Consistency graph workflow merges all the entities by ID
parent
36173c13a5
commit
dfd6205b95
@ -0,0 +1,93 @@
|
||||
|
||||
package eu.dnetlib.dhp.oa.dedup;
|
||||
|
||||
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
||||
|
||||
import java.util.Optional;
|
||||
|
||||
import org.apache.commons.io.IOUtils;
|
||||
import org.apache.commons.lang3.StringUtils;
|
||||
import org.apache.spark.SparkConf;
|
||||
import org.apache.spark.api.java.function.FilterFunction;
|
||||
import org.apache.spark.api.java.function.MapFunction;
|
||||
import org.apache.spark.sql.Encoders;
|
||||
import org.apache.spark.sql.SaveMode;
|
||||
import org.apache.spark.sql.SparkSession;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
|
||||
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
||||
import eu.dnetlib.dhp.common.HdfsSupport;
|
||||
import eu.dnetlib.dhp.schema.oaf.Oaf;
|
||||
import eu.dnetlib.dhp.schema.oaf.OafEntity;
|
||||
|
||||
public class DispatchEntitiesSparkJob {
|
||||
|
||||
private static final Logger log = LoggerFactory.getLogger(DispatchEntitiesSparkJob.class);
|
||||
|
||||
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
||||
|
||||
public static void main(String[] args) throws Exception {
|
||||
|
||||
String jsonConfiguration = IOUtils
|
||||
.toString(
|
||||
DispatchEntitiesSparkJob.class
|
||||
.getResourceAsStream(
|
||||
"/eu/dnetlib/dhp/oa/dedup/dispatch_entities_parameters.json"));
|
||||
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
||||
parser.parseArgument(args);
|
||||
|
||||
Boolean isSparkSessionManaged = Optional
|
||||
.ofNullable(parser.get("isSparkSessionManaged"))
|
||||
.map(Boolean::valueOf)
|
||||
.orElse(Boolean.TRUE);
|
||||
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
|
||||
|
||||
String inputPath = parser.get("inputPath");
|
||||
log.info("inputPath: {}", inputPath);
|
||||
|
||||
String outputPath = parser.get("outputPath");
|
||||
log.info("outputPath: {}", outputPath);
|
||||
|
||||
String graphTableClassName = parser.get("graphTableClassName");
|
||||
log.info("graphTableClassName: {}", graphTableClassName);
|
||||
|
||||
Class<? extends OafEntity> entityClazz = (Class<? extends OafEntity>) Class.forName(graphTableClassName);
|
||||
|
||||
SparkConf conf = new SparkConf();
|
||||
runWithSparkSession(
|
||||
conf,
|
||||
isSparkSessionManaged,
|
||||
spark -> {
|
||||
HdfsSupport.remove(outputPath, spark.sparkContext().hadoopConfiguration());
|
||||
dispatchEntities(spark, inputPath, entityClazz, outputPath);
|
||||
});
|
||||
}
|
||||
|
||||
private static <T extends Oaf> void dispatchEntities(
|
||||
SparkSession spark,
|
||||
String inputPath,
|
||||
Class<T> clazz,
|
||||
String outputPath) {
|
||||
|
||||
spark
|
||||
.read()
|
||||
.textFile(inputPath)
|
||||
.filter((FilterFunction<String>) s -> isEntityType(s, clazz))
|
||||
.map((MapFunction<String, String>) s -> StringUtils.substringAfter(s, "|"), Encoders.STRING())
|
||||
.map(
|
||||
(MapFunction<String, T>) value -> OBJECT_MAPPER.readValue(value, clazz),
|
||||
Encoders.bean(clazz))
|
||||
.write()
|
||||
.mode(SaveMode.Overwrite)
|
||||
.option("compression", "gzip")
|
||||
.json(outputPath);
|
||||
}
|
||||
|
||||
private static <T extends Oaf> boolean isEntityType(final String s, final Class<T> clazz) {
|
||||
return StringUtils.substringBefore(s, "|").equals(clazz.getName());
|
||||
}
|
||||
|
||||
}
|
||||
@ -0,0 +1,202 @@
|
||||
|
||||
package eu.dnetlib.dhp.oa.dedup;
|
||||
|
||||
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
||||
import static eu.dnetlib.dhp.utils.DHPUtils.toSeq;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.util.List;
|
||||
import java.util.Objects;
|
||||
import java.util.Optional;
|
||||
import java.util.stream.Collectors;
|
||||
|
||||
import org.apache.commons.io.IOUtils;
|
||||
import org.apache.commons.lang3.StringUtils;
|
||||
import org.apache.spark.SparkConf;
|
||||
import org.apache.spark.api.java.JavaSparkContext;
|
||||
import org.apache.spark.api.java.function.FilterFunction;
|
||||
import org.apache.spark.api.java.function.MapFunction;
|
||||
import org.apache.spark.sql.*;
|
||||
import org.apache.spark.sql.expressions.Aggregator;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import com.jayway.jsonpath.Configuration;
|
||||
import com.jayway.jsonpath.DocumentContext;
|
||||
import com.jayway.jsonpath.JsonPath;
|
||||
import com.jayway.jsonpath.Option;
|
||||
|
||||
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
||||
import eu.dnetlib.dhp.common.HdfsSupport;
|
||||
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
||||
import eu.dnetlib.dhp.schema.oaf.*;
|
||||
import scala.Tuple2;
|
||||
|
||||
/**
|
||||
* Groups the graph content by entity identifier to ensure ID uniqueness
|
||||
*/
|
||||
public class GroupEntitiesSparkJob {
|
||||
|
||||
private static final Logger log = LoggerFactory.getLogger(GroupEntitiesSparkJob.class);
|
||||
|
||||
private final static String ID_JPATH = "$.id";
|
||||
|
||||
private final static String SOURCE_JPATH = "$.source";
|
||||
|
||||
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
||||
|
||||
public static void main(String[] args) throws Exception {
|
||||
|
||||
String jsonConfiguration = IOUtils
|
||||
.toString(
|
||||
GroupEntitiesSparkJob.class
|
||||
.getResourceAsStream(
|
||||
"/eu/dnetlib/dhp/oa/dedup/group_graph_entities_parameters.json"));
|
||||
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
||||
parser.parseArgument(args);
|
||||
|
||||
Boolean isSparkSessionManaged = Optional
|
||||
.ofNullable(parser.get("isSparkSessionManaged"))
|
||||
.map(Boolean::valueOf)
|
||||
.orElse(Boolean.TRUE);
|
||||
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
|
||||
|
||||
String graphInputPath = parser.get("graphInputPath");
|
||||
log.info("graphInputPath: {}", graphInputPath);
|
||||
|
||||
String outputPath = parser.get("outputPath");
|
||||
log.info("outputPath: {}", outputPath);
|
||||
|
||||
SparkConf conf = new SparkConf();
|
||||
conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer");
|
||||
conf.registerKryoClasses(ModelSupport.getOafModelClasses());
|
||||
|
||||
runWithSparkSession(
|
||||
conf,
|
||||
isSparkSessionManaged,
|
||||
spark -> {
|
||||
HdfsSupport.remove(outputPath, spark.sparkContext().hadoopConfiguration());
|
||||
groupEntities(spark, graphInputPath, outputPath);
|
||||
});
|
||||
}
|
||||
|
||||
private static void groupEntities(
|
||||
SparkSession spark,
|
||||
String inputPath,
|
||||
String outputPath) {
|
||||
|
||||
final TypedColumn<OafEntity, OafEntity> aggregator = new GroupingAggregator().toColumn();
|
||||
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
||||
spark
|
||||
.read()
|
||||
.textFile(toSeq(listPaths(inputPath, sc)))
|
||||
.map((MapFunction<String, OafEntity>) s -> parseOaf(s), Encoders.kryo(OafEntity.class))
|
||||
.filter((FilterFunction<OafEntity>) e -> StringUtils.isNotBlank(ModelSupport.idFn().apply(e)))
|
||||
.groupByKey((MapFunction<OafEntity, String>) oaf -> ModelSupport.idFn().apply(oaf), Encoders.STRING())
|
||||
.agg(aggregator)
|
||||
.map(
|
||||
(MapFunction<Tuple2<String, OafEntity>, String>) t -> t._2().getClass().getName() +
|
||||
"|" + OBJECT_MAPPER.writeValueAsString(t._2()),
|
||||
Encoders.STRING())
|
||||
.write()
|
||||
.option("compression", "gzip")
|
||||
.mode(SaveMode.Overwrite)
|
||||
.text(outputPath);
|
||||
}
|
||||
|
||||
public static class GroupingAggregator extends Aggregator<OafEntity, OafEntity, OafEntity> {
|
||||
|
||||
@Override
|
||||
public OafEntity zero() {
|
||||
return null;
|
||||
}
|
||||
|
||||
@Override
|
||||
public OafEntity reduce(OafEntity b, OafEntity a) {
|
||||
return mergeAndGet(b, a);
|
||||
}
|
||||
|
||||
private OafEntity mergeAndGet(OafEntity b, OafEntity a) {
|
||||
if (Objects.nonNull(a) && Objects.nonNull(b)) {
|
||||
return OafMapperUtils.mergeEntities(b, a);
|
||||
}
|
||||
return Objects.isNull(a) ? b : a;
|
||||
}
|
||||
|
||||
@Override
|
||||
public OafEntity merge(OafEntity b, OafEntity a) {
|
||||
return mergeAndGet(b, a);
|
||||
}
|
||||
|
||||
@Override
|
||||
public OafEntity finish(OafEntity j) {
|
||||
return j;
|
||||
}
|
||||
|
||||
@Override
|
||||
public Encoder<OafEntity> bufferEncoder() {
|
||||
return Encoders.kryo(OafEntity.class);
|
||||
}
|
||||
|
||||
@Override
|
||||
public Encoder<OafEntity> outputEncoder() {
|
||||
return Encoders.kryo(OafEntity.class);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
private static OafEntity parseOaf(String s) {
|
||||
|
||||
DocumentContext dc = JsonPath
|
||||
.parse(s, Configuration.defaultConfiguration().addOptions(Option.SUPPRESS_EXCEPTIONS));
|
||||
final String id = dc.read(ID_JPATH);
|
||||
if (StringUtils.isNotBlank(id)) {
|
||||
|
||||
String prefix = StringUtils.substringBefore(id, "|");
|
||||
switch (prefix) {
|
||||
case "10":
|
||||
return parse(s, Datasource.class);
|
||||
case "20":
|
||||
return parse(s, Organization.class);
|
||||
case "40":
|
||||
return parse(s, Project.class);
|
||||
case "50":
|
||||
String resultType = dc.read("$.resulttype.classid");
|
||||
switch (resultType) {
|
||||
case "publication":
|
||||
return parse(s, Publication.class);
|
||||
case "dataset":
|
||||
return parse(s, eu.dnetlib.dhp.schema.oaf.Dataset.class);
|
||||
case "software":
|
||||
return parse(s, Software.class);
|
||||
case "other":
|
||||
return parse(s, OtherResearchProduct.class);
|
||||
default:
|
||||
throw new IllegalArgumentException(String.format("invalid resultType: '%s'", resultType));
|
||||
}
|
||||
default:
|
||||
throw new IllegalArgumentException(String.format("invalid id prefix: '%s'", prefix));
|
||||
}
|
||||
} else {
|
||||
throw new IllegalArgumentException(String.format("invalid oaf: '%s'", s));
|
||||
}
|
||||
}
|
||||
|
||||
private static <T extends OafEntity> OafEntity parse(String s, Class<T> clazz) {
|
||||
try {
|
||||
return OBJECT_MAPPER.readValue(s, clazz);
|
||||
} catch (IOException e) {
|
||||
throw new RuntimeException(e);
|
||||
}
|
||||
}
|
||||
|
||||
private static List<String> listPaths(String inputPath, JavaSparkContext sc) {
|
||||
return HdfsSupport
|
||||
.listFiles(inputPath, sc.hadoopConfiguration())
|
||||
.stream()
|
||||
.filter(f -> !f.equals("relation"))
|
||||
.collect(Collectors.toList());
|
||||
}
|
||||
|
||||
}
|
||||
@ -0,0 +1,26 @@
|
||||
[
|
||||
{
|
||||
"paramName": "issm",
|
||||
"paramLongName": "isSparkSessionManaged",
|
||||
"paramDescription": "when true will stop SparkSession after job execution",
|
||||
"paramRequired": false
|
||||
},
|
||||
{
|
||||
"paramName": "i",
|
||||
"paramLongName": "inputPath",
|
||||
"paramDescription": "the source path",
|
||||
"paramRequired": true
|
||||
},
|
||||
{
|
||||
"paramName": "o",
|
||||
"paramLongName": "outputPath",
|
||||
"paramDescription": "path of the output graph",
|
||||
"paramRequired": true
|
||||
},
|
||||
{
|
||||
"paramName": "c",
|
||||
"paramLongName": "graphTableClassName",
|
||||
"paramDescription": "the graph entity class name",
|
||||
"paramRequired": true
|
||||
}
|
||||
]
|
||||
Loading…
Reference in New Issue