WIP: logic for promoting action sets added

This commit is contained in:
Przemysław Jacewicz 2020-02-17 18:18:24 +01:00
parent bea1a94346
commit 958f0693d6
6 changed files with 477 additions and 78 deletions

View File

@ -5,7 +5,7 @@
<parent> <parent>
<groupId>eu.dnetlib.dhp</groupId> <groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-workflows</artifactId> <artifactId>dhp-workflows</artifactId>
<version>1.0.5-SNAPSHOT</version> <version>1.1.6-SNAPSHOT</version>
</parent> </parent>
<artifactId>dhp-actionmanager</artifactId> <artifactId>dhp-actionmanager</artifactId>
@ -52,11 +52,34 @@
<version>2.25.0</version> <version>2.25.0</version>
<scope>test</scope> <scope>test</scope>
</dependency> </dependency>
<dependency> <dependency>
<groupId>eu.dnetlib.dhp</groupId> <groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-schemas</artifactId> <artifactId>dhp-schemas</artifactId>
<version>1.0.5-SNAPSHOT</version> <version>${project.version}</version>
<scope>compile</scope> </dependency>
<dependency>
<groupId>eu.dnetlib</groupId>
<artifactId>dnet-actionmanager-common</artifactId>
<version>[6.0.0, 7.0.0)</version>
<exclusions>
<!-- duplicate with different groupId -->
<exclusion>
<groupId>apache</groupId>
<artifactId>commons-logging</artifactId>
</exclusion>
<exclusion>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
</exclusion>
</exclusions>
</dependency>
<dependency>
<groupId>eu.dnetlib</groupId>
<artifactId>dnet-openaire-data-protos</artifactId>
<version>[3.0.0, 4.0.0)</version>
</dependency> </dependency>
</dependencies> </dependencies>
</project> </project>

View File

@ -0,0 +1,45 @@
package eu.dnetlib.dhp.actionmanager;
import eu.dnetlib.dhp.schema.oaf.OafEntity;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.api.java.function.ReduceFunction;
import org.apache.spark.sql.Column;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import scala.Tuple2;
import java.util.Optional;
import java.util.function.BiFunction;
public class PromoteActionSetFromHDFSFunctions {
public static <T extends OafEntity> Dataset<T> groupEntitiesByIdAndMerge(Dataset<T> entityDS,
Class<T> clazz) {
return entityDS
.groupByKey((MapFunction<T, String>) OafEntity::getId, Encoders.STRING())
.reduceGroups((ReduceFunction<T>) (x1, x2) -> {
x1.mergeFrom(x2);
return x1;
})
.map((MapFunction<Tuple2<String, T>, T>) pair -> pair._2, Encoders.bean(clazz));
}
public static <T extends OafEntity, S> Dataset<T> joinEntitiesWithActionPayloadAndMerge(Dataset<T> entityDS,
Dataset<S> actionPayloadDS,
BiFunction<Dataset<T>, Dataset<S>, Column> entityToActionPayloadJoinExpr,
BiFunction<S, Class<T>, T> actionPayloadToEntityFn,
Class<T> clazz) {
return entityDS
.joinWith(actionPayloadDS, entityToActionPayloadJoinExpr.apply(entityDS, actionPayloadDS), "left_outer")
.map((MapFunction<Tuple2<T, S>, T>) pair -> Optional
.ofNullable(pair._2())
.map(x -> {
T entity = actionPayloadToEntityFn.apply(x, clazz);
pair._1().mergeFrom(entity);
return pair._1();
})
.orElse(pair._1()), Encoders.bean(clazz));
}
}

View File

@ -1,24 +1,24 @@
package eu.dnetlib.dhp.actionmanager; package eu.dnetlib.dhp.actionmanager;
import com.fasterxml.jackson.databind.JsonNode;
import com.fasterxml.jackson.databind.ObjectMapper; import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.data.proto.OafProtos;
import eu.dnetlib.dhp.application.ArgumentApplicationParser; import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.OafEntity; import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.dhp.schema.oaf.Software;
import org.apache.commons.io.IOUtils; import org.apache.commons.io.IOUtils;
import org.apache.hadoop.io.Text; import org.apache.hadoop.io.Text;
import org.apache.spark.SparkConf; import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.api.java.function.MapFunction; import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.api.java.function.ReduceFunction;
import org.apache.spark.sql.*; import org.apache.spark.sql.*;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.types.*; import org.apache.spark.sql.types.*;
import scala.Tuple2;
import java.io.IOException; import java.io.IOException;
import java.util.Collections; import java.util.Arrays;
import java.util.List;
import static eu.dnetlib.dhp.actionmanager.PromoteActionSetFromHDFSFunctions.*;
import static org.apache.spark.sql.functions.*; import static org.apache.spark.sql.functions.*;
public class PromoteActionSetFromHDFSJob { public class PromoteActionSetFromHDFSJob {
@ -28,73 +28,209 @@ public class PromoteActionSetFromHDFSJob {
PromoteActionSetFromHDFSJob.class PromoteActionSetFromHDFSJob.class
.getResourceAsStream("/eu/dnetlib/dhp/actionmanager/actionmanager_input_parameters.json"))); .getResourceAsStream("/eu/dnetlib/dhp/actionmanager/actionmanager_input_parameters.json")));
parser.parseArgument(args); parser.parseArgument(args);
String inputActionSetPath = parser.get("input");
String outputPath = parser.get("output");
final SparkConf conf = new SparkConf(); String inputGraphPath = parser.get("inputGraphPath");
List<String> inputActionSetPaths = Arrays.asList(parser.get("inputActionSetPaths").split(","));
String outputGraphPath = parser.get("outputGraphPath");
SparkConf conf = new SparkConf();
conf.setMaster(parser.get("master")); conf.setMaster(parser.get("master"));
conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer"); conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer");
try (SparkSession spark = SparkSession.builder().config(conf).getOrCreate()) { try (SparkSession spark = SparkSession.builder().config(conf).getOrCreate()) {
// reading actions as RDD
JavaRDD<Row> actionsRDD = JavaSparkContext
.fromSparkContext(spark.sparkContext())
.sequenceFile(inputActionSetPath, Text.class, Text.class)
.map(x -> RowFactory.create(x._2().toString()));
// converting actions to DataFrame and deserializing content of TargetValue // ----- READ -----
// using unbase64 on TargetValue content to get String representation // dataset
StructType rowSchema = StructType$.MODULE$.apply( Dataset<eu.dnetlib.dhp.schema.oaf.Dataset> datasetDS = readGraphTable(
Collections.singletonList( spark, String.format("%s/dataset", inputGraphPath), eu.dnetlib.dhp.schema.oaf.Dataset.class)
StructField$.MODULE$.apply("value", DataTypes.StringType, false, Metadata.empty())
));
Dataset<Row> deserializedTargetValue = spark.createDataFrame(actionsRDD, rowSchema)
.withColumn("TargetValue", get_json_object(col("value"), "$.TargetValue"))
.select(unbase64(col("TargetValue")).cast(DataTypes.StringType).as("target_value_json"))
.cache(); .cache();
datasetDS.printSchema();
datasetDS.show();
// printing: only for testing // datasource
deserializedTargetValue.printSchema(); Dataset<Datasource> datasourceDS =
deserializedTargetValue.show(); readGraphTable(spark, String.format("%s/datasource", inputGraphPath), Datasource.class)
System.out.println(deserializedTargetValue.first().toString()); .cache();
datasourceDS.printSchema();
datasourceDS.show();
// grouping and merging: should be generic // organization
Dataset<Software> softwareDS = deserializedTargetValue Dataset<Organization> organizationDS =
.map((MapFunction<Row, Software>) PromoteActionSetFromHDFSJob::rowToOafEntity, Encoders.kryo(Software.class)) readGraphTable(spark, String.format("%s/organization", inputGraphPath), Organization.class)
.groupByKey((MapFunction<Software, String>) OafEntity::getId, Encoders.STRING()) .cache();
.reduceGroups((ReduceFunction<Software>) (software1, software2) -> { organizationDS.printSchema();
software1.mergeFrom(software2); organizationDS.show();
return software1;
})
.map((MapFunction<Tuple2<String, Software>, Software>) pair -> pair._2, Encoders.kryo(Software.class));
// otherresearchproduct
Dataset<OtherResearchProduct> otherResearchProductDS =
readGraphTable(spark, String.format("%s/otherresearchproduct", inputGraphPath), OtherResearchProduct.class)
.cache();
otherResearchProductDS.printSchema();
otherResearchProductDS.show();
// project
Dataset<Project> projectDS =
readGraphTable(spark, String.format("%s/project", inputGraphPath), Project.class)
.cache();
projectDS.printSchema();
projectDS.show();
// publication
Dataset<Publication> publicationDS =
readGraphTable(spark, String.format("%s/publication", inputGraphPath), Publication.class)
.cache();
publicationDS.printSchema();
publicationDS.show();
// relation
Dataset<Relation> relationDS =
readGraphTable(spark, String.format("%s/relation", inputGraphPath), Relation.class)
.cache();
relationDS.printSchema();
relationDS.show();
// software
Dataset<Software> softwareDS =
readGraphTable(spark, String.format("%s/software", inputGraphPath), Software.class)
.cache();
softwareDS.printSchema(); softwareDS.printSchema();
softwareDS.show(); softwareDS.show();
// save // actions
// softwareDS.toDF() Dataset<String> actionPayloadDS = inputActionSetPaths.stream()
// .write() .map(inputActionSetPath -> readActionSetPayload(spark, inputActionSetPath))
// .partitionBy("id") .reduce(Dataset::union)
// .save(outputPath); .get()
.cache();
actionPayloadDS.printSchema();
actionPayloadDS.show();
System.out.println(String.join("\n", actionPayloadDS.takeAsList(20)));
// another approach: using only DataFrames i.e. DataSet<Row>, not DataSets<Software> Dataset<String> relationActionPayloadDS = filterActionPayloadForRelations(actionPayloadDS)
.cache();
relationActionPayloadDS.printSchema();
relationActionPayloadDS.show();
Dataset<String> entityActionPayloadDS = filterActionPayloadForEntity(actionPayloadDS)
.cache();
entityActionPayloadDS.printSchema();
entityActionPayloadDS.show();
// ----- LOGIC -----
Dataset<eu.dnetlib.dhp.schema.oaf.Dataset> processedDatasetDS =
processEntityDS(datasetDS, entityActionPayloadDS, eu.dnetlib.dhp.schema.oaf.Dataset.class);
Dataset<Datasource> processedDatasourceDS =
processEntityDS(datasourceDS, entityActionPayloadDS, Datasource.class);
Dataset<Organization> processedOrganizationDS =
processEntityDS(organizationDS, entityActionPayloadDS, Organization.class);
Dataset<OtherResearchProduct> processedOtherResearchProductDS =
processEntityDS(otherResearchProductDS, entityActionPayloadDS, OtherResearchProduct.class);
Dataset<Project> processedProjectDS =
processEntityDS(projectDS, entityActionPayloadDS, Project.class);
Dataset<Publication> processedPublicationDS =
processEntityDS(publicationDS, entityActionPayloadDS, Publication.class);
Dataset<Relation> processedRelationDS =
processRelationDS(relationDS, relationActionPayloadDS);
Dataset<Software> processedSoftwareDS =
processEntityDS(softwareDS, entityActionPayloadDS, Software.class);
// ----- SAVE -----
processedDatasetDS.write()
.save(String.format("%s/dataset", outputGraphPath));
processedDatasourceDS.write()
.save(String.format("%s/datasource", outputGraphPath));
processedOrganizationDS.write()
.save(String.format("%s/organization", outputGraphPath));
processedOtherResearchProductDS.write()
.save(String.format("%s/otherresearchproduct", outputGraphPath));
processedProjectDS.write()
.save(String.format("%s/project", outputGraphPath));
processedPublicationDS.write()
.save(String.format("%s/publication", outputGraphPath));
processedRelationDS.write()
.save(String.format("%s/relation", outputGraphPath));
processedSoftwareDS.write()
.save(String.format("%s/software", outputGraphPath));
} }
} }
private static Software rowToOafEntity(Row row) { private static final StructType KV_SCHEMA = StructType$.MODULE$.apply(
// converts row with JSON into Software object: should be generic Arrays.asList(
// currently extracts only "entity.id" field from JSON StructField$.MODULE$.apply("key", DataTypes.StringType, false, Metadata.empty()),
ObjectMapper objectMapper = new ObjectMapper(); StructField$.MODULE$.apply("value", DataTypes.StringType, false, Metadata.empty())
));
private static <T extends Oaf> Dataset<T> readGraphTable(SparkSession spark, String path, Class<T> clazz) {
JavaRDD<Row> rows = JavaSparkContext
.fromSparkContext(spark.sparkContext())
.sequenceFile(path, Text.class, Text.class)
.map(x -> RowFactory.create(x._1().toString(), x._2().toString()));
return spark.createDataFrame(rows, KV_SCHEMA)
.map((MapFunction<Row, T>) row -> new ObjectMapper().readValue(row.<String>getAs("value"), clazz),
Encoders.bean(clazz));
}
private static Dataset<String> readActionSetPayload(SparkSession spark, String inputActionSetPath) {
JavaRDD<Row> actionsRDD = JavaSparkContext
.fromSparkContext(spark.sparkContext())
.sequenceFile(inputActionSetPath, Text.class, Text.class)
.map(x -> RowFactory.create(x._1().toString(), x._2().toString()));
return spark.createDataFrame(actionsRDD, KV_SCHEMA)
.select(unbase64(get_json_object(col("value"), "$.TargetValue"))
.cast(DataTypes.StringType).as("target_value_json"))
.as(Encoders.STRING());
}
private static Dataset<String> filterActionPayloadForRelations(Dataset<String> actionPayloadDS) {
return actionPayloadDS
.where(get_json_object(col("target_value_json"), "$.kind").equalTo("relation"));
}
private static Dataset<String> filterActionPayloadForEntity(Dataset<String> actionPayloadDS) {
return actionPayloadDS
.where(get_json_object(col("target_value_json"), "$.kind").equalTo("entity"));
}
private static <T extends OafEntity> Dataset<T> processEntityDS(Dataset<T> entityDS,
Dataset<String> actionPayloadDS,
Class<T> clazz) {
Dataset<T> groupedAndMerged = groupEntitiesByIdAndMerge(entityDS, clazz);
Dataset<T> joinedAndMerged = joinEntitiesWithActionPayloadAndMerge(groupedAndMerged,
actionPayloadDS,
PromoteActionSetFromHDFSJob::entityToActionPayloadJoinExpr,
PromoteActionSetFromHDFSJob::actionPayloadToEntity,
clazz);
return groupEntitiesByIdAndMerge(joinedAndMerged, clazz);
}
private static <T extends OafEntity> Column entityToActionPayloadJoinExpr(Dataset<T> left,
Dataset<String> right) {
return left.col("id").equalTo(
get_json_object(right.col("target_value_json"), "$.entity.id"));
}
public static <T extends OafEntity> T actionPayloadToEntity(String actionPayload,
Class<T> clazz) {
try { try {
JsonNode jsonNode = objectMapper.readTree(row.getString(0)); OafProtos.Oaf oldEntity = new ObjectMapper().readValue(actionPayload, OafProtos.Oaf.class);
String id = jsonNode.at("/entity/id").asText(); return entityOldToNew(oldEntity, clazz);
Software software = new Software();
software.setId(id);
return software;
} catch (IOException e) { } catch (IOException e) {
e.printStackTrace();
throw new RuntimeException(e); throw new RuntimeException(e);
} }
} }
//TODO
private static <T extends OafEntity> T entityOldToNew(OafProtos.Oaf old,
Class<T> clazz) {
return null;
}
//TODO
private static Dataset<Relation> processRelationDS(Dataset<Relation> relationDS,
Dataset<String> actionPayloadDS) {
return null;
}
} }

View File

@ -6,15 +6,21 @@
"paramRequired": true "paramRequired": true
}, },
{ {
"paramName": "i", "paramName": "ig",
"paramLongName": "input", "paramLongName": "inputGraphPath",
"paramDescription": "the path of the input sequential file to read", "paramDescription": "#TODO: input graph path",
"paramRequired": true "paramRequired": true
}, },
{ {
"paramName": "o", "paramName": "ia",
"paramLongName": "output", "paramLongName": "inputActionSetPaths",
"paramDescription": "the path of the result DataFrame on HDFS", "paramDescription": "#TODO: comma separated list of paths to input action sets",
"paramRequired": true
},
{
"paramName": "og",
"paramLongName": "outputGraphPath",
"paramDescription": "#TODO: the path of the result DataFrame on HDFS",
"paramRequired": true "paramRequired": true
} }
] ]

View File

@ -0,0 +1,169 @@
package eu.dnetlib.dhp.actionmanager;
import com.fasterxml.jackson.databind.JsonNode;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.schema.oaf.OafEntity;
import org.apache.spark.SparkConf;
import org.apache.spark.sql.Column;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import org.junit.AfterClass;
import org.junit.BeforeClass;
import org.junit.Test;
import java.io.IOException;
import java.io.Serializable;
import java.util.Arrays;
import java.util.List;
import java.util.function.BiFunction;
import java.util.stream.Collectors;
import static org.apache.spark.sql.functions.get_json_object;
import static org.junit.Assert.assertEquals;
public class PromoteActionSetFromHDFSFunctionsTest {
private static SparkSession spark;
@BeforeClass
public static void beforeClass() {
SparkConf conf = new SparkConf();
conf.setMaster("local");
conf.setAppName(PromoteActionSetFromHDFSFunctionsTest.class.getSimpleName());
conf.set("spark.driver.host", "localhost");
spark = SparkSession.builder().config(conf).getOrCreate();
}
@AfterClass
public static void afterClass() {
spark.stop();
}
@Test
public void shouldGroupOafEntitiesByIdAndMergeWithinGroup() {
// given
String id1 = "id1";
String id2 = "id2";
String id3 = "id3";
List<OafEntityImpl> entityData = Arrays.asList(
createOafEntityImpl(id1),
createOafEntityImpl(id2), createOafEntityImpl(id2),
createOafEntityImpl(id3), createOafEntityImpl(id3), createOafEntityImpl(id3)
);
Dataset<OafEntityImpl> entityDS = spark.createDataset(entityData, Encoders.bean(OafEntityImpl.class));
// when
List<OafEntityImpl> results = PromoteActionSetFromHDFSFunctions
.groupEntitiesByIdAndMerge(entityDS, OafEntityImpl.class)
.collectAsList();
System.out.println(results.stream().map(x -> String.format("%s:%d", x.getId(), x.merged)).collect(Collectors.joining(",")));
// then
assertEquals(3, results.size());
results.forEach(result -> {
switch (result.getId()) {
case "id1":
assertEquals(1, result.merged);
break;
case "id2":
assertEquals(2, result.merged);
break;
case "id3":
assertEquals(3, result.merged);
break;
}
});
}
@Test
public void shouldJoinWithActionPayloadUsingIdAndMerge() {
// given
String id1 = "id1";
String id2 = "id2";
String id3 = "id3";
String id4 = "id4";
List<OafEntityImpl> entityData = Arrays.asList(
createOafEntityImpl(id1), createOafEntityImpl(id2), createOafEntityImpl(id3), createOafEntityImpl(id4)
);
Dataset<OafEntityImpl> entityDS = spark.createDataset(entityData, Encoders.bean(OafEntityImpl.class));
List<String> actionPayloadData = Arrays.asList(
actionPayload(id1),
actionPayload(id2), actionPayload(id2),
actionPayload(id3), actionPayload(id3), actionPayload(id3)
);
Dataset<String> actionPayloadDS = spark.createDataset(actionPayloadData, Encoders.STRING());
BiFunction<Dataset<OafEntityImpl>, Dataset<String>, Column> entityToActionPayloadJoinExpr = (left, right) ->
left.col("id").equalTo(get_json_object(right.col("value"), "$.id"));
BiFunction<String, Class<OafEntityImpl>, OafEntityImpl> actionPayloadToEntityFn =
(BiFunction<String, Class<OafEntityImpl>, OafEntityImpl> & Serializable) (s, clazz) -> {
try {
JsonNode jsonNode = new ObjectMapper().readTree(s);
String id = jsonNode.at("/id").asText();
OafEntityImpl x = new OafEntityImpl();
x.setId(id);
return x;
} catch (IOException e) {
throw new RuntimeException(e);
}
};
// when
List<OafEntityImpl> results = PromoteActionSetFromHDFSFunctions
.joinEntitiesWithActionPayloadAndMerge(entityDS,
actionPayloadDS,
entityToActionPayloadJoinExpr,
actionPayloadToEntityFn,
OafEntityImpl.class)
.collectAsList();
System.out.println(results.stream().map(x -> String.format("%s:%d", x.getId(), x.merged)).collect(Collectors.joining(",")));
// then
assertEquals(7, results.size());
results.forEach(result -> {
switch (result.getId()) {
case "id1":
assertEquals(2, result.merged);
break;
case "id2":
assertEquals(2, result.merged);
break;
case "id3":
assertEquals(2, result.merged);
break;
case "id4":
assertEquals(1, result.merged);
break;
}
});
}
public static class OafEntityImpl extends OafEntity {
private int merged = 1;
@Override
public void mergeFrom(OafEntity e) {
merged += ((OafEntityImpl) e).merged;
}
public int getMerged() {
return merged;
}
public void setMerged(int merged) {
this.merged = merged;
}
}
private static OafEntityImpl createOafEntityImpl(String id) {
OafEntityImpl x = new OafEntityImpl();
x.setId(id);
return x;
}
private static String actionPayload(String id) {
return String.format("{\"id\":\"%s\"}", id);
}
}

View File

@ -1,7 +1,5 @@
package eu.dnetlib.dhp.actionmanager; package eu.dnetlib.dhp.actionmanager;
import org.apache.commons.io.FileUtils;
import org.junit.After;
import org.junit.Before; import org.junit.Before;
import org.junit.Test; import org.junit.Test;
@ -14,42 +12,64 @@ import java.util.Objects;
public class PromoteActionSetFromHDFSJobTest { public class PromoteActionSetFromHDFSJobTest {
private ClassLoader cl = getClass().getClassLoader(); private ClassLoader cl = getClass().getClassLoader();
private Path workingDir; private Path workingDir;
private Path inputActionSetDir; private Path inputDir;
private Path outputDir; private Path outputDir;
@Before @Before
public void before() throws IOException { public void before() throws IOException {
workingDir = Files.createTempDirectory("promote_action_set"); workingDir = Files.createTempDirectory("promote_action_set");
inputActionSetDir = workingDir.resolve("input"); inputDir = workingDir.resolve("input");
outputDir = workingDir.resolve("output"); outputDir = workingDir.resolve("output");
} }
@After // @After
public void after() throws IOException { // public void after() throws IOException {
FileUtils.deleteDirectory(workingDir.toFile()); // FileUtils.deleteDirectory(workingDir.toFile());
} // }
@Test @Test
public void shouldReadAtomicActionsFromHDFSAndWritePartitionedAsParquetFiles() throws Exception { public void shouldReadAtomicActionsFromHDFSAndWritePartitionedAsParquetFiles() throws Exception {
// given // given
// NOTE: test resource should contain atomic actions in a human readable form, probably as json files; here the // NOTE: test resource should contain atomic actions in a human readable form, probably as json files; here the
// files should be converted to a serialized format and written out to workingDir/input // files should be converted to a serialized format and written out to workingDir/input
// for current testing: actions from software export, given as sequence file are copied to workingDir/input/ // for current testing: actions from iis export, given as sequence file are copied to workingDir/input/
Path exportedActionSetDir = Paths.get(Objects.requireNonNull(cl.getResource("entities/entities_software")).getFile());
Path inputDir = inputActionSetDir.resolve("entities_software"); //graph
Files.createDirectories(inputDir); Path inputGraphDir = inputDir.resolve("graph");
copyFiles(exportedActionSetDir, inputDir); Files.createDirectories(inputGraphDir);
copyFiles(Paths.get(Objects.requireNonNull(cl.getResource("graph")).getFile()), inputGraphDir);
//actions
Path inputActionsDir = inputDir.resolve("actions");
Files.createDirectories(inputActionsDir);
Path inputEntitiesPatentDir = inputActionsDir.resolve("entities_patent");
Files.createDirectories(inputEntitiesPatentDir);
copyFiles(Paths.get(Objects.requireNonNull(cl.getResource("actions/entities_patent")).getFile()), inputEntitiesPatentDir);
Path inputEntitiesSoftwareDir = inputActionsDir.resolve("entities_software");
Files.createDirectories(inputEntitiesSoftwareDir);
copyFiles(Paths.get(Objects.requireNonNull(cl.getResource("actions/entities_software")).getFile()), inputEntitiesSoftwareDir);
String inputActionSetPaths = String.join(",", inputEntitiesSoftwareDir.toString()); //inputEntitiesPatentDir.toString(),
PromoteActionSetFromHDFSJob.main(new String[]{ PromoteActionSetFromHDFSJob.main(new String[]{
"-mt", "local[*]", "-master", "local[*]",
"-i", inputDir.toString(), "-inputGraphPath", inputGraphDir.toString(),
"-o", outputDir.toString() "-inputActionSetPaths", inputActionSetPaths,
"-outputGraphPath", outputDir.toString()
}); });
} }
private static void copyFiles(Path source, Path target) throws IOException { private static void copyFiles(Path source, Path target) throws IOException {
Files.list(source).forEach(f -> { Files.list(source).forEach(f -> {
try { try {
Files.copy(f, target.resolve(f.getFileName())); if (Files.isDirectory(f)) {
Path subTarget = Files.createDirectories(target.resolve(f.getFileName()));
copyFiles(f, subTarget);
} else {
Files.copy(f, target.resolve(f.getFileName()));
}
} catch (IOException e) { } catch (IOException e) {
e.printStackTrace(); e.printStackTrace();
throw new RuntimeException(e); throw new RuntimeException(e);