESTATUS[0]||$FAILED))

    # Fail and exit on the first test failure.
    if [[ $FAILED != 0 ]]; then
        cat $LOG_FILE | grep -v "^[0-9][0-9]*" # filter all lines starting with a number.
        echo -en "\033[31m"  # Red
        echo "Had test failures; see logs."
        echo -en "\033[0m"  # No color
        exit -1
    else
        now=$(date +"%s")
        echo "ok ($(($now - $start))s)"
    fi
}

function run_core_tests() {
    echo "Run core tests ..."
    run_test "pyspark.rdd"
    run_test "pyspark.context"
    run_test "pyspark.conf"
    run_test "pyspark.broadcast"
    run_test "pyspark.accumulators"
    run_test "pyspark.serializers"
    run_test "pyspark.profiler"
    run_test "pyspark.shuffle"
    run_test "pyspark.tests"
}

function run_sql_tests() {
    echo "Run sql tests ..."
    run_test "pyspark.sql.types"
    run_test "pyspark.sql.context"
    run_test "pyspark.sql.column"
    run_test "pyspark.sql.dataframe"
    run_test "pyspark.sql.group"
    run_test "pyspark.sql.functions"
    run_test "pyspark.sql.readwriter"
    run_test "pyspark.sql.window"
    run_test "pyspark.sql.tests"
}

function run_mllib_tests() {
    echo "Run mllib tests ..."
    run_test "pyspark.mllib.classification"
    run_test "pyspark.mllib.clustering"
    run_test "pyspark.mllib.evaluation"
    run_test "pyspark.mllib.feature"
    run_test "pyspark.mllib.fpm"
    run_test "pyspark.mllib.linalg"
    run_test "pyspark.mllib.random"
    run_test "pyspark.mllib.recommendation"
    run_test "pyspark.mllib.regression"
    run_test "pyspark.mllib.stat._statistics"
    run_test "pyspark.mllib.tree"
    run_test "pyspark.mllib.util"
    run_test "pyspark.mllib.tests"
}

function run_ml_tests() {
    echo "Run ml tests ..."
    run_test "pyspark.ml.feature"
    run_test "pyspark.ml.classification"
    run_test "pyspark.ml.recommendation"
    run_test "pyspark.ml.regression"
    run_test "pyspark.ml.tuning"
    run_test "pyspark.ml.tests"
    run_test "pyspark.ml.evaluation"
}

function run_streaming_tests() {
    echo "Run streaming tests ..."

    KAFKA_ASSEMBLY_DIR="$FWDIR"/external/kafka-assembly
    JAR_PATH="${KAFKA_ASSEMBLY_DIR}/target/scala-${SPARK_SCALA_VERSION}"
    for f in "${JAR_PATH}"/spark-streaming-kafka-assembly-*.jar; do
      if [[ ! -e "$f" ]]; then
        echo "Failed to find Spark Streaming Kafka assembly jar in $KAFKA_ASSEMBLY_DIR" 1>&2
        echo "You need to build Spark with " \
             "'build/sbt assembly/assembly streaming-kafka-assembly/assembly' or" \
             "'build/mvn package' before running this program" 1>&2
        exit 1
      fi
      KAFKA_ASSEMBLY_JAR="$f"
    done

    export PYSPARK_SUBMIT_ARGS="--jars ${KAFKA_ASSEMBLY_JAR} pyspark-shell"
    run_test "pyspark.streaming.util"
    run_test "pyspark.streaming.tests"
}

echo "Running PySpark tests. Output is in python/$LOG_FILE."

export PYSPARK_PYTHON="python"

# Try to test with Python 2.6, since that's the minimum version that we support:
if [ $(which python2.6) ]; then
    export PYSPARK_PYTHON="python2.6"
fi

echo "Testing with Python version:"
$PYSPARK_PYTHON --version

run_core_tests
run_sql_tests
run_mllib_tests
run_ml_tests
run_streaming_tests

# Try to test with Python 3
if [ $(which python3.4) ]; then
    export PYSPARK_PYTHON="python3.4"
    echo "Testing with Python3.4 version:"
    $PYSPARK_PYTHON --version

    run_core_tests
    run_sql_tests
    run_mllib_tests
    run_ml_tests
    run_streaming_tests
fi

# Try to test with PyPy
if [ $(which pypy) ]; then
    export PYSPARK_PYTHON="pypy"
    echo "Testing with PyPy version:"
    $PYSPARK_PYTHON --version

    run_core_tests
    run_sql_tests
    run_streaming_tests
fi

if [[ $FAILED == 0 ]]; then
    now=$(date +"%s")
    echo -e "\033[32mTests passed \033[0min $(($now - $START)) seconds"
fi

# TODO: in the long-run, it would be nice to use a test runner like `nose`.
# The doctest fixtures are the current barrier to doing this.