memgraph/tests/public_benchmark/ldbc/build_dataset

#!/bin/bash

# Generate SNB dataset.

function print_help () {
    echo "Usage: $0 [OPTION]"
    echo "Optional arguments:"
    echo -e "  -h|--help -> Prints help."
    echo -e "  --scale-factor Positive_Integer -> Defines the dataset size."
    echo -e "  --neo4j-home Neo4j home directory, overrides NEO4J_HOME"
    echo -e "  --memgraph-home Memgraph home directory."
    echo -e "  --skip-generating Only transform generated dataset"
}

set -e
script_dir="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"

# Read the arguments.
scale_factor=1
skip_generating=false
memgraph_dir="${script_dir}/../../.."
while [[ $# -gt 0 ]]
do
    case $1 in
        -h|--help)
        print_help
        exit 1
        ;;
        --scale-factor)
        scale_factor=$2
        shift
        ;;
        --neo4j-home)
        NEO4J_HOME=$2
        shift
        ;;
        --memgraph-home)
        memgraph_dir=$2
        shift
        ;;
        --skip-generating)
        skip_generating=true
        ;;
        *)
        # unknown option
        ;;
    esac
    shift # past argument or value
done

echo "Using scale_factor" $scale_factor
# Prepare the folder structure.
dataset_folder_prefix="neo4j_csv_dataset"
dataset_folder="${script_dir}/${dataset_folder_prefix}_scale_${scale_factor}"
mkdir -p ${dataset_folder}

# Define scale factor.
cat > ${script_dir}/ldbc_snb_datagen/params.ini <<EOF
ldbc.snb.datagen.generator.scaleFactor:snb.interactive.${scale_factor}
ldbc.snb.datagen.serializer.personSerializer:ldbc.snb.datagen.serializer.snb.interactive.CSVPersonSerializer
ldbc.snb.datagen.serializer.invariantSerializer:ldbc.snb.datagen.serializer.snb.interactive.CSVInvariantSerializer
ldbc.snb.datagen.serializer.personActivitySerializer:ldbc.snb.datagen.serializer.snb.interactive.CSVPersonActivitySerializer
EOF

ldbc_snb_datagen_folder=${script_dir}/ldbc_snb_datagen
if [[ ${skip_generating} = false ]]; then
  # Generate the dataset.
  rm -rf ${dataset_folder}/*
  cd ${ldbc_snb_datagen_folder}
  export HADOOP_OPTS="$HADOOP_OPTS -Xmx20G"
  if [[ -d "/usr/lib/jvm/default-java/jre" ]]; then
    export JAVA_HOME=/usr/lib/jvm/default-java/jre
  elif [[ -d "/usr/lib/jvm/default-runtime/" ]]; then
    export JAVA_HOME=/usr/lib/jvm/default-runtime/
  else
    echo "Unable to find JRE under /usr/lib/jvm"
    exit 1
  fi
  echo "Using JAVA_HOME" $JAVA_HOME
  HADOOP_HOME=/usr/local/hadoop LDBC_SNB_DATAGEN_HOME=${ldbc_snb_datagen_folder} ./run.sh || exit 1

  # Transform the dataset into Neo4j CSV format.
  cd ${script_dir}/ldbc-snb-impls/snb-interactive-neo4j
  mvn exec:java \
      -Dexec.mainClass="net.ellitron.ldbcsnbimpls.interactive.neo4j.util.DataFormatConverter" \
      -Dexec.args="${ldbc_snb_datagen_folder}/social_network ${dataset_folder}" || exit 1
fi

rm -rf ${dataset_folder}/social_network
cp -r ${ldbc_snb_datagen_folder}/social_network ${dataset_folder}/social_network

rm -rf ${dataset_folder}/substitution_parameters
cp -r ${ldbc_snb_datagen_folder}/substitution_parameters ${dataset_folder}/substitution_parameters

csv_dataset="
--nodes ${dataset_folder}/comment_0_0.csv \
--nodes ${dataset_folder}/forum_0_0.csv \
--nodes ${dataset_folder}/organisation_0_0.csv \
--nodes ${dataset_folder}/person_0_0.csv \
--nodes ${dataset_folder}/place_0_0.csv \
--nodes ${dataset_folder}/post_0_0.csv \
--nodes ${dataset_folder}/tag_0_0.csv \
--nodes ${dataset_folder}/tagclass_0_0.csv \
--relationships ${dataset_folder}/comment_hasCreator_person_0_0.csv \
--relationships ${dataset_folder}/comment_hasTag_tag_0_0.csv \
--relationships ${dataset_folder}/comment_isLocatedIn_place_0_0.csv \
--relationships ${dataset_folder}/comment_replyOf_comment_0_0.csv \
--relationships ${dataset_folder}/comment_replyOf_post_0_0.csv \
--relationships ${dataset_folder}/forum_containerOf_post_0_0.csv \
--relationships ${dataset_folder}/forum_hasMember_person_0_0.csv \
--relationships ${dataset_folder}/forum_hasModerator_person_0_0.csv \
--relationships ${dataset_folder}/forum_hasTag_tag_0_0.csv \
--relationships ${dataset_folder}/organisation_isLocatedIn_place_0_0.csv \
--relationships ${dataset_folder}/person_hasInterest_tag_0_0.csv \
--relationships ${dataset_folder}/person_isLocatedIn_place_0_0.csv \
--relationships ${dataset_folder}/person_knows_person_0_0.csv \
--relationships ${dataset_folder}/person_likes_comment_0_0.csv \
--relationships ${dataset_folder}/person_likes_post_0_0.csv \
--relationships ${dataset_folder}/person_studyAt_organisation_0_0.csv \
--relationships ${dataset_folder}/person_workAt_organisation_0_0.csv \
--relationships ${dataset_folder}/place_isPartOf_place_0_0.csv \
--relationships ${dataset_folder}/post_hasCreator_person_0_0.csv \
--relationships ${dataset_folder}/post_hasTag_tag_0_0.csv \
--relationships ${dataset_folder}/post_isLocatedIn_place_0_0.csv \
--relationships ${dataset_folder}/tag_hasType_tagclass_0_0.csv \
--relationships ${dataset_folder}/tagclass_isSubclassOf_tagclass_0_0.csv"

# Convert to neo4j internal format.
if [[ ! -d "${NEO4J_HOME}" ]]; then
  NEO4J_HOME="/usr/share/neo4j"
fi
echo "Using NEO4J_HOME" ${NEO4J_HOME}
mkdir -p ${dataset_folder}/neo4j
cd ${dataset_folder}/neo4j
echo "Converting CSV dataset to '${dataset_folder}/neo4j/graph.db'"
rm -rf graph.db
${NEO4J_HOME}/bin/neo4j-import --into graph.db ${csv_dataset} --delimiter "|" --array-delimiter ";"
# Convert to memgraph internal format.
echo "Using MEMGRAPH_HOME" ${memgraph_dir}
mkdir -p ${dataset_folder}/memgraph
cd ${dataset_folder}/memgraph
echo "Converting CSV dataset to '${dataset_folder}/memgraph/graph.snapshot'"
rm -rf graph.snapshot
${memgraph_dir}/tools/csv_to_snapshot -o graph.snapshot ${csv_dataset} --csv-delimiter "|" --array-delimiter ";"